Most guides treat AI Overviews like a new game with secret rules. The data says otherwise: Google links at least one top-10 domain in 92.36% of AI Overviews, according to SE Ranking research.
You don’t need tricks. You need three layers: a page that ranks, answers an AI can lift, and a technical layer most sites skip. Here’s how to build all three across every client site you manage.
What are Google AI Overviews?
AI Overviews are AI-generated summaries that appear at the top of Google results, answering the query directly and citing a handful of source pages. Google builds them with its Gemini models and shows them mainly on informational and long-tail queries, almost always alongside other SERP features.
They change the click math. More searches end without a click, but a citation puts your brand at the top of the page, and Google reports that clicks from results pages with AI Overviews are higher quality, with users spending more time on the site (Google Search Central).
They also rarely travel alone. AI Overviews appear next to at least one other SERP feature 99.25% of the time, with People Also Ask present in 98.54% of cases (SE Ranking). If you report to clients on generative engine optimization, that context matters: the AIO is one feature in a crowded page, not the whole game.
Caption: An AI Overview cites three sources above the first organic result: the citation slot sits higher than position 1.
How Google picks the sources it cites
Google’s official position is short: a page only needs to be indexed and eligible to appear in Search with a snippet. Per Google Search Central, “there are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary.”
The controls that do exist are the standard snippet controls: nosnippet, data-nosnippet, max-snippet, and noindex. Those limit what Google can show from your pages, in AI Overviews and everywhere else.
One thing that does not exist: an AI Overview submission process. If a vendor offers to “submit” a client’s site to AI Overviews, that service is fiction. Send them Google’s documentation and keep your budget.
The uncomfortable truth: you still have to rank
The anchor stat: 92.36% of AI Overviews link to at least one domain ranking in the organic top 10, and 63.19% of the time they pull from pages in the top 10 for that query (SE Ranking).
The 2026 nuance: Ahrefs analyzed 863K SERPs in March 2026 and found that only 37.9% of the individual URLs cited in AI Overviews rank in the top 10 for that exact query, down from roughly 76% a year earlier (Ahrefs). The cause is query fan-out: Google splits the search into related sub-queries and cites pages that rank for those. Ranking still matters, but Google now rewards covering the whole topic, not just the head keyword.
Our position: if your page doesn’t rank in the top 10-20 for the topic, optimizing “for AI” is starting the house from the roof. First improve your site’s SEO and fix your SEO issues. The SEO retainer you already run for clients is still 80% of the AI Overview game.
How to show up in AI Overviews, step by step
Here is the full sequence. Each step gets its own section below.
Rank the page in Google’s top 10 for the query.
Answer the question in the first one or two sentences.
Structure the content with clear headings, lists, and tables.
Add relevant schema markup to remove ambiguity.
Publish an llms.txt file at your domain root.
Build author bios and brand mentions that AI systems recognize.
Track your citations and reformat the pages that miss.
Steps 1 through 3 are content work. Steps 4 and 5 are the machine-readability layer. Steps 6 and 7 are what keep citations coming after the first win.
Write answers an AI can lift
AI Overviews don’t cite pages. They cite fragments: the two sentences, the list, or the table that answers a sub-question cleanly.
Your job is to leave liftable fragments on every page. For an agency running 10+ client sites, this isn’t a page-by-page tweak. It’s an editorial standard you apply to every new piece and retrofit onto the pages that already rank.
Answer the question in the first two sentences
The pattern is answer-first: state the answer in the first one or two sentences under the heading, then expand. Definition sections that follow a “What is X” heading with an immediate answer are among the most-cited fragments in AI Overviews.
Compare the two openings for a client’s service page section titled “How much does lawn aeration cost?”:
Liftable: “Lawn aeration costs $75 to $200 for most residential yards. Price scales with lot size and soil condition.”
Not liftable: “Every lawn is different. Before we talk numbers, it’s worth understanding what aeration does for your soil, and why the timing matters…”
The second version is how most service pages are written. Three paragraphs of context, then the answer. An AI summarizer skips it and lifts the competitor’s version instead.
Structure content for extraction: headings, lists, tables
Formatting is extraction engineering. The standard to impose across every client site:
Paragraphs of 2-3 sentences, one idea each
A clean H2/H3 hierarchy that mirrors real sub-questions
Numbered lists for steps and processes
Tables for anything with two or more attributes
A key takeaways section on long pieces
This is the same formatting that wins featured snippets, and the overlap is not a coincidence. AI Overviews appear next to People Also Ask 98.54% of the time (SE Ranking), so optimizing for PAA and snippets is optimizing for AI Overviews with the same hour of work.
Target long-tail, question-based queries
Long queries are AI Overview territory. Searches of four or more words trigger an AI Overview in 60.85% of cases (SE Ranking), and question queries are their natural habitat.
The tactic: mine People Also Ask and the real questions in each client’s niche, then give each question its own H2 or H3 with a direct answer underneath. Match the phrasing to the search intent behind the question, not to the keyword tool’s phrasing. An e-commerce client’s questions (“does X fit Y”) look nothing like a local service client’s (“how much does X cost near me”), so the mining is per niche, not per template.
Does schema markup get you into AI Overviews?
Most of the pages ranking for this question say yes, and most of them sell schema tools. The honest answer has more nuance: schema is not a ticket into AI Overviews, but it is the machine-readability layer that removes ambiguity about what your content is.
We build a schema plugin, so read the evidence yourself before taking our word for it. Both directions of it.
What the data actually says
The experimental evidence is sobering. A 2026 Ahrefs test tracked 1,885 pages that added JSON-LD between August 2025 and March 2026 against 4,000 control pages, and found no meaningful citation lift on any platform: AI Overviews moved -4.6%, AI Mode +2.4%, ChatGPT +2.2% (via Search Engine Journal).
The correlational evidence points the other way. Roughly 71% of pages cited by ChatGPT and 65% of pages cited by Google AI Mode carry structured data (Stackmatix), and the same Ahrefs analysis found across 6 million URLs that AI-cited pages are about three times more likely to include JSON-LD.
Study
What it measured
What it found
What it means
Ahrefs controlled test (2026)
Citation change after adding JSON-LD
No lift vs. control group
Schema alone doesn’t earn citations
Ahrefs correlation (6M URLs)
JSON-LD presence on AI-cited pages
Cited pages ~3x more likely to carry schema
Well-run sites tend to have schema
Stackmatix (2026)
Structured data on cited pages
~71% (ChatGPT), ~65% (AI Mode)
Machine-readability travels with citations
Google Search Central docs
Requirements for AI features
No special markup required
Schema is for clarity, not admission
Correlation is not causation. Sites that implement schema also publish stronger content, earn more links, and keep their technical setup clean. Schema won’t substitute for ranking or for liftable fragments. What it does is remove ambiguity about entities: who the author is, what the product is, which question each FAQ block answers. That clarity powers rich results and supports E-E-A-T, which is why what structured data is and how to use schema markup still belong in your technical checklist.
Our position: apply schema for the complete machine-readability layer, not for a promised citation.
Applying schema across a WordPress site without writing code
Auditing schema by hand across 10 client sites is the kind of work that eats a retainer. This is the workflow we built Schemafy for.
Open WP Admin → Schemafy → Auto Schema Generator and click Scan Site. The scan analyzes every page and detects schema opportunities per page. Filter by Post Type (Post, Page, Product) and Status (Needs Schema, Partially Covered, Fully Covered), then review the suggested schemas and match percentages page by page. Prioritize Article with a real author, FAQPage, Organization, and Product. No JSON-LD to write.
For complex or custom types, the AI Schema Generator builds the JSON-LD from the page content and shows a validation status before you save.
The pitch for agency operators is repeatability: the same scan, the same filters, the same priorities, on every one of your 10+ client sites.
Caption: Auto Schema Generator flags 37 product pages missing schema after one Scan Site pass, with the suggested type per page.
[CTA_DOWNLOAD]
Make your site readable to AI systems
Beyond schema, there are two technical pieces specific to AI search that the top-ranking guides barely develop: an llms.txt file and a crawler check. Both are cheap, fast, and easy to package into a technical retainer.
Add an llms.txt file
An llms.txt file is a markdown file at your domain root that summarizes what your site is and points AI systems to your most important content.
Honest status first: llms.txt is an emerging proposal, not a confirmed signal. Google states plainly that “you don’t need to create new machine readable files, AI text files, or markup” to appear in its AI features (Google Search Central). Publish one anyway, for three reasons: it costs minutes, other answer engines and LLM crawlers can read it today, and writing it forces you to define each site’s core pages.
The manual workflow: write the summary and the key URLs in markdown, save the file as llms.txt, and upload it to the domain root through your host’s file manager or FTP. Build one template and reuse it across clients. Budget 15 minutes per site.
Check which AI crawlers actually visit your site
You can’t appear in AI answers if the crawlers behind them never read your pages. The check is manual but simple: open your server logs or your host’s raw access stats (AWStats or equivalent) and search for the AI user agents and their visit frequency. If the logs show bots burning visits on thin pages, optimize crawl budget before anything else.
Crawler
Operator
What it feeds
Googlebot
Google
The Search index, including AI Overviews
Google-Extended
Google
Gemini training and grounding (a robots.txt token, not a separate crawler)
GPTBot
OpenAI
ChatGPT training and search
ClaudeBot
Anthropic
Claude
PerplexityBot
Perplexity
Perplexity answers
Then read each client’s robots.txt. Sites inherited from a previous developer often block AI bots by accident, and unblocking one is a reportable quick win. Know the trade-off before you touch Google-Extended: blocking it limits Gemini training and grounding in some of Google’s other systems, but it does not remove you from AI Overviews in Search. That’s controlled by Googlebot access plus the snippet rules (Google Search Central). If a page is stuck earlier in the pipeline, start with crawling vs indexing instead.
Build the authority signals AI Overviews reward
Every section so far happens on the page. The last layer mostly happens off it.
Real authors. Replace “admin” bylines with a named author, a bio, and credentials on every client site. Person schema connects that author entity to the content, and it’s one of the types Schemafy generates.
First-hand data. Original numbers, real examples, and client-anonymized results are the fragments AI systems can’t find in the other nine posts on the SERP. One proprietary stat outworks a page of paraphrased advice.
Brand mentions. In Ahrefs’ analysis of 75,000 brands, brand web mentions correlate with AI Overview visibility at 0.664, against 0.218 for backlinks, nearly three times stronger (Ahrefs). Unlinked mentions now do work that link builders used to charge for, which changes how you prioritize off-page SEO signals.
Freshness. More than half of AI Overview citations come from content published in the last two years: 28.76% from 2025 and 26.85% from 2024 (SE Ranking). A refresh calendar for each client’s top pages is now a defensible retainer line.
How to track whether you’re showing up
Start with the honest limitation: Search Console does not separate AI Overview traffic. Impressions and clicks from AI features are counted inside the regular “Web” search type in the Performance report (Google Search Central), so there is no filter to pull.
What works instead:
Run manual checks on the client keywords that trigger AI Overviews, and log who gets cited.
Watch CTR on pages ranking top 10 for AIO queries. Falling CTR with stable position is the signature of an AI Overview absorbing clicks, and it’s the honest way to explain a traffic dip to a client without panic.
Those three feeds answer the only question that matters operationally: which pages to reformat first. The top-10 pages losing CTR are next sprint’s work list.
Caption: Stable position, falling CTR: the page still ranks 4-5 but the AI Overview above it is taking the clicks.
Final thoughts
Showing up in AI Overviews is not a new discipline. It’s ranking, plus fragments an AI can lift, plus an honest technical layer of schema, llms.txt, and crawler access. No secret rules, no submission form.
The concrete next step: for each client, pull the five pages already ranking in the top 10 for queries that trigger an AI Overview. Reformat them answer-first, then run a Schemafy scan to close the schema gaps the audit finds. That short list is where citations come from fastest.
[CTA_DOWNLOAD]
Frequently asked questions
How do I get my website to show up in AI Overviews?
Rank the page in Google’s top 10 for the query, answer the question directly in the first sentences, and structure the content with clear headings and lists. Google links at least one top-10 domain in over 92% of AI Overviews, so traditional SEO remains the foundation. Schema markup and llms.txt add machine-readability on top.
Does schema markup help you appear in AI Overviews?
Not on its own. A 2026 Ahrefs test found no citation lift from adding schema alone, yet most pages cited by AI systems do carry structured data. Schema removes ambiguity about your content and authors and powers rich results, so treat it as a readability layer, not a shortcut.
What triggers an AI Overview in Google?
Google shows AI Overviews when it predicts a generated summary will help, most often on informational, question-style, and long-tail queries. Searches of four or more words trigger them in roughly 61% of cases. Transactional and simple navigational queries trigger them far less often.
How long does it take to show up in AI Overviews?
There is no fixed timeline. If a page already ranks in the top 10, formatting changes like answer-first paragraphs can get it cited once Google recrawls and processes the page. If the page does not rank yet, expect the normal SEO timeline of months before AI Overview visibility follows.
Can I see AI Overview traffic in Google Search Console?
Not separately. Google counts AI Overview impressions and clicks inside the regular Web search data, with no dedicated filter. To estimate impact, track your keywords that trigger AI Overviews, monitor CTR changes on stable rankings, and use an AI visibility tool to log where your pages get cited.
Can I opt out of AI Overviews?
You cannot opt out of AI Overviews appearing on queries, but you can control how your content is used. The nosnippet and max-snippet rules limit or block what Google shows from your pages, and blocking Google-Extended affects Gemini training rather than AI Overviews in Search.
Every SEO audit template repeats the same rule: keep every page within 3 clicks of the homepage. The rule is real. The way most guides apply it is not.
Depth has a measurable cost: the deeper a page sits, the less often bots crawl it and the later it gets indexed. But “everything at 3 clicks” is not the fix. Flattening the right pages is. Here’s how to tell the difference.
Crawl depth is the number of clicks it takes a search engine bot to reach a page from your homepage. The homepage is depth 0, pages linked from it are depth 1, and so on. The deeper the page, the less often bots crawl it.
Picture a typical client store. The homepage links to a category page (depth 1), the category links to a subcategory (depth 2), and the subcategory finally links to a product (depth 3). That product needed three hops of discovery before any bot could even know it exists. Add a color variant or a paginated product grid, and you’re at depth 4 before anyone has done anything wrong.
A typical store hierarchy puts products at depth 3 before a single filter or pagination link is added.
Crawl depth vs click depth vs page depth
You’ll see three terms used as if they were one. The difference is perspective, not math.
Term
Perspective
Who uses it
Click depth
The user: minimum clicks to reach a page from the homepage
UX teams, most SEO tools
Crawl depth
The bot: link hops a crawler follows from its entry point
Log-file tools, crawl documentation
Page depth
Umbrella term for either
Casual usage in audits
In practice, tools like Screaming Frog and Sitebulb measure click depth and label it “Crawl Depth” in their reports. Operationally, the distinction changes almost nothing: you flatten architecture the same way regardless. Knowing it exists mainly saves you from confusion when a report from your team or a client’s previous agency uses the terms differently.
Why crawl depth matters for SEO
Depth hits three things in a chain: crawl frequency, then indexing speed, then rankings and traffic. Each extra level weakens the first link, and the rest follow.
It’s also a UX problem wearing an SEO costume. If a bot needs five clicks to find a page, so does a human. Buried pages don’t just get crawled less. They get visited less, linked less, and bought from less.
How depth eats your crawl budget
Googlebot enters through your homepage and sitemap, then discovers everything else by following links. Google describes its crawl capacity as crawl budget: a combination of how much it can crawl (rate limit) and how much it wants to (demand). Deep URLs are last in that queue. The budget often runs out before the crawler gets to them.
Before you panic: Google’s own guidance says crawl budget concerns sites with 1 million+ pages that change weekly, or 10,000+ pages that change daily, or sites where much of the inventory sits in “Discovered – currently not indexed”. Most client stores are under those thresholds. Their real problem is discovery: pages so deep that Googlebot rarely reaches them at all.
If budget itself is your bottleneck, that’s a separate job. We cover it in our guide to optimizing crawl budget.
Deep pages get crawled less, indexed later, and ranked worse
Botify’s log-file analyses show that Googlebot’s crawl rate drops mechanically with each level of depth (Botify). The pattern is consistent: depth 1–2 pages absorb most of the crawl activity, and everything below gets a fraction of it.
Even the shallow layers don’t get full coverage. Across 413 million pages and 6 billion Googlebot requests, Botify measured crawl ratios between 31% and 73% of indexable pages crawled in a 30-day window, depending on site type. Deep pages compete for whatever is left of that.
The consequence chain is predictable. Less crawl means Google takes weeks to see your content updates. Slow discovery means pages pile up in “Crawled – currently not indexed”.
None of this makes depth a direct ranking factor. It doesn’t need to be: a page Google rarely visits and slowly indexes underperforms either way. The mechanics of that pipeline are worth understanding, and we break them down in crawling vs indexing.
Crawl depth and AI crawlers: the %currentyear% wrinkle
Here’s the part the classic guides skip. AI crawlers like GPTBot, ClaudeBot, and PerplexityBot discover content the same way Googlebot does: by following links. They just do it with far less capacity.
The gap is not subtle. A Botify analysis of 7 billion+ log files found 18.2 billion Google crawl events against 887 million from OpenAI’s bots in the most recent measured month, roughly 4% of Google’s volume, even after OpenAI tripled its crawling since August 2025 (Botify × Chris Long, 2026). A crawler operating at 4% of Googlebot’s scale reaches far less of your deep inventory. Depth punishes you twice.
The response is the same discipline, applied harder: keep the content you want cited in AI answers at shallow levels, and expose it through discovery files that skip the click chain entirely, like your XML sitemap and the emerging llms.txt convention. Once a bot arrives, structured data helps AI systems understand what the page actually is, which is where a schema plugin like Schemafy earns its place in the stack. If AI visibility is on your roadmap, start with generative engine optimization and the answer engine optimization tools that measure it.
What is a good crawl depth?
Keep your important pages within 3 clicks of the homepage. Google has never published an official depth limit; the 3-click threshold is practitioner consensus, built on the observation that crawl activity drops sharply with each level. Important pages means the ones that earn traffic or revenue, not every URL you own.
That last sentence is the whole trick. “Important” is the load-bearing word, and it’s the one most audits ignore. A 3,000-URL store does not need 3,000 URLs at depth 3.
When deep pages are actually fine
Some pages can sit at depth 4+ without costing you anything: old blog archives, legal pages, seasonal content in the off-season, and product variants that carry a canonical tag pointing at the main product.
The operational rule for a portfolio of client stores: rank each site’s URLs by traffic and conversion potential, then flatten only the top of that list. Money pages (services, top categories, best-selling products) belong at depth 1–3. The rest can stay deep, and nobody, bot or human, will miss them. Trying to flatten everything across 10+ stores burns hours you don’t have on pages that don’t pay.
How to measure crawl depth on your site
Crawl the site with a desktop crawler. In Screaming Frog, every URL gets a Crawl Depth value, and the Site Structure view shows the distribution of URLs per level. Sitebulb draws the same picture as a crawl map. One caveat so nobody asks: Schemafy is not a crawler; this step runs on the tools above.
Then validate against reality in Google Search Console: Crawl Stats shows how Googlebot actually spends its visits, and the indexing report tells you whether deep pages are landing in “Discovered – currently not indexed”. If they are, depth is no longer a theory. It’s your diagnosis, and it usually travels with other fixable SEO issues.
The practical move: export URLs per depth level, cross them with traffic or revenue data, and flag every money page sitting at level 4+. That crossed list is your work queue.
The depth distribution is the first export to pull: anything money-page-shaped at level 4+ is a discovery problem, not a content problem.
How to reduce crawl depth: 7 fixes that flatten your architecture
Flattening architecture is mostly linking work. Here are the seven fixes, in the order most stores should apply them:
Link key pages from the homepage.
Add hub or category pages that group related content.
Fix internal links from your highest-authority pages.
Add breadcrumbs across the site.
Flatten long pagination chains.
Fix orphan pages.
Keep your XML sitemap fresh.
Fixes 1 and 2 are pure discovery shortcuts: a homepage link makes any page depth 1, and a well-built hub pulls a whole cluster up a level. Fix 7 is the safety net: a fresh sitemap gives crawlers a second door, though it never replaces internal links. The three fixes that move depth the most get their own sections below.
Fix internal linking from high-authority pages
This is the fix with the most impact on the list. Identify the pages with the most authority (backlinks and traffic; on client stores that’s almost always the homepage plus two or three categories) and link from them to the deep pages you want lifted. The math is immediate: one link from a depth 1 page makes any target depth 2.
The mechanisms are boring and effective: related-product and related-post modules, plus contextual links inside evergreen content that already ranks. Internal linking is step one of most site-wide improvements anyway, and it slots into the broader sequence in improving SEO on your website.
Use breadcrumbs (and BreadcrumbList schema)
Breadcrumbs do two jobs at once. Every page gets an upward link path (home → category → product), which gives crawlers alternative routes through the site and cuts effective depth across thousands of pages with one template change. And users get visible hierarchy, which matters on a 1,000-product store.
Add BreadcrumbList schema and Google can show the breadcrumb trail in the search result instead of a raw URL. BreadcrumbList is one of the schema types a plugin like Schemafy generates and applies in WordPress, so this fix rarely requires touching code. Implement it once per theme, and it pays across every client store running that theme.
Tame pagination, filters, and orphan pages
These are the three classic depth generators.
Long pagination first: page 14 of a blog archive sits at enormous depth by construction. Use “load more” patterns with crawlable links, or route discovery through category archives instead of endless page chains.
Faceted filters are the WooCommerce special: color × size × price combinations can mint near-infinite URL chains. Contain them with noindex or robots rules based on which facets deserve to rank. There is no absolute recipe here; the right call depends on whether the facet has real search intent behind it.
Orphan pages are the terminal case: no internal links pointing in means infinite depth for practical purposes. Detect them by crossing your crawl export against your sitemap. Anything in the sitemap that the crawler never found is orphaned.
Final thoughts
The 3-click rule survives because it’s half right. Depth genuinely suppresses crawling and indexing, and Botify’s data and Google’s own docs back that up. But the rule applies to your important pages, not your whole site. Flattening everything is effort spent on URLs nobody searches for; ranking pages by value and lifting only the top is the version of the rule that scales across a client portfolio.
The next step takes one afternoon: crawl your highest-revenue store today, export URLs by depth, cross with traffic, and flatten whatever money pages you find at level 4+. Then repeat on store number two.
[CTA_DOWNLOAD]
Frequently asked questions
What is crawl depth in SEO?
Crawl depth is the number of clicks it takes a search engine bot to reach a page starting from your homepage. The homepage sits at depth 0, pages linked directly from it are depth 1, and so on. Deeper pages get crawled less often and indexed more slowly.
What is a good crawl depth?
Keep your important pages within 3 clicks of the homepage. Google has never published an official limit, but crawl frequency drops sharply with each level of depth. Pages that drive traffic or revenue belong at depth 1–3; archives and low-priority pages can sit deeper without much harm.
Does crawl depth affect rankings?
Not directly, but the chain is real: deep pages are crawled less often, indexed later, and refreshed more slowly, which suppresses their organic visibility. Log-file analyses of large sites by Botify show pages at depth 4 and beyond receive a fraction of the crawl activity that depth 1–2 pages get.
What is the difference between crawl depth and click depth?
Click depth counts the minimum clicks a user needs to reach a page from the homepage. Crawl depth is the same distance from a bot’s perspective as it follows links. In practice, most SEO tools measure click depth and label it crawl depth, so the terms overlap.
How do I check the crawl depth of my website?
Crawl your site with a tool like Screaming Frog or Sitebulb and check the crawl depth column or visualization. Then cross-check Google Search Console’s Crawl Stats and the “Discovered – currently not indexed” report to see whether deep pages are actually being missed by Googlebot.
How do I reduce crawl depth?
Link important pages from your homepage or top-level category pages, add breadcrumbs, fix orphan pages, and flatten long pagination. Internal links from your highest-authority pages do the most work: each new link path can pull a buried page several levels closer to the homepage.
URL Rating (UR) is the 0 to 100 score Ahrefs attaches to every page, and it gets misread more than any other metric in the tool. This guide covers what UR actually measures, the three calculation rules Ahrefs documents publicly, and the one lever that raises it for free: internal links.
URL Rating (UR) is an Ahrefs metric that measures the strength of a single page’s backlink profile on a logarithmic scale from 0 to 100. It is a third-party metric, not a Google ranking factor, and it works on principles similar to Google’s PageRank (Ahrefs SEO Glossary).
Two clarifications before anything else. Google does not use UR, or any Ahrefs metric, to rank pages. And UR is strictly page-level: it scores one URL, not the domain it lives on.
That page-level focus is what makes it useful. When you need a fast read on how much link equity a competing URL has earned, through backlinks and other off-page SEO signals, UR is the quickest proxy available.
How Ahrefs calculates UR
Ahrefs doesn’t publish the exact UR formula. It does document the rules, and the calculation follows the same principles as PageRank: counting links between pages, weighting them differently, and applying a damping factor (Ahrefs’ SEO metrics glossary).
Three of those rules change how you should act on the metric.
Only dofollow links pass UR
Ahrefs states it plainly: “Only ‘dofollow’ links pass UR” (Ahrefs). Links tagged nofollow, sponsored, or UGC transfer nothing.
The practical implication: 100 nofollow mentions from press coverage won’t move your UR. Five dofollow links from strong pages will.
That doesn’t make nofollow links worthless. They still send referral traffic and help discovery. They just don’t feed this metric.
Internal links count too
Ahrefs confirms that “both internal and external backlinks are taken into account when calculating URL Rating (although ‘weighted’ differently)” (Ahrefs).
Half the articles ranking for this topic mention that in passing. Almost none draw the conclusion: internal links affect UR but do not affect DR, which makes them the only UR lever you control 100%, at zero cost.
Link from your highest-UR page to the page you want to push, and that page’s UR benefits. No outreach, no budget, no waiting on a journalist. The step-by-step is in the how to increase section below.
The scale is logarithmic
Moving from UR 10 to 20 takes far less link equity than moving from 60 to 70. Each step up the scale gets progressively harder.
The benchmarking consequence: never compare absolute deltas across different ranges. A competitor jumping from UR 55 to 60 gained far more equity than your page moving from 15 to 20, even though both moved 5 points.
URL Rating vs Domain Rating: what’s the difference?
Both metrics run on a 0 to 100 logarithmic scale. The difference is scope, and it changes what each one is for.
Aspect
URL Rating (UR)
Domain Rating (DR)
Scope
One page’s backlink profile
The whole domain’s backlink profile
Internal links
Count toward UR
Don’t count toward DR
Typical use
Sizing up a competing URL in the SERP
Vetting a domain for link building
Speed of change
Faster: page-level links and internal linking move it
Slower: needs new referring domains
Can one exceed the other?
A viral page’s UR can top its domain’s DR
A DR 80 site can host UR 0 pages
The table’s last row deserves emphasis because it trips people up in both directions. A DR 80 domain routinely publishes new pages that start at UR 0 to 5: domain authority is not inherited per page.
The inverse also happens. A page that goes viral and attracts dozens of dofollow links of its own can carry a UR higher than its domain’s DR. Free tools and data studies do this constantly.
In practice, SEOs use UR to evaluate specific competing URLs in a SERP and DR to qualify domains during link prospecting. Different questions, different metric.
What is a good UR score?
As a rough rule: UR 20+ is decent, and UR 40+ indicates a strong page-level backlink profile.
Treat those numbers as loose calibration, not targets. The only benchmark that matters is the UR of the pages ranking in the top 5 for your keyword. If the top 5 sits at UR 8 to 15, a UR 25 page is heavily armed. If the top 5 sits at UR 60+, UR 40 is underpowered.
No UR value guarantees a ranking. The metric is relative by design, the same argument Ahrefs makes about judging DR in its own documentation. Check your SERP first, then decide whether the page needs more equity.
Does URL Rating affect Google rankings?
No. UR is not a ranking factor. Google doesn’t consume Ahrefs metrics, and no third-party score sits in the algorithm.
But the correlation is real and documented:
“Our studies show that the URL Rating has a clear positive correlation with Google rankings.” (Ahrefs SEO Glossary)
The explanation is unglamorous. UR approximates what PageRank measures, so pages with strong link equity tend to score high on both. Ahrefs is careful with the same nuance on its authority studies, noting its 218,713-domain correlation study “does not prove causation.” Even Google’s John Mueller, while denying “domain authority” as a factor, has acknowledged a sitewide score that “maps to similar things” (via Ahrefs).
So read it this way: links are the cause, UR is the summary. Build the links, and both UR and your organic traffic tend to follow.
How to check the UR of any page
If you pay for Ahrefs, paste any URL into Site Explorer and UR appears at the top of the Overview. For competitor spying, the SERP overview inside Keywords Explorer lists the UR of every top-ranking result for a keyword, which turns benchmarking the top 5 into a ten-second job.
Without a subscription, the Ahrefs SEO Toolbar browser extension paired with a free Ahrefs account shows DR and UR for any page you visit. Note that Ahrefs’ free Website Authority Checker returns domain-level DR, not page-level UR.
How to increase your URL Rating
There are two levers: one free (internal links) and one that costs time or money (backlinks). Internal linking comes first here deliberately, because it’s the one you can pull this afternoon, ideally as part of a broader plan to improve SEO on your website.
Strengthen your internal linking
Since internal links pass UR, the most actionable fix is redistributing equity you already own:
Open the Best by links report in Ahrefs Site Explorer, sorted by UR, to find your strongest pages.
Add contextual dofollow links from those pages to the page you want to push.
Keep the target page within 3 clicks of the homepage.
Re-check UR after the next Ahrefs crawl update.
Pages buried 4+ clicks deep receive little internal equity and are slower to get discovered, a problem that overlaps with crawling and indexing issues. Contextual links from body copy beat footer or sidebar links, which is exactly the pattern the weighting rewards.
Caption: Best by links sorted by UR reveals which pages have equity to spare: the top three here are the donor pages for internal links.
Earn quality backlinks to the page
The paid-in-effort lever. The tactics that reliably earn dofollow links to a specific page are the unglamorous ones: linkable assets (original data, free tools, reference guides), digital PR, and broken link building.
The evidence favors focusing here: across roughly 920 million pages studied, Ahrefs found the number of referring domains to a page is the strongest correlating backlink factor for rankings (Ahrefs).
One filter before you celebrate any placement: only dofollow links pass UR. A nofollow mention in a major publication may be great for the brand, but it won’t move this metric. No tactic here works on a predictable timeline, and anyone promising one is selling something.
Fix broken links and redirect chains
Equity leaks. When links point at 404s, or pass through chains of redirects before reaching the destination, part of the value never arrives at the page.
Crawl the site with Ahrefs Site Audit (or your crawler of choice), consolidate redirect chains into single 301s, and repoint internal links straight at final URLs. It’s the same hygiene that helps you fix common SEO issues and optimize crawl budget, and it protects the equity you already earned.
High UR is only half the battle: make Google understand the page
UR tells you how much authority a page has accumulated. It says nothing about whether search engines understand what the page is. That job belongs to structured data: the JSON-LD markup that tells Google and AI engines whether your URL is a product, a review, an FAQ, or a how-to, and makes it eligible for rich results and AI citations.
A page with UR 45 and no schema competes as one more blue link.
If your pages run on WordPress, Schemafy generates valid JSON-LD across 16 schema types without writing code, and you can verify the output with Google’s Rich Results Test. Pair the authority you’ve built with markup that machines can parse; how to use schema markup covers the implementation side. Learn more at schemafy.net.
FAQs about URL Rating
Is URL Rating a Google ranking factor?
No. URL Rating is a third-party metric created by Ahrefs, and Google doesn’t use it. However, UR approximates the same link signals Google’s PageRank measures, which is why Ahrefs’ studies show a clear positive correlation between higher UR and better organic rankings.
What is the difference between UR and DR?
UR (URL Rating) measures the backlink strength of a single page, while DR (Domain Rating) measures the strength of an entire domain’s backlink profile. Internal links affect UR but not DR, so a strong site can still have individual pages with very low UR.
What is a good URL Rating?
As a rough benchmark, a UR of 20+ is decent and 40+ indicates a strong page-level backlink profile. The more useful benchmark is competitive: check the UR of the pages ranking in the top 5 for your target keyword and aim for that range.
Why is my page’s UR higher than my site’s DR?
Because they measure different things. A single page can attract many strong dofollow backlinks (a viral post, a free tool) while the rest of the domain has few. Internal links also boost UR without touching DR, widening the gap further.
How can I increase URL Rating for free?
Through internal linking. Find your highest-UR pages in Ahrefs (Best by links), then add contextual internal links from them to the page you want to strengthen. Only followed links pass UR, and internal links are the one lever you fully control at zero cost.
Final thoughts
UR is a thermometer for page-level link equity: read it against the top 5 for your keyword, never as an absolute number, and remember that the cheapest way to raise it is the internal linking lever most sites ignore.
The next step takes ten minutes. Pull the UR of the pages outranking you, compare, and if the gap is closable, start moving internal links from your strongest pages this week. And once the authority is there, make sure Google actually understands the page: with a tool like Schemafy, adding the structured data layer takes minutes, not a sprint.
Most guides treat a Google Merchant Center account as a form you fill once. Create it, upload a feed, done in 20 minutes.
The account is the easy part. Keeping products approved is the real work, because Google crawls your product pages and checks that your feed, your landing page, and your structured data all say the same thing. That check is where most new WooCommerce stores fail. This guide covers both halves.
What is a Google Merchant Center account?
A Google Merchant Center account is a free account where you upload and manage the product data Google shows across Search, Shopping, Maps, and YouTube. It is the data source behind both free product listings and paid Shopping ads, not an advertising channel by itself.
Think of it as the layer between your store and every Google surface that displays products. Your WooCommerce catalog feeds Merchant Center, and Merchant Center feeds Google. It is also where Google checks that your prices, availability, and images match what shoppers actually see on your site. Google’s own Get started with Merchant Center doc covers the basics; this guide covers the WooCommerce specifics it skips.
Is Google Merchant Center free?
Yes. Creating the account, uploading products, and appearing in free listings costs $0, per Google’s documentation.
You only pay if you run Shopping ads, and that billing happens in a linked Google Ads account, not in Merchant Center. The two get confused constantly: Merchant Center holds your product data, Google Ads spends your budget. You can use the first without ever touching the second.
Free listings vs Shopping ads: what the account actually gets you
One account powers two very different distribution channels. Here is how they compare:
Free listings
Shopping ads
Where they appear
Shopping tab, Search, Maps, YouTube
Shopping tab, Search results, partner sites
Cost
$0
Pay-per-click budget
Feed requirements
Same product data spec
Same product data spec
Google Ads account
Not required
Required (linked)
For a WooCommerce store with 100 to 1,000 products, free listings alone justify the account. Your products become eligible to appear in the Shopping tab with zero ad spend, and you can add paid campaigns later for the products that earn impressions. If you manage stores for clients, this is also the cleanest pitch for setting up Merchant Center on day one, before there is any ad budget to argue about.
What you need before creating your account
Have these five things ready before you start. Missing any of them causes rejections later:
A Google account (one Merchant Center account per Google account)
An active website with product pages shoppers can buy from online
A visible returns policy and contact information
A working checkout with SSL
Product data on the site that matches what you plan to submit
One warning from Google’s setup documentation deserves bold type: the country you select during account setup cannot be changed later. If you sell across markets, or you are setting up accounts for clients in different countries, choose deliberately. Fixing a wrong country means creating a new account from scratch.
How to create a Google Merchant Center account, step by step
The four steps below take roughly 20 minutes in the current interface (Merchant Center Next). If you run an agency, budget that 20 minutes per client store: the process repeats identically for every domain.
Caption: Merchant Center Next onboarding asks for business details first. The name you enter here appears on your listings.
Step 1: Sign in and add your business details
Go to merchants.google.com and sign in with the Google account that should own the store’s data. One Google account can create one Merchant Center account, per Google’s documentation, so agencies should use the client’s account rather than their own. You will enter a business name, country, and time zone.
The business name is not cosmetic. It appears on your listings, and Google cross-references it against your website. Use your brand name exactly as it appears on your site. A store called “Maple & Main” on the site and “MM Retail LLC” in Merchant Center reads as an inconsistency, and inconsistencies feed the misrepresentation checks covered later in this guide.
Step 2: Verify and claim your website
Verification proves you control the domain. Google’s Configure your accounts doc lists four methods: an HTML tag added to your homepage, Google Tag or Google Analytics if already installed, Google Search Console, or an HTML file upload.
For WordPress stores, Search Console is the shortcut. If your site is already verified there, Merchant Center picks up the verification with one click and nothing new touches your theme. Otherwise, the HTML tag goes in your site’s head section.
Two details trip people up. First, verifying and claiming are separate actions: verification proves control, claiming binds the domain to this specific Merchant Center account. Complete both. Second, only one account can claim a domain at a time, which matters the moment an agency and a client both create accounts.
Step 3: Configure shipping, returns, and tax
These are the settings that silently block listings when they are missing or wrong. Configure shipping with the rates you actually charge, not optimistic estimates. Add your return policy and link it to the policy page on your site. United States sellers also configure tax settings here.
Google compares the shipping cost a shopper sees at your checkout with the cost you declared in Merchant Center, as part of its landing page requirements. A $4.99 declared rate against a $7.99 checkout rate is a mismatch, and mismatches become disapprovals. When in doubt, declare the higher rate.
Step 4: Add your WooCommerce products
This is where the generic guides go Shopify-first and leave WooCommerce owners guessing. You have three ways to get products into Merchant Center:
A file or Google Sheets feed built to the product data specification, maintained by hand or by a feed plugin.
Manual entry in the Merchant Center interface, viable only for catalogs under a couple dozen products.
For a store with 100 to 1,000 products, use the extension. Then audit the attributes Google requires before you sync: id, title, description, link, image_link, price, availability, brand, and GTIN or MPN. The extension maps most of these from WooCommerce fields automatically, but only if those fields are filled. Empty brand and GTIN fields are the single most common gap, and they matter enough to get their own section below.
How Google verifies your product data (and where stores fail)
Submitting a feed is not the finish line. After you sync, Google crawls your product landing pages and compares what it finds against what you submitted. Price, availability, images, all of it. If you want the background on how that crawling works, we cover it in how Google crawls your pages. Large catalogs should also make sure those recrawls aren’t wasting crawl budget on duplicate or filtered URLs.
This creates a triangle that has to stay consistent: the feed, the visible page, and the page’s structured data. Most disapprovals for new stores trace back to one corner of that triangle drifting from the other two.
Caption: Google checks all three corners. A price change that reaches one corner but not the other two becomes a disapproval.
Feed, landing page, and Product schema must tell the same story
Google does not only read your page visually. Its structured data documentation for Merchant Center states that markup must be present in the HTML returned by the server, cannot be generated by JavaScript after the page loads, and must match the values shown to the user. Google recommends JSON-LD and uses schema.org markup to read price, availability, and product identifiers directly from your pages.
The technical requirements are specific. Your product page should carry one Offer, or, if it lists several, each offer needs a SKU or GTIN that matches the feed. The Offer itself must include price, priceCurrency, and availability. If you are new to this layer, start with what structured data is. You can also build a valid Product block by hand with a free schema markup generator.
This is where a schema plugin earns its place in an e-commerce stack. A plugin like Schemafy generates the Product schema with its offers by reading price, stock, and ratings directly from your WooCommerce fields, so the markup updates when the product updates and never tells Google a different story than your feed. You can monitor how Google reads that markup in Search Console’s Merchant Listings report.
GTIN, MPN, and Brand: the identifiers that gate your listings
Three identifiers decide how well Google matches your products. GTIN (the global barcode number), MPN (the manufacturer’s part number), and Brand. The rule from Google’s supported attributes documentation: submit the GTIN if the product has one; if it does not, submit MPN plus Brand as the fallback.
Skip them and you pay three times. Matching gets worse, so your products land next to the wrong competitors. You lose eligibility for enriched listings. And in strict categories, missing identifiers become outright disapprovals.
The WooCommerce-specific failure is quieter: WooCommerce has shipped a native “GTIN, UPC, EAN, or ISBN” field on the product Inventory tab since version 9.2, supported by the Google for WooCommerce extension, for both simple products and variations. Most stores never fill it. Those same fields are what your Product schema should expose, and Schemafy maps them into the markup without manual editing. For the broader context on markup strategy, see our schema markup guide.
Common Merchant Center account problems and how to fix them
Three problems cover most of what goes wrong with a new account: item disapprovals, misrepresentation suspensions, and lost website claims. Each has a distinct symptom, cause, and fix. The debugging mindset is the same one you would use to fix common SEO issues: find the mismatch, correct the source, request re-review.
Caption: Products > All products is the first place to look. Disapproved items list the specific policy reason when clicked.
Item disapprovals: price and availability mismatch
Symptom: products flip to “Disapproved” in Products > All products, usually citing a price or availability mismatch. Per Google’s product status documentation, reviews take 3 to 5 business days for Shopping ads and can take weeks for free listings, so each disapproval-fix cycle is expensive.
Cause: in WooCommerce stores, the usual suspect is a page that no longer matches the feed. A cache or CDN serving yesterday’s price. A sale that started in WooCommerce but has not reached the crawled page. Variations whose prices differ from what the schema reports.
Fix: let the extension handle feed sync (it re-syncs on product edits automatically), purge your cache every time prices change, and run the affected URL through Google’s Rich Results Test to confirm the schema on the live page shows the same price and availability as the feed.
Misrepresentation suspensions
This is the account-level suspension that scares store owners most, partly because Google explains it least.
Cause: Google cannot verify the business is what it claims. Common triggers are missing contact information, missing or thin policy pages, a new domain with no external signals, and inconsistencies between feed, site, and checkout. Google’s guidance on keeping products approved points repeatedly at data consistency and transparency.
Fix: complete your business identity in Merchant Center, make contact details and policies visible on the site, and align your branding across feed, storefront, and checkout. Then request review and wait it out. Be realistic: reviews are slow, repeated failed reviews make things worse, and nobody can honestly promise fast reinstatement. Agencies onboarding brand-new client domains should set that expectation before the first sync, because new domains with no history trip these checks most often.
Website claim lost or verification failed
Symptom: Merchant Center reports the website is no longer verified or claimed, and listings stop.
Cause: the HTML verification tag was deleted during a redesign or theme change, a redirect chain broke the homepage check, or another Merchant Center account claimed the domain.
Fix: re-verify through Search Console, which survives theme changes because it is tied to the property rather than a tag in your template. If another account holds the claim, resolve ownership before fighting Google about it. For agencies this is a contract question as much as a technical one: decide up front whether the agency’s account or the client’s account owns the claim, and document it.
What changed in Merchant Center in 2026
The product data spec gets an annual update, and the 2026 round, published by Google, has three changes worth acting on.
Image quality became enforceable policy. The minimum resolution for image_link and additional_image_link rose to 500×500 pixels across all categories, with warnings running since April 14, 2026 and enforcement starting January 31, 2027. Audit your thumbnails now, before warnings become disapprovals.
Video became a real feed asset. Since June 30,2026, videos submitted through video_link are eligible to serve, and Google validates them for policy and quality. A failed video blocks the video, not the offer.
Shipping got product-level controls, including a handling cutoff time and a minimum order value attribute.
The direction is consistent: Google keeps tightening how product data is validated while Shopping surfaces expand into AI experiences. Consistent, machine-readable product data stops being a nice-to-have and becomes the entry ticket.
Final thoughts
Creating a Google Merchant Center account takes 20 minutes. What keeps your products live afterward is consistency: the feed, the landing page, and the Product schema telling Google the same price, the same stock, the same brand, every day.
If you are starting today, the order matters. Create the account, verify through Search Console, connect the Google for WooCommerce extension, and fill in GTIN and Brand on your products before you sync anything. Stores that do the identifier work first skip most of the disapproval cycle entirely. If you manage multiple client stores, turn that order into your standard onboarding checklist, because every shortcut taken during setup resurfaces later as a policy flag.
[CTA_DOWNLOAD]
Frequently asked questions
What is a Google Merchant Center account?
A Google Merchant Center account is a free account where you upload and manage the product data Google shows across Search, Shopping, Maps, and YouTube. It feeds both free product listings and paid Shopping ads, and it is where Google checks that your prices, availability, and images match your website.
Is a Google Merchant Center account free?
Yes. Creating the account, uploading products, and appearing in free listings cost nothing. You only pay when you run Shopping ads through a linked Google Ads account. For many WooCommerce stores, free listings alone justify setting up Merchant Center even with no ad budget.
Do I need a website to use Google Merchant Center?
Yes, in practice. Google requires an active website with product pages you verify and claim during setup, and it crawls those pages to confirm your feed data matches. Your site also needs visible contact information, a returns policy, and a secure checkout to stay in good standing.
Do I need Google Ads to use Merchant Center?
No. Free product listings work through Merchant Center alone. You only need a linked Google Ads account if you want to run paid Shopping campaigns. Many stores start with free listings, measure which products get impressions, and add paid campaigns later for the winners.
How long does Merchant Center verification take?
Website verification is usually instant if you use Google Search Console or a tag already on your site. Product review takes 3 to 5 business days for Shopping ads and can take a few weeks for free listings, and account-level reviews such as a misrepresentation check can take longer still.
Can I change my Merchant Center account country later?
No. The country you select during account setup cannot be changed afterward, so choose carefully if you sell across markets. You can add additional target countries for your feeds, but the account’s base country is permanent; changing it means creating a new account.
Five numbers written down before any changes are made. That snapshot is what turns the traffic went upinto this change worked.
You have probably read five different lists of SEO tips. The problem was never the tips. It was that nobody told you what order to do them in.
This guide on how to improve SEO on your website is sequenced by dependency, not popularity. Publishing new content does nothing if Google cannot index the site.
In 2026 that sequence ends somewhere new: not just ranking, but getting cited by AI systems. A copyable checklist waits at the end.
How to Improve SEO on a Website (Short Answer)
To improve SEO on a website: fix crawling and indexing issues, match each page to search intent, optimize titles, headings and internal links, refresh existing content, add schema markup, earn authority signals, and track results in Search Console. Work in that order, because later steps depend on earlier ones.
Each step assumes the one before it is done. Skipping ahead is how sites end up with beautifully optimized pages that Google never indexed.
Step 1: Get an SEO Baseline Before You Change Anything
Before you touch a single title tag, write down five numbers:
Clicks over the last 28 days (Search Console)
Impressions over the last 28 days (Search Console)
Average position across the whole property
Pages indexed versus pages not indexed
Core Web Vitals status: how many URLs are Good, Needs Improvement, and Poor
This takes about fifteen minutes and it saves months of guessing.
Without a snapshot you cannot prove a gain was yours, and you cannot tell a drop you caused from a drop that came with a core update. When a client asks why traffic fell in week six, “here is what the site looked like in week one” is the only answer that holds up.
Everything you need is free: Google Search Console, Google Analytics 4, and PageSpeed Insights. Save the five numbers in a dated row in a spreadsheet. You will compare against that same row in step 12.
Step 2: Fix Your Technical SEO Foundation
The best page on the internet does not rank if Google cannot crawl it, index it, or load it. Technical SEO is not the part that wins you rankings. It is the part that makes every other step possible.
Three checks handle most of it: what Google has actually indexed, how fast the site loads, and whether duplicate URLs are splitting your signals.
Check What Google Actually Has Indexed
Open Search Console and go to Indexing > Pages. The first number to read is not clicks. It is indexed versus not indexed.
Having excluded pages is normal. Not knowing why is the problem. Open each exclusion reason and match it to an action.
Exclusion reason
What it means
What to do
Crawled, currently not indexed
Google saw the page and chose not to index it
Deepen the content or fix the intent match, or consolidate it into a stronger page
Discovered, currently not indexed
Google knows the URL exists but has not crawled it
Add internal links and reduce how many clicks it sits from the home page
Duplicate without user-selected canonical
Equivalent versions exist and Google picked one for you
Add a self-referencing canonical tag (next section)
Excluded by ‘noindex’ tag
A noindex directive is on the page
Confirm it is intentional; if not, remove it
Redirect error
The redirect chain fails or loops
Fix the chain and leave a single hop
Blocked by robots.txt
A robots.txt rule stops the crawl
Review the rules and unblock anything Google needs to render the page
Two housekeeping tasks close this out. Submit your XML sitemap in Search Console so Google has a clean list of the URLs you care about. Then open robots.txt and confirm it is not blocking CSS or JavaScript, because Google renders pages the way a browser does. This is the same diagnostic order Semrush recommends before touching content.
The exclusion reasons matter more than the totals. Each one maps to a different fix.
Improve Page Speed and Core Web Vitals
Three metrics carry the weight, and each one is simpler than its acronym suggests. LCP measures how long the main element takes to appear. INP measures how long the page takes to respond when someone interacts with it. CLS measures how much the layout jumps around while loading.
Fix them in this order, because that is roughly the order of impact:
Compress images and serve them in WebP with lazy loading.
Delete plugins you are not actively using.
Turn on caching.
Minify CSS and JavaScript.
Put a CDN in front of the site.
Check how heavy your WordPress theme is before you blame anything else.
Run PageSpeed Insights on the mobile tab, not desktop. Google indexes the mobile version of your site first, so the desktop score is the one that does not count.
One honest note, because most guides skip it. Speed is a tiebreaker, not a multiplier. Fixing Core Web Vitals will not move you ten positions on its own.
What it does do is stop you losing to an equally good page that loads faster, and stop visitors leaving before they read anything. Both are worth having. Neither is a ranking strategy.
Clean Up Duplicates With Canonical Tags
The same content is usually reachable at more URLs than you think. With and without a trailing slash. With UTM parameters attached from a paid campaign. Across paginated versions. On http and https.
Google sees each variation as a candidate and picks one. It does not always pick the one you wanted.
That is a self-referencing canonical: the page names itself as the version to index. For the full syntax and the CMS-specific steps, see how to add a canonical tag.
Two things to know. A canonical is a signal, not a command, so Google can override it. And you can check whether it did: if Search Console reports “Duplicate, Google chose different canonical than user” for a URL, your tag was read and ignored, which usually means the two pages are more similar than you assumed.
Step 3: Match Every Page to Search Intent
The most common reason an optimized page does not rank has nothing to do with optimization. The page answers a different question than the one the search results reward.
There are four intents behind any query:
Informational: the searcher wants to understand something. “what is schema markup”
Navigational: the searcher wants a specific site or page. “search console login”
Commercial: the searcher is comparing options before buying. “best crm software”
Transactional: the searcher is ready to act. “buy standing desk”
Diagnosing intent takes two minutes and no tools. Search your keyword in an incognito window and look at what format dominates the top ten: a long guide, a listicle, a product page, a comparison, a video. That format is Google telling you what it has already decided the query deserves.
Then match it before a single word gets written.
“Best crm software” is the clearest example. The intent is commercial, and every result on page one is a comparison of multiple tools. A product landing page for one CRM will not rank there no matter how well it is written, because it answers a question nobody asked. If you want the full breakdown of the four types of search intent with more examples, that is covered separately.
For anyone briefing writers, this check belongs before the brief, not after the draft. Format decisions made after the writing are expensive.
Step 4: Tighten Your On-Page SEO
On-page is the only group of signals you control completely, and it is the fastest to fix. No third party has to agree, no algorithm has to recrawl a link graph.
Three areas cover almost all of it: what shows in the search result, how the page is structured, and how its URLs and images are named.
Title Tags and Meta Descriptions
The title tag rules are short. Keep it to 60 characters or fewer so it does not get truncated. Put the keyword near the front. Make it unique across the site.
And write it for the person deciding whether to click, not for the crawler.
Meta descriptions follow the same logic at 155 characters or fewer: state the benefit, then give a reason to click.
Here is the nuance almost nobody spells out. The meta description is not a ranking factor, and Google rewrites it most of the time anyway. Neither fact is a reason to skip it.
When Google keeps yours, it is the only sentence of sales copy you get in the search results. When it rewrites yours, it pulls from your page copy, so a page written with a clear benefit still wins.
Write it, then check how it renders before publishing. A free SERP simulator shows you the truncation point on both mobile and desktop, which is faster than publishing and squinting at the live result. If you are implementing these on WordPress, the mechanics of adding meta descriptions in WordPress are covered step by step.
Across a portfolio of client sites, this is also the one on-page task that batches well. Titles and descriptions can be rewritten in bulk without touching a single page’s content.
Heading Hierarchy (H1, H2, H3)
The rules fit in four lines. One H1 per page, containing the primary keyword. H2s for sections. H3s for subsections inside them.
Never skip a level.
The rule people break most often is using headings for styling. If a line is bold and large because it looks good, it should be styled, not marked up as an H3.
This matters more now than it did three years ago. AI systems extract answers block by block, and a clean hierarchy is what makes a single section quotable on its own. A page with a broken heading structure forces the model to guess where an idea starts and stops, and a guess is easy to skip in favor of a competitor who made it obvious.
Correct:
H1: How to Improve SEO on Your Website
H2: Fix Your Technical Foundation
H3: Check What Google Has Indexed
H3: Improve Page Speed
H2: Match Every Page to Search Intent
Broken:
H1: How to Improve SEO on Your Website
H3: Check What Google Has Indexed
H2: Fix Your Technical Foundation
H4: Improve Page Speed
The second version says the same things. It just makes the relationships unreadable.
URL Slugs and Image Alt Text
Slugs should be short, lowercase, hyphenated, and built around the keyword. Drop stop words. Drop dates, because a slug with 2024 in it ages badly and gives you a reason to break a URL later.
The warning that matters: never change the slug of a page that already ranks without putting a 301 redirect in place. You reset the signals that URL accumulated, and rebuilding them takes months. If you want the full set of rules, we covered what an SEO slug is and how to change one safely.
Alt text has one job: describe the image for someone who cannot see it. That is it. Keyword stuffing alt attributes is a habit left over from 2012 and it helps nothing.
Every informative image needs alt text. Decorative images, the ones that carry no information, take an empty alt attribute so screen readers skip them instead of reading a filename aloud.
While you are in there, name the file before uploading. blue-cotton-tshirt-front.jpg is worth more than IMG_4471.jpg, and it costs three seconds.
Step 5: Update and Consolidate Your Existing SEO Content
This is the highest-return step in the guide, and it is the one most people skip in favor of publishing something new. The reason it works is simple: you are improving pages Google has already crawled, already indexed, and already assigned some level of trust.
Part A: refresh what is close.
In Search Console, filter your queries to positions 8 through 20. These are the queries one push away from the part of page one people actually click. Find the page receiving each one, then do four things to it:
Update figures, screenshots and references that have gone stale.
Add the subtopics the current top five cover and you do not.
Fix the intent match if the format is wrong for the query.
Change the publish date only if the content genuinely changed.
That last one is not a formality. Updating a date without updating the content is a well-known trick and it does not work.
Part B: consolidate what competes.
Keyword cannibalization is when several of your own URLs compete for the same query, and it is more common on older sites than anyone expects. To find it, open Search Console, filter by a specific query, then check the Pages tab. If more than one of your URLs is pulling impressions and clicks for that query, you have cannibalization.
The fix is uncomfortable but mechanical. Pick the strongest page, merge whatever is genuinely useful from the others into it, and 301 the rest.
Two mediocre pages competing for one query split every signal they have earned, internal links included. One consolidated page concentrates them. You are not deleting work, you are stopping it from competing with itself.
Step 6: Target Low-Competition Keywords and Build Topic Clusters
Two ideas that only work together.
Low-competition keywords first. Google Keyword Planner is free and gets you started. If you have Ahrefs or Semrush, filter for keyword difficulty under 30 with volume at or above 100. Those thresholds are not magic numbers, they are a starting filter that keeps you out of fights you cannot win yet.
A young site should not attack head terms. The pages ranking there have years of accumulated links and topical history behind them, and you will spend six months learning that. Long-tail queries convert better anyway, because the searcher has already narrowed what they want.
Topic clusters second. A cluster is one broad pillar page plus 10 to 30 subtopic pieces, where every subtopic links up to the pillar and across to its siblings. The structure is what builds topical authority, and topical authority is what separates a site that ranks from a site that publishes.
For a WordPress site selling a WooCommerce theme, a cluster might look like this. Pillar: a complete guide to WooCommerce product pages. Subtopics: product schema markup, product image optimization, variation handling, category page structure, and product page speed.
Each one is a real query someone searches. Together they tell Google you are not guessing about WooCommerce product pages.
Publishing five connected pieces on one subject beats publishing five disconnected pieces on five subjects, and the gap widens over time. More strategies for increasing organic traffic build on the same principle.
Step 7: Strengthen Your Internal Linking for SEO
Internal linking is the only authority lever you fully control. Backlinks depend on someone else saying yes. Internal links depend on you opening an editor.
Five rules cover it:
Link from your highest-authority pages to the pages you want to lift, not the other way around.
Use descriptive anchor text that names the destination. Never “click here” or “read more.”
Give every important page at least 3 to 5 internal links pointing at it.
Find your orphan pages, the ones with no internal links at all. Google barely crawls them and users never find them.
Keep key pages within three clicks of the home page.
You can audit this without paying for anything. Search Console has a Links report that lists your internally most-linked pages in descending order. The pages you consider important that do not appear near the top of that list are the ones your site structure is quietly ignoring.
Fixing internal links costs an afternoon and no approvals. That combination is rare enough in SEO that it deserves to be higher on most people’s list than it is.
Step 8: Add Schema Markup So Search Engines and AI Can Read Your Pages
Schema markup, written in a format called JSON-LD, is a layer of structured data that tells Google and AI models what each thing on your page actually is.
This is a product. This is its price. This is the author. This is a review, and this is its rating.
Without that layer, machines infer all of it from your text. They are decent at inferring, and “decent” is not what you want deciding whether your product’s price shows up in a search result.
Three reasons it earns a place in this list, and one honest caveat.
It makes you eligible for rich results. Star ratings, FAQ dropdowns, prices, breadcrumbs. These raise click-through rate without your position changing at all. It is the only lever in this guide that improves the outcome without moving the ranking.
It makes you readable to AI search. AI Overviews are no longer an experiment. How often they appear depends heavily on who is measuring: Conductor’s analysis of 21.9 million searches in the first quarter of 2026 put the figure at 25.11%, while trackers measuring through March 2026 report figures approaching 48%. The spread comes from different keyword sets, different geographies, different sample sizes and different measurement dates, which is why a single number is the wrong thing to quote.
What is consistent is the gap between being cited and not being cited. Seer Interactive, analyzing 5.47 million queries across 53 brands between January 2025 and February 2026, found that pages cited in AI Overviews get 120% more clicks per impression than uncited pages, though they still trail pages on AI-free result pages by 38%.
It disambiguates entities. A model needs to know that “Apple” in your article is the company, that the review belongs to the product and not the page, that the author is a person with credentials. Structured data states all of that explicitly. Explicit is what a model needs before it cites you with confidence instead of citing someone clearer.
Now the caveat, because it gets overstated constantly. Schema markup is not a direct ranking factor. It improves understanding, it makes you eligible for rich results, and it makes you easier to cite. It does not move you up the page by itself.
For agencies, schema has one property nothing else in this guide has: it standardizes at the template level. One correct Article template applies to every post on a site, and the same template can be replicated across a portfolio without rewriting a word of content. Our full guide on how to use schema markup goes deeper on each type.
The top two results are not ranked higher because of schema. They just take up more of the screen and answer more of the question before the click.
Which Schema Types to Add First
This list is ordered by return, not alphabetically. If you only get to three, do the first three.
Schema type
When it applies
What it can earn you
Organization (or LocalBusiness)
On the home page of any site. Use LocalBusiness instead if there is a physical address
Knowledge panel eligibility, logo and contact details in search
Article
On every blog post
News and article carousel eligibility, author attribution
Product + Review
On every ecommerce product page
Price, availability and star ratings in the result
FAQPage
On pages with real questions visible to the reader
Expandable questions under your result
BreadcrumbList
Site-wide
A readable navigation path instead of a raw URL
Person
On author pages and bios
Reinforces the authorship signals covered in step 11
One rule overrides all of them: only mark up what is visibly on the page. Google’s structured data guidelines are direct about it. Do not add structured data about information that is not visible to the user, even if the information is accurate, and violating a quality guideline can stop syntactically correct markup from producing a rich result or get it flagged as spam.
The temptation is real, especially with reviews and prices. It is also the fastest way to lose rich result eligibility across an entire site.
How to Add Schema in WordPress Without Code
Three routes, with honest trade-offs.
Route
Control
Effort per page
Scales to a whole site?
Hand-written JSON-LD in the <head>
Total
High
No
Your SEO plugin’s schema module
Low
Low
Partially
A dedicated schema plugin
High
Low
Yes
Route 1, writing JSON-LD by hand. Total control, and a good exercise once. It also breaks the moment someone edits a template, and nobody maintains it across 400 product pages.
Route 2, your SEO plugin’s schema module. Yoast, Rank Math and AIOSEO all ship one. They cover the basics competently. They also support a small number of types and give you limited control over individual fields, which is fine until you need Product with variations or a Person with real credentials.
Route 3, a dedicated schema plugin. This is the category built for the problem: generate JSON-LD per template, inject it automatically, manage it in one place.
Schemafy is one option in that category. It includes a visual builder covering 16 Schema.org types, AI-assisted generation that runs on your own ChatGPT or Claude API key, a manual JSON editor for the edge cases the builder does not cover, a site scan that finds pages missing schema with filters by content type and bulk selection, automatic handling of WooCommerce products, and a single screen where every applied schema can be reviewed and edited.
The objection worth answering directly, because it stops most agencies from testing anything: you do not have to uninstall your current SEO plugin. Compatibility is declared with Yoast SEO, Rank Math, AIOSEO and WooCommerce, which means this is a change you can trial on one client site without renegotiating the stack on the other eleven.
A site scan turns “we should probably add schema” into a list of 412 specific pages and the type each one needs.
Validate Before You Move On
Run every template through two validators: Google’s Rich Results Test and the Schema Markup Validator at schema.org. Templates, not pages. If the Article template is correct, all 300 posts using it are correct.
Then wait. One to two weeks after deployment, open the Enhancements section in Search Console. That is when Google has reprocessed enough of the site to tell you what it actually accepted, which is not always what the validator approved.
Fix errors before warnings. Errors block the rich result entirely. Warnings are recommended fields you left empty, which reduce your chances without eliminating them.
The most common failure has nothing to do with syntax. It is duplicate schema: your SEO plugin and your schema plugin both emit an Article block for the same page, and Google gets two conflicting descriptions of one thing. If a validator shows two of the same type on one URL, turn one of them off before you debug anything else.
Step 9: Win Featured Snippets and AI Overview Citations
Featured snippets look like luck. They are closer to a format-matching exercise, and the process repeats across every client you have.
Find the queries where you already rank in the top 10 but do not hold the snippet. Search Console gives you the ranking queries; a manual search tells you who holds position zero.
Read the format of the current snippet. Paragraph, numbered list, bulleted list, or table. Google has already decided which one this query deserves.
Rebuild that format on your page. A question-form heading, followed immediately by a direct answer of 40 to 60 words, with no warm-up sentence in between.
The reason this works is worth understanding. Google is not picking the best page on the internet and quoting it. It is extracting the block that best fits a format it already chose. Your job is to hand it a clean block.
The same discipline carries over to AI systems, with one addition. Models pull self-contained passages, so every section needs to survive being read in isolation. That means explicit definitions instead of pronouns pointing back three paragraphs, numbers with a citable source attached, and consistent structure from section to section. The generative engine optimization playbook goes deeper on this.
Set expectations honestly while you do it. Being cited in an AI Overview does not return the traffic a blue link produced in 2019. The Seer data above shows the gap between cited and uncited pages is large, and the gap between AI result pages and clean ones is also real. Both facts are true at once.
Step 10: Build SEO Authority Off Your Site
Off-site authority is where most SEO advice gets either vague or reckless. Five approaches are worth your time, and none of them involve buying anything.
Convert unlinked mentions. Someone already wrote your brand name without linking it. A polite email asking for the link converts a meaningful share of these, and it is the cheapest link acquisition that exists.
Be a source. Journalist request platforms like Qwoted and Help a B2B Writer send daily queries from writers who need an expert quote. Answering three a week from a real practitioner earns placements that no outreach sequence will.
Build linkable assets. Original data, a calculator, a template someone can use. People link to things they need again.
Guest post on real sites in your niche. Real means it has its own audience and would exist without link sellers.
Fix your directory profiles. For local clients, identical name, address and phone across every listing. Inconsistent NAP is a slow leak that nobody notices until a rankings audit.
One update for 2026: in AI search, unlinked brand mentions carry weight as an entity signal, not just linked ones. That changes the economics of outreach. Getting mentioned is cheaper than getting linked, and it is no longer worth nothing. The full breakdown of off-page signals that count covers how to measure them.
And the obvious warning, stated once: buying links violates Google’s spam policies. The upside is temporary and the downside lands on a client’s site, not yours.
Step 11: Make E-E-A-T Visible on the Page
Start with the correction, because this is one of the most misrepresented ideas in SEO. E-E-A-T, meaning Experience, Expertise, Authoritativeness and Trust, is not a measurable ranking factor. It is the framework human evaluators at Google apply when they rate results, and those ratings do not directly affect rankings. They feed back into algorithm refinement.
That does not make it useless. It makes it indirect. The signals that communicate E-E-A-T are entirely implementable, and most sites implement none of them:
Real author bios with actual credentials and a link to a full profile page
Person markup on those profiles, and a populated author field inside the Article schema
Publish and last-updated dates visible to readers, not buried in metadata
Sources cited with working links, especially for any claim carrying a number
Complete About and Contact pages with a real address and a real human named
A published editorial policy explaining who reviews content and how
Look at that list again and notice how much of it is structured data. An author bio at the bottom of a post is a paragraph. The same bio with Person markup and an author reference from the article is a machine-readable statement that this named individual with these credentials wrote this specific piece.
That is the link between this step and step 8. Trust signals a human can read are worth something. Trust signals a machine can parse are worth something in a search result.
Step 12: Track the SEO Metrics That Tell You It’s Working
Different signals move at different speeds, so checking everything weekly just generates noise.
Cadence
What to check
Where
Weekly
Impressions, clicks, and queries you have not ranked for before
Search Console, Performance report
Monthly
Average position for target pages, CTR per page, indexed page count, conversions from organic
Search Console plus your analytics
Quarterly
Referring domains, authority metrics, visibility in AI answers
A third-party SEO tool
Two vanity metrics deserve to be ignored. The rank of one isolated keyword moves daily for reasons unrelated to your work. Total traffic without segmentation can rise on a viral piece that converts nobody while your commercial pages quietly decline.
Then close the loop. Open the spreadsheet row from step 1 and compare the same five numbers. Not similar numbers, the same ones, measured the same way, over the same 28-day window. That comparison is the only thing that tells you whether the last three months of work did anything, and it is why the baseline came first.
For tracking visibility inside AI answers specifically, the tooling is still young and uneven. We tested AI SEO tools by use case to sort out which ones measure something real.
The average hides a lot of variance, and the variance is mostly about what you changed.
Technical fixes and title tag rewrites can show up within days or weeks, because Google only has to recrawl a page it already trusts. New content takes months, because it has to be discovered, indexed, evaluated and then compared against pages that have been there for years. Authority takes longest of all, because it depends on other people acting.
One number is worth showing any client who thinks six months is slow. Ahrefs studied 2 million random keywords and found the average top 10 page was over 2 years old, with pages in position one averaging close to 3 years. Only 5.7% of the pages studied reached the top 10 within a year for even one keyword. A later Ahrefs update put the figures higher still.
You are not competing against pages published last month. You are competing against pages that have been compounding since 2022. WebFX reaches similar conclusions on typical timelines.
Your Website SEO Improvement Checklist
Copy this into a project ticket and work down it.
This week
Record the five baseline numbers in a dated spreadsheet row
Review indexed versus not indexed in Search Console and open every exclusion reason
Submit the XML sitemap
Confirm robots.txt is not blocking CSS or JavaScript
Add self-referencing canonical tags
Run PageSpeed Insights on mobile
This month
Check search intent against the live SERP for every target page
Rewrite title tags and meta descriptions
Fix broken heading hierarchy
Clean up slugs and add real alt text
Add internal links pointing at the pages you want to lift
Add schema markup to the home page and every post
This quarter
Refresh the pages ranking in positions 8 to 20
Consolidate cannibalizing URLs and 301 the losers
Publish one new topic cluster
Chase featured snippets on queries where you already rank top 10
Convert unlinked brand mentions and pitch two source requests a week
Make E-E-A-T signals visible: bios, dates, sources, editorial policy
Ongoing
Compare against the baseline monthly
Revalidate schema after every template change
Update figures and screenshots before they go stale
Start With the Layer Most Sites Skip
Most sites work through steps 1 to 7 and stop there. The content is good, the technical foundation is clean, and Google and every AI model are still inferring what each page means from raw text.
Structured data is the cheap step that makes everything above it legible. Try the free schema generator on one page to see what valid markup looks like, or install Schemafy to generate, validate and inject JSON-LD across WordPress without writing code.
[CTA_DOWNLOAD]
Frequently Asked Questions About Improving SEO
These answers are marked up with FAQPage schema, which is exactly what step 8 recommends doing on any page with real questions on it.
How can I improve my website’s SEO for free?
Most high-impact SEO work costs nothing. Use Google Search Console to find indexing errors and pages ranking 8–20, rewrite title tags and meta descriptions, fix internal links, refresh outdated content, compress images, and add schema markup with a free generator. Paid tools speed up research, not results.
What is the fastest way to improve SEO?
Updating pages that already rank on page two. Find queries in positions 8–20 in Search Console, improve the matching page’s intent alignment, title tag and depth, then request reindexing. These pages already have authority, so gains often show in weeks instead of months.
How long does it take to improve SEO?
Most sites see measurable movement in three to six months, and six to twelve months in competitive niches. Google states meaningful results typically take four to twelve months. Technical fixes and title tag changes can show within days; new content and authority building take considerably longer.
Does schema markup improve SEO rankings?
Schema markup is not a direct ranking factor. It helps search engines and AI systems understand what your page contains, which makes you eligible for rich results like star ratings, FAQs and breadcrumbs. Those richer listings raise click-through rate and improve your odds of being cited in AI answers.
How often should I update my website for SEO?
Audit technical health monthly, review top-performing content quarterly, and refresh pages with outdated statistics or declining traffic at least once a year. Update only when you’re adding real value. Changing a publish date without changing the content does nothing for rankings.
Can I improve my website’s SEO myself without an agency?
Yes. Steps one through eight of this guide, from the baseline and technical fixes through intent matching, on-page optimization, content refreshes, internal linking and schema markup, are all DIY with free tools. Agencies mainly add speed, link building capacity and strategy at scale, not secret techniques.
Does page speed affect SEO?
Yes, but as a tiebreaker rather than a main driver. Google uses Core Web Vitals as a page experience signal, and slow pages lose visitors before they read anything. Test on mobile with PageSpeed Insights, since Google indexes the mobile version of your site first.
Most crawl budget guides assume you already have a crawl budget problem. Across ten client stores, maybe two actually do.
Crawl budget is the set of URLs Google can and wants to crawl on your site. Optimizing it means Google spends those requests on the pages that make money instead of on filter URLs nobody searches for.
This guide starts by telling you honestly whether it applies to you, then gives you the eight fixes that matter.
Caption: Three Search Console checks tell you which client sites deserve a crawl budget audit and which do not.
What is crawl budget?
Crawl budget is the set of URLs Google can and wants to crawl on your site over a given period. Google sets it with two things: how much your server can handle, and how much Google wants your content. Google treats each unique hostname as its own site with its own budget.
That last detail catches people managing portfolios. Your store and your blog subdomain do not share a budget, so a bloated subdomain is not stealing crawls from the money pages (Google Search Central).
Here is the rough arithmetic worth running before you do anything else. Take your indexable URLs and divide them by the average daily requests in your Crawl Stats report. If the answer is close to ten, Google needs more than a week to walk your site once. That is a field heuristic, not official guidance, but it is the fastest way to decide whether the rest of this guide is worth your afternoon.
Crawl capacity vs. crawl demand
Crawl capacity is how many simultaneous connections Googlebot will hold open without hurting your server. It goes up when your responses stay fast and stable, and it comes down when latency climbs or you return server errors and rate-limiting codes. This is the hosting lever.
Crawl demand is how much Google wants to crawl you. It is driven by how many URLs Google thinks you have, how popular those URLs are, and how stale they look. Google is direct about which part you own: perceived inventory is “the factor that you can positively control the most.” This is the URL inventory lever, and it is where seven of the eight fixes below live.
One detail almost nobody writes down: the crawl capacity limit is shared across all of Google’s crawlers, while crawl demand is calculated per crawler. If AdsBot is hammering a client store, it is taking capacity away from Googlebot.
Does crawl budget actually matter for your site?
For most sites, no. Google says so plainly at the top of its own guide: if your pages get crawled the same day you publish them, you do not need to read it.
The guide is written for two profiles. Large sites with 1 million or more unique pages that change roughly weekly, and medium or larger sites with 10,000 or more unique pages that change daily. Search Console is blunter still: “if you have a site with fewer than a thousand pages, you should not need to use this report or worry about this level of crawling detail” (Search Console Help).
This is not new. When Gary Illyes published Google’s canonical explanation back in 2017, the first thing he did was warn that crawl budget is not something most publishers need to worry about, and that sites with up to a few thousand URLs get crawled efficiently. Nine years later the message has not moved.
Site profile
Typical crawlable URLs
Change frequency
Does crawl budget matter?
Brochure or service site
Under 100
Rarely
No
Blog or small store
100 to 1,000
Weekly
No
Mid-size WooCommerce store
1,000 to 10,000
Weekly
Rarely, unless facets are crawlable
Large store with open faceted navigation
10,000+
Daily
Yes
Marketplace or programmatic site
1M+
Weekly or faster
Yes
Read the second column carefully, because it says crawlable URLs, not products. A WooCommerce store with 800 products and five open filters is not an 800-page site. It is a site with tens of thousands of crawlable URLs wearing an 800-product costume. That is how a portfolio of ten clients produces two or three real cases.
If none of your clients qualify, the principles still earn their keep. The same URL sprawl that burns budget on a large site delays indexing on a small one.
How to check if you have a crawl budget problem
Four checks, ordered by what they cost you in minutes. With ten accounts, the order matters more than the thoroughness.
Open the Crawl Stats report. It lives under property settings and only works on root-level properties. Read three numbers: total crawl requests, average response time, and host status. The report covers the last 90 days.
Run the ratio. Compare average daily requests against your indexable URL count. Then check the Crawl purpose breakdown, which splits requests into Discovery (a URL Google has never crawled) and Refresh (a recrawl of something known). If Refresh eats almost everything while Discovery sits near zero and you are publishing new products weekly, you have found your symptom.
Check the Pages report. “Discovered – currently not indexed” and “Crawled – currently not indexed” are normal in small numbers. Thousands of them, with product URLs inside, is the clearest signal Search Console will give you.
Read the server logs. Log-file analysis is the gold standard because it is the only method that tells you what share of Googlebot’s requests went to junk. It is also the only one that needs hosting access, a parser and a couple of hours per account. The red flag is unmistakable: most requests landing on URLs with filter and sort parameters instead of on product and category pages.
The first three checks take about fifteen minutes per property. Across ten clients that is one morning, and you finish it knowing exactly which two accounts deserve the log analysis.
Caption: Refresh at 94% and Discovery at 6% on a store publishing weekly is what a crawl budget symptom looks like.
How to optimize your crawl budget: 8 steps
The fixes below are ordered by impact and follow the principle Google puts first in its own guide: manage your URL inventory before you touch anything else.
Block low-value URLs in robots.txt.
Consolidate duplicate content onto one canonical URL.
Tame faceted navigation and URL parameters.
Flatten redirect chains to a single hop.
Fix soft 404s and return proper status codes.
Clean and update your XML sitemap.
Link internally to your priority pages.
Speed up server response and rendering.
1. Block low-value URLs in robots.txt
This is the highest-impact move because it removes entire URL spaces from consideration instead of fixing URLs one at a time. On WordPress those spaces are predictable: admin paths, cart and checkout, internal site search, staging subdirectories and sort parameters.
One clarification that trips up half the industry: robots.txt controls crawling, not indexing. Blocking a URL does not remove it from the index if Google already has it. Google does note that blocking “significantly decreases the chance the URLs will be processed by other Google systems,” so it is not useless for index hygiene, but it is not a deindexing tool.
Two warnings. Do not block CSS or JavaScript that Google needs to render your pages, which on a WooCommerce theme usually means leaving /wp-content/ alone. And do not reach for crawl-delay: the non-standard rule is not processed by Google’s crawlers at all.
In a portfolio, this file is the one fix that replicates almost unchanged across clients running the same stack.
2. Consolidate duplicate content
Duplicates make Google spend several requests on the same content. Google lists consolidation as the first item under URL inventory management for exactly that reason: the goal is to focus crawling on unique content rather than unique URLs.
On WordPress the usual sources are http and https or www and non-www resolving in parallel, inconsistent trailing slashes, session IDs appended by a plugin, and above all the tag and category archives your theme generates automatically. A WooCommerce store with 40 product tags is quietly publishing 40 near-identical listing pages.
Fix it in this order: canonical tags pointing at the version you want, 301 redirects for versions that should not exist at all, and internal links that consistently use the chosen URL. A clean URL structure makes the third step much cheaper.
The nuance that separates a technical SEO from the average guide: canonical is a signal, not a directive. Google may choose a different URL, and it will keep crawling the non-canonical versions for a while before the volume drops.
3. Tame faceted navigation and URL parameters
Do the arithmetic once and this section sells itself. Five filters with four values each, combinable in any order, produce over a thousand URL combinations before you count sorting and pagination. Google’s own term for the result is an infinite URL space, and it causes two specific harms: overcrawling of useless URLs, and slower discovery of the new ones you care about.
Fix it in Google’s recommended order.
Start by disallowing the filter parameters in robots.txt if you do not need those URLs indexed. Google’s own example blocks each parameter and whitelists the unfiltered listing:
If your theme allows it, switch filters to URL fragments instead of parameters. Google generally does not crawl fragments, so the filtering stops affecting crawling entirely.
Canonical tags to the clean category and rel="nofollow" on facet links are the second-line options, and Google is explicit that both are less effective long term. nofollow in particular only works if every anchor pointing at that URL carries it, anywhere on the web.
Two details worth writing on a sticky note. The URL Parameters tool in Search Console is gone, so any process that still expects it needs updating. And when a filter combination returns nothing, serve a 404 at that URL rather than redirecting to a shared error page.
Caption: Five filters with four values each generate over a thousand URLs from a single category page.
4. Flatten redirect chains and fix broken links
Google counts every hop in a redirect chain as a separate request. If page1 redirects to page2 which redirects to page3, Search Console records three requests to deliver one page. A migrated client site with three-hop chains is spending triple to serve the same catalog.
Point internal links straight at the final URL, flatten inherited chains to a single hop, and crawl the site with Screaming Frog, Sitebulb or Ahrefs to find them, and if you are still assembling that stack there is a breakdown of SEO tools by use case. On sites you inherit after a migration, chains are the default finding rather than the exception.
Now the correction, because most guides on this topic get it backwards. Google states that pages serving 4xx status codes, except 429, do not waste crawl budget. Google attempted the request, got a status code and no content, and moved on. Broken internal links are still worth fixing for user experience and internal link equity, but they are not a crawl budget line item. If you want the full treatment of those, fix the underlying SEO issues as a separate pass.
What genuinely hurts is 5xx errors and timeouts. They do not spend your budget, they shrink it: a meaningful volume of server errors signals poor health and Google slows crawling in response.
5. Fix soft 404s and error pages
A soft 404 is a page that returns HTTP 200 while being effectively empty or missing. On WooCommerce the classic case is a category page whose products all went out of stock and which now returns a successful response with zero results.
These are expensive because the status code tells Google there is content worth revisiting. In Google’s words, “soft 404 pages will continue to be crawled, and waste your budget.”
Return a 404 or 410 for pages that are genuinely gone, 301 the ones with a real replacement, and watch the soft 404 rows in the Search Console Pages report. The contrast is worth internalizing: a 404 is a strong signal not to crawl that URL again, while a blocked URL stays in the crawl queue much longer and gets recrawled once the block lifts.
6. Keep your XML sitemap clean and current
Your sitemap is the list of URLs you are declaring important, and Google reads it regularly. That only works if the list is honest.
A sitemap stuffed with noindexed, redirected and 404ing URLs does two kinds of damage. It spends crawl requests on URLs you did not want crawled, and it erodes the file’s usefulness as a priority signal.
Four rules cover it. Include only canonical, indexable, 200-status URLs. Keep <lastmod> accurate rather than stamping it with your last deploy date, which is what most WordPress sitemap plugins do by default. Split large sitemaps into a sitemap index. Submit the file in Search Console.
One myth dies here too: compressing your sitemap does not increase your crawl budget, because the server still has to serve the file either way.
7. Strengthen internal linking to priority pages
Internal links are how Googlebot discovers pages and how it infers which ones matter. A page with no internal links pointing at it gets crawled rarely, if at all.
On WooCommerce, orphans are usually products reachable only through internal search or through facet URLs you just blocked in step 1. That is the side effect worth auditing right after you touch robots.txt: check that blocking the facets did not orphan the products behind them.
Deciding which pages are priority is a search intent question before it is a crawling one. Link those pages from your highest-authority pages, keep them within a few clicks of the home page, and find orphans by crawling the site and diffing the result against your sitemap.
The honest caveat, in Google’s own framing: pages linked from the home page “may be seen as more important, and therefore crawled more often. However, this doesn’t mean that these pages will be ranked more highly.” Internal linking buys crawl frequency, not position.
8. Improve server response time and page speed
Faster responses let Googlebot fetch more pages per session, because the constraint is time and bot count rather than a fixed page quota. Serve more pages in the same window and Google crawls more of them.
Here is the part generic speed advice leaves out: time spent rendering counts as much as time spent requesting. A JavaScript-heavy WooCommerce theme burns budget even when your TTFB looks fine, because rendering the page is part of the crawl.
Reduce TTFB, put caching and a CDN in front of the site, support 304 responses so Google can reuse its cached copy of unchanged pages, and keep host status green in Crawl Stats. Across a portfolio, shared hosting is the most common root cause and the most uncomfortable conversation to have with the client.
Calibrate your expectations, though. Google may spend more time on a slower site that holds more important information, and it says outright that making your site faster for users matters more than making it faster for crawl coverage.
Mistakes that waste crawl budget (and myths to ignore)
Half the work in this section is stopping you from spending an afternoon on the wrong fix.
Myth
What actually happens
What to do instead
noindex saves crawl budget
Google must crawl the page to see the tag, so the request is spent anyway
Use noindex to deindex, robots.txt to stop crawling
robots.txt removes pages from the index
It blocks crawling; a known URL can stay indexed
Allow the crawl and serve noindex, or remove the URL
More crawling means better rankings
Crawling is required for indexing but is not a ranking signal
Optimize which pages get crawled, not how many
Broken 404 links waste crawl budget
Pages returning 4xx, except 429, do not waste budget
Fix them for UX and link equity, not for crawl budget
Compressing your sitemap increases crawl budget
The file still has to be fetched from your server
Keep the sitemap accurate instead of small
crawl-delay controls Googlebot
Google’s crawlers do not process the rule
Return 503 or 429 temporarily if you need relief
Two of those deserve more than a table cell.
On noindex, Google’s position is more precise than the internet’s version of it. Google tells you not to use it as a crawl budget tool, because it “will still request, but then drop the page when it sees a noindex meta tag.” At the same time, Google acknowledges that removing URLs from the index can indirectly free up budget over the long run, since crawlers can then focus elsewhere. So: noindex for index control, robots.txt for crawl control, and no expectation of an overnight effect either way.
On crawling and rankings, Google is unambiguous: improving your crawl rate will not necessarily improve your positions. Crawling is a precondition, not a signal. It is the same category of misunderstanding as believing meta descriptions are a ranking factor, which they are also not a ranking factor.
Where the budget actually goes: infinite URL spaces from facets, duplicate content across parallel URLs, multi-hop redirect chains, and soft 404s that never stop looking alive. Add one nobody counts, alternate URLs and embedded resources. Your CSS, JavaScript and XHR fetches consume crawl budget too.
How structured data and clean metadata help crawlers
Everything above decides which pages Google visits. None of it decides what Google understands once it arrives. That is a separate problem with separate tools.
Valid JSON-LD gives a crawler the entity type, the price, the availability and the rating without making it infer any of that from your HTML, and it is the entry requirement for rich results. On a store where the product template is the same across a thousand pages, that difference is the whole game. Structured data is doing the interpretation work your markup otherwise leaves ambiguous, and there is a full walkthrough of how to use schema markup if you are starting from zero.
Clean metadata does the same job at the SERP layer. Unique titles and descriptions stop hundreds of pages from introducing themselves identically, which on a templated WooCommerce catalog is the default state rather than the exception. Writing a meta description in WordPress one page at a time is fine for a blog and unworkable for a catalog, and you can preview how the snippet renders before you commit to a pattern.
Be clear on the limit: none of this increases your crawl budget. It makes each crawl more useful, which is a different claim and the only one worth making.
Optimize crawl budget with Schemafy
Schemafy does not manage your crawling. It does not edit robots.txt, generate sitemaps, set canonicals or read server logs, and any plugin that claims to solve crawl budget for you is selling you something.
What it does is the layer after the crawl. Schemafy generates valid JSON-LD across 16 schema types, through a template builder, an automatic site scan or an AI generator, so the pages Google does fetch are immediately legible. And its bulk meta editing works from a CSV import, which is how hundreds of product pages get unique titles and descriptions instead of the same boilerplate repeated across a catalog.
Caption: Character counters and a live Google preview on a single product page; the same fields import in bulk from CSV.
Final thoughts
Crawl budget is an inventory problem wearing a technical costume. You are not looking for a trick that makes Googlebot work harder, you are deciding which URLs deserve to exist and removing the rest from consideration.
The next step is fifteen minutes, not a project. Open Crawl Stats on your largest client property, run the ratio, and look at the Discovery versus Refresh split before you change a single line of robots.txt. Once the crawl is going where you want it, the follow-on work is making those pages earn their place, which is where the broader playbook to increase organic traffic picks up.
Frequently asked questions
Does crawl budget matter for small websites?
For most small sites, no. Google can easily crawl and index sites under about 1,000 pages, and Google has said crawl budget isn’t a concern for most publishers. It becomes important on large sites (1M+ pages) or medium sites (10k+) that change daily. Small sites still benefit from clean architecture.
Does noindex save crawl budget?
Not directly. Google has to crawl the page to see the tag, so the request is spent either way. Over the long run, removing URLs from the index can indirectly free up crawling for other pages. Use noindex to deindex and robots.txt to stop crawling.
How do I increase my crawl budget?
You raise crawl budget by improving crawl capacity and demand: speed up server response, fix 5xx errors, publish and update valuable content, and build internal and external links to important pages. You can’t set a number directly, but a fast, healthy, popular site gets crawled more.
How often does Google crawl my site?
It varies by site. Google crawls popular, frequently updated pages more often and low-value pages rarely. Check the Crawl Stats report in Google Search Console to see your actual crawl frequency, total requests, and average response time over the last 90 days.
Does page speed affect crawl budget?
Yes. Faster pages let Googlebot fetch more of your site in each session, effectively increasing crawl capacity. Slow responses and server errors make Googlebot slow down to avoid overloading your server, so it crawls fewer pages. Rendering time counts as much as response time.
If your rankings stalled or your traffic slipped, the cause is rarely a mystery. Most SEO problems fall into a handful of buckets, and nearly all of them are fixable without an agency. This guide shows you how to fix SEO issues by symptom: find each one, understand why it hurts, and follow the steps to clear it. The starting point is a free crawl in Google Search Console, Ahrefs, or Screaming Frog.
In this guide:
How to Find SEO Issues Before You Fix Them
How to Fix Slow Page Speed and Poor Core Web Vitals
How to Fix Broken Links and 404 Errors
How to Fix Duplicate Content
How to Fix Pages That Aren’t Getting Indexed
How to Fix Missing or Weak Title Tags and Meta Descriptions
How to Fix Mobile Usability Problems
How to Fix Thin or Low-Value Content
How to Fix Missing or Broken Schema Markup
How to Prevent SEO Issues From Coming Back
Fix Your Schema Issues in Minutes with Schemafy
Frequently Asked Questions
How to Find SEO Issues Before You Fix Them
You can’t fix what you can’t see. Before you touch a single page, measure. Guessing wastes hours and usually fixes the wrong thing.
Three free tools surface almost every issue on this list. Google Search Console reports indexing problems and Core Web Vitals errors. A crawler like Screaming Frog or Ahrefs finds broken links and duplicate content. Google PageSpeed Insights scores speed and gives page-by-page fix instructions.
Run all three, then work down the list below. Each section is one issue your audit will flag.
Run a Google Search Console audit for indexing and Core Web Vitals errors.
Crawl your site with Screaming Frog or Ahrefs to surface broken links and duplicate content.
Test key pages in Google PageSpeed Insights for speed and Core Web Vitals.
Validate your structured data with Google’s Rich Results Test.
Work down the list below, fixing one issue type at a time.
If you run several client sites, this three-tool pass is the standard check to run before you bill a single hour of fixes. Here is the full map of what you will find and where to fix it.
Responsive theme, larger tap targets, no interstitials
Thin content
High bounce, low time-on-page
Expand with data, consolidate near-duplicates
Missing or broken schema
Rich Results Test, GSC Enhancements
Add valid JSON-LD for the right type
How to Fix Slow Page Speed and Poor Core Web Vitals
Speed is part of how Google judges a page, and mobile speed matters most. Page experience is not a single ranking signal, but Google’s core systems weigh Core Web Vitals, HTTPS, and mobile-friendliness together, according to Google’s page experience documentation.
Aim for the thresholds Google calls “good”: Largest Contentful Paint at or below 2.5 seconds, Interaction to Next Paint below 200 milliseconds, and Cumulative Layout Shift below 0.1. Interaction to Next Paint replaced First Input Delay in March 2024, so if you are still tracking FID, you are watching a retired metric (web.dev).
Work through these fixes in order:
Compress and serve next-gen images. Convert heavy JPEGs and PNGs to WebP or AVIF. Images are the most common cause of a slow Largest Contentful Paint.
Cut unused scripts and plugins. A stack of 4 or 5 overlapping SEO and marketing plugins loads scripts on every page. Deactivate what you do not use.
Enable caching and pick a lightweight theme. A bloated theme can undo every other fix.
Lazy-load below-the-fold media so images and videos load only when the reader scrolls to them.
Re-test in PageSpeed Insights after each change. It tells you exactly which resource is slowing the page, so you fix the real bottleneck instead of guessing.
How to Fix Broken Links and 404 Errors
Broken internal and external links waste crawl budget, frustrate readers, and signal a neglected site. Every dead link is a small tax on how efficiently Google crawls you.
Crawl the site with Screaming Frog or Ahrefs to list every 404 and every redirect chain. Then fix them:
301-redirect broken internal links to the closest live page.
Update or remove dead external links pointing off your site.
Fix redirect chains by pointing the original link straight to the final URL, not through two or three hops.
On a WooCommerce store, deleted and out-of-stock products are the most common source of 404s. When you retire a product, redirect its URL to the parent category or a close replacement instead of letting it 404. If you manage client sites, keep a redirect map for every migration so old URLs never dead-end.
How to Fix Duplicate Content
This is the highest-value section for any store owner. Duplicate content is where WooCommerce and WordPress sites quietly bleed rankings, and almost nobody audits for it.
Near-identical pages come from three places. Product variants for color and size can each spin up their own URL. Filtered, tag, and category archive URLs create thin variations of the same list. And a single product reachable through several category paths can exist at several addresses. Google then has to guess which version to rank.
Point Google at the right version:
Add canonical tags so every duplicate points to the primary version of the page.
301-redirect true duplicates that should not exist at all.
Use hreflang if you run regional stores in different languages or currencies.
Apply noindex to thin tag and archive pages that add no unique value.
Canonicalization matters most on stores with product variations. You set canonical tags through your existing SEO plugin, such as Yoast or Rank Math, or in your theme templates. A consistent canonical policy across client sites saves you from relitigating this on every audit.
How to Fix Pages That Aren’t Getting Indexed
Indexing is eligibility to rank. If a page is not indexed, it cannot earn a single visit, no matter how good it is.
Open the Page indexing report in Google Search Console and read the exclusion reason for each affected URL. Then work through the usual causes:
Confirm the page is not set to noindex and is not blocked in robots.txt.
Add internal links pointing to the page so Google can discover and prioritize it.
Improve thin content, since low-value pages are often left unindexed on purpose.
Once the page is worth indexing, submit it through URL Inspection and click Request Indexing.
The usual culprits are thin, low-quality content and orphan pages with no internal links. On large client sites, orphan pages pile up fast, so clean URL slugs and a sensible internal-link structure prevent most indexing gaps before they start.
How to Fix Missing or Weak Title Tags and Meta Descriptions
Title tags influence rankings. Meta descriptions influence click-through. Weak or duplicate tags leave both on the table.
Write a unique, keyword-first title tag for every page, kept to about 60 characters so it does not truncate in search. Write a compelling meta description of roughly 155 characters with a clear benefit or call to action. Front-load the primary keyword in both, and never reuse the same title across pages. If you want the background on why descriptions still matter, see whether meta descriptions are a ranking factor and the step-by-step guide to add a meta description in WordPress.
Editing tags one page at a time is fine for a homepage. It does not scale to hundreds of posts or products. In Schemafy, open WP Admin > Schemafy > Meta Tags, search for the page, and edit the Meta Title and Meta Description fields. A character counter and validation status keep you inside the limits, and the Google Preview shows the live search snippet before you save. Click Save Meta Tags to apply.
Schemafy shows the live Google Preview as you rewrite a product page’s title and description, so you catch truncation before it hits the SERP.
To rewrite tags in bulk, open Meta Tags > Bulk Import CSV, click Download Template for a file with the required url, meta_title, and meta_description columns, fill in your rows, and upload it. The import preview flags each row as Valid, Warning, or Invalid, then Import Rows applies only the valid ones. You can also test any title and description length against a live snippet with the SERP preview tool before you commit.
How to Fix Mobile Usability Problems
Google uses mobile-first indexing, which means it evaluates the mobile version of your site for ranking, as explained in Google’s page experience guidance. If your mobile experience is broken, your desktop polish will not save you.
Fix the fundamentals:
Use a responsive theme that adapts cleanly to phone screens.
Make tap targets and font sizes large enough to use without pinching or zooming.
Remove intrusive interstitials and pop-ups that cover the content on load.
Simplify navigation so the core paths work with a thumb.
Test on a real device and run Lighthouse in Chrome, since Google retired the standalone Mobile Usability report in Search Console. Because mobile Core Web Vitals feed the same page-experience signals, most speed fixes from earlier in this guide also improve mobile usability.
How to Fix Thin or Low-Value Content
Thin content underperforms, and Google may decline to index it at all. Pages that say very little give search engines very little reason to rank them.
Start by finding pages with high bounce and low time-on-page in your analytics. Then decide, page by page, whether to expand or consolidate:
Expand pages worth keeping with real data, answers to the follow-up questions readers ask, images, and internal links.
Consolidate or 301-redirect near-duplicate thin pages into one strong page.
Add unique descriptions to category and collection pages instead of leaving them bare.
The test for every page is whether it matches search intent better than what already ranks. If it does not, expand it until it does, or fold it into a page that does.
How to Fix Missing or Broken Schema Markup
Most SEO guides mention schema in passing. That is a mistake, because broken structured data is one of the few issues that changes how your listing looks in search, not just where it ranks.
Missing, invalid, or incomplete structured data means no rich results: no star ratings, no price, no FAQ, no breadcrumbs. You lose click-through even when your ranking is fine. The impact is measurable. Rotten Tomatoes added structured data to 100,000 pages and saw a 25% higher click-through rate on marked-up pages, and Nestlé measured pages shown as rich results getting an 82% higher click-through rate than non-rich pages, both reported in Google’s structured data documentation.
That makes schema a fixable, high-return SEO issue rather than a technical footnote.
A single product with valid Product and Review schema earns stars and a price in the SERP, while the plain results below compete on text alone.
How to check your structured data
Check your key pages before you change anything. Run them through Google’s Rich Results Test and the Schema Markup Validator, then open the Enhancements and Rich results reports in Search Console to see errors and warnings across the whole site. Look for missing required properties, such as a product price or a review rating, and for invalid syntax.
Most schema problems fall into four buckets:
No schema at all on pages that qualify for rich results.
The wrong @type for the page, such as Article on a product page.
Missing required fields, so the markup is present but ineligible.
Markup that does not match the visible content on the page.
That last one matters: Google does not guarantee rich results even for valid markup, and it requires your structured data to match what users actually see, per Google’s structured data guidelines. If you want to inspect or hand-edit the raw code, a JSON-LD editor validates it against the same specification. Inside Schemafy, the Rich Snippets screen lists every schema already applied across your site, so you can see coverage and gaps in one place.
How to fix it fast on WordPress
You have two paths. The first is editing JSON-LD by hand in your theme files, such as functions.php or template files. It works, but it is error-prone, usually needs a developer, and breaks when the theme updates. The second is a dedicated plugin that injects validated JSON-LD without touching your theme.
Schemafy takes the second path. To cover a whole site, open WP Admin > Schemafy > Auto Schema Generator, click Scan Site, and filter by Post Type and Status, for example Product and Needs Schema. The scan lists every page with its current and suggested schemas so you can see coverage at a glance. For a specific page or a complex type, open AI Schema Generator, select the page, choose a type such as Product or FAQPage, click Generate Schema with AI, review the generated JSON-LD and its validation status, and click Save to Website. You can also start from a template with the schema markup generator or follow the full walkthrough on how to use schema markup. Either way, you add Product, Review, BreadcrumbList, and FAQPage schema that stays valid, without writing code.
Schemafy’s Auto Schema Generator flags every product page missing schema after a single scan, so you fix coverage in bulk instead of page by page.
[CTA_DOWNLOAD]
How to Prevent SEO Issues From Coming Back
SEO is not one-and-done. Most sites fix their issues once, stop monitoring, and watch broken links, speed regressions, and missing schema creep back within months. Maintenance beats emergencies.
Build a short routine you can repeat:
Schedule a monthly crawl with your usual tool.
Watch Search Console for new coverage and Core Web Vitals errors.
Re-validate your schema after every theme or plugin update, since updates are what silently break markup. In Schemafy, the Rich Snippets screen shows what is still applied so you can spot anything that dropped.
Keep a simple audit checklist so nothing gets skipped.
If you manage client sites, turn that checklist into a repeatable monthly pass and you will grow organic traffic more steadily than any one-time cleanup ever could.
Fix Your Schema Issues in Minutes with Schemafy
Of every issue in this guide, schema is one of the fastest to fix and one of the very few that directly changes how your listing looks in search. Rich results are a click-through win you can capture this week.
Schemafy is the no-code way for WordPress sites and WooCommerce stores to add valid JSON-LD, including Product, Review, BreadcrumbList, and FAQPage, without editing theme code. It complements the SEO work you are already doing. The fastest first step is to scan your site and see exactly what schema is missing.
[CTA_DOWNLOAD]
Frequently Asked Questions
Quick answers to the questions store owners and agencies ask most about fixing SEO issues.
How do I check what SEO issues my site has?
Start with three free tools: Google Search Console for indexing and Core Web Vitals errors, a crawler like Screaming Frog or Ahrefs to find broken links and duplicate content, and Google PageSpeed Insights for speed. Together they surface almost every fixable SEO issue on your site.
How long does it take to fix SEO issues?
Individual fixes, like redirects, meta tags, or schema, take minutes to hours. Seeing ranking or traffic changes takes longer: Google must recrawl and reprocess your pages, which can be days for small fixes and several weeks for content or indexing changes across a larger site.
What are the most common SEO issues?
The most common SEO issues are slow page speed, broken links, duplicate content, pages not getting indexed, weak title tags and meta descriptions, poor mobile usability, thin content, and missing or invalid schema markup. Most sites have several at once, and nearly all are fixable without a developer.
Can I fix SEO issues myself without a developer?
Yes. Most SEO issues, like meta tags, redirects, image compression, indexing, and content, can be fixed in your CMS or WordPress settings without code. Schema markup is the main exception, but no-code plugins let you add valid structured data without editing theme files.
Does schema markup help fix SEO issues?
Schema markup doesn’t boost rankings directly, but it fixes a real visibility issue: without valid structured data your pages can’t show rich results like star ratings, prices, or FAQs. Adding correct JSON-LD schema improves how your listing looks and lifts click-through rate.
Two people type “jaguar” into Google. One wants the animal. The other wants the car. Same word, two completely different goals. Search intent is the real question hiding behind the words, and it decides which page Google shows each of them. Get it right and you rank. Get it wrong and no amount of keywords will save you.
Search intent (also called user intent or keyword intent) is the goal behind a search query: the reason someone typed it and what they hope to find. Every Google search is a question or a task, and search engines rank the pages that best satisfy that goal, not just the pages that repeat the keyword.
You will also see it called audience intent. The label matters less than the idea: behind every query is a person trying to do something specific.
Think of it like ordering at a restaurant. The words on the menu matter, but the waiter’s job is to bring what you actually want. Google is the waiter. If you ask for a definition and it hands you a checkout page, it did its job badly, and it knows it.
Why search intent matters for SEO
Google ranks pages on relevance, and relevance is mostly about intent. When your content matches the goal behind a query, it earns the spot. When it doesn’t, it slips, no matter how strong the page looks on paper.
Google says this plainly. “Does your content match intent?” sits among the self-assessment questions in its helpful content guidance for creators. Matching intent against the query is treated as a central ranking criterion in most practitioner guides too, including Ahrefs’ search intent guide.
Miss the intent and you can have the best backlinks in your niche and still not rank. Visitors land, see the wrong kind of page, and bounce straight back to Google. That pattern, called pogo-sticking, tells search engines your page did not answer the question.
There is a business side too. The right intent brings traffic that converts instead of traffic that just visits. A page that meets the searcher where they are is how you turn rankings into leads and increase organic traffic that actually pays off.
The 4 types of search intent
Almost every query falls into one of four types: informational, navigational, commercial, and transactional. Learn to spot them and you can predict what Google wants to show before you write a single word. Most keywords sit cleanly in one type. A few are mixed, which we will cover later.
Informational intent
The user wants to learn something or answer a question. These queries lean on modifiers like “how,” “what,” “why,” “when,” “guide,” and “tutorial.” Examples: “what is search intent” (yes, the query that brought you here) or “how does SEO work.”
Informational is the biggest bucket by far. It is widely cited that around 70% of all searches are informational (Neil Patel), though more recent analysis puts the figure closer to 57% of Google queries (Ahrefs). Either way, it dominates. The ideal format is a blog post or a guide.
Navigational intent
The user is looking for a specific site or page. They already know where they want to go and use search as a shortcut. Examples: “yoast login,” “semrush pricing,” or any brand name typed straight into the search bar.
One practical note: it rarely makes sense to chase navigational queries for a brand you do not own. Someone searching “semrush login” wants Semrush, not your comparison post.
Commercial intent
The user is researching before a purchase. They want to buy eventually, but they are still comparing options. Modifiers give it away: “best,” “top,” “review,” “vs,” and “comparison.” Examples: “best SEO tools” or “Semrush vs Ahrefs.”
These searchers reward pages that help them decide. The ideal format is a comparison, a listicle, or an in-depth review that lays out real trade-offs instead of a sales pitch.
Transactional intent
The user is ready to act, usually to buy or sign up. The query signals a decision already made. Modifiers include “buy,” “price,” “cheap,” “discount,” “coupon,” and “near me.” Examples: “buy SEO software” or “SEO agency near me.”
Serve these searchers a product page, a service page, or a focused landing page. A long explainer here just gets in the way of the thing they came to do.
Search intent examples (keyword by keyword)
Intent is easier to see with real keywords. Each one below belongs to a type, and that type dictates the page format Google rewards. If your page format does not match the right-hand column, you will struggle to rank no matter how good the writing is.
Keyword
Search intent
Page type Google rewards
what is search intent
Informational
Guide / blog post
how to add schema markup
Informational
Tutorial / how-to
meta description length
Informational
Guide with a quick answer
yoast login
Navigational
Specific site page
schemafy blog
Navigational
Brand page
best seo plugins for wordpress
Commercial
Listicle / comparison
semrush vs ahrefs
Commercial
Head-to-head comparison
buy seo software
Transactional
Product / pricing page
seo agency near me
Transactional
Local service / landing page
wordpress hosting coupon
Transactional
Deals / pricing page
Read the table top to bottom and the pattern is clear: the words tell you the goal, and the goal tells you the page.
How to identify the search intent behind a keyword
You do not have to guess. The SERP is the source of truth, because Google has already classified the intent for you and ranked the results that prove it. Here are three ways to read it, from most reliable to least.
Analyze the SERP
Type the keyword into Google and look at what dominates the first page. This is the most direct signal you have, because Google has already done the classification work (Search Engine Land). If the top 10 are guides and blog posts, the intent is informational. If they are product and pricing pages, it is transactional. If comparison articles rank, it is commercial.
SERP features are signals too. People Also Ask boxes and featured snippets point to informational intent. Shopping carousels point to transactional. A local pack means Google reads a “near me” goal even if you did not type those words.
Google’s own SERP is the clearest intent signal: guide-style results and a People Also Ask box mark this as an informational query.
Read the modifiers in the query
The extra words around your main keyword reveal the likely intent before you even open Google. Use this as a quick first read, then confirm with the SERP.
Modifier
Likely intent
how, what, why, guide, tutorial
Informational
best, top, review, vs, comparison
Commercial
buy, price, cheap, coupon, discount, near me
Transactional
brand name, login
Navigational
One rule keeps you honest: SERP evidence overrides the modifier (Ahrefs). If the words say one thing but the ranking pages say another, believe the pages.
Use a keyword tool’s intent labels
Tools like Ahrefs and Semrush now tag intent for you automatically, which is handy when you are sorting hundreds of keywords at once. The same is true of newer AI SEO tools that add intent scoring. Treat these labels as a starting point, not a verdict. A tool guesses from patterns; the live SERP shows what Google actually decided. Always validate against the real results.
How to optimize your content for search intent
Once you know the intent, three moves align your page with it. None of them require a rewrite from scratch.
Match the content type and format
Give the searcher the format their intent expects. A guide for informational queries. A comparison for commercial ones. A product or landing page for transactional ones. The single most common mistake in SEO is publishing a blog post for a transactional keyword, then wondering why a page full of product results outranks it. The content type has to match before anything else you do matters.
Cover the whole topic (content depth)
Answer the main question, then answer the questions around it. People Also Ask boxes are a free checklist of the subtopics Google associates with your query, so mine them and cover what is missing. Keep your headings clear and scannable so a reader can confirm in seconds that your page solves their search. Depth is not word count. It is coverage of what the searcher actually needs.
Align title tags and meta descriptions
Your title tag and meta description are the first intent check a searcher runs, right there in the results. They should reflect the goal and promise the outcome, which lifts click-through rate and reinforces relevance. It also helps to see how the snippet looks before you publish. A no-code plugin like Schemafy lets you edit the meta title and description and preview your snippet the way Google shows it, so you can match the wording to intent without touching code. If you are unsure how much these tags move the needle, here is the reality on meta descriptions and CTR and how to add a meta description in WordPress.
Search intent in the age of AI Overviews and AEO
Search intent is shifting under our feet. More and more informational queries now get answered inside the SERP itself through AI Overviews, so a growing share of searches end without a click. The old goal, rank first, is no longer the whole game.
The new goal is to be the source the AI cites. That means writing answer-first: lead with a direct answer, back it with concrete data, and name your entities clearly so a language model can lift and attribute your point. Structured data helps here, which is why schema markup is becoming table stakes for AI search.
Intent itself is expanding. SE Ranking now lists a sixth type, generative AI intent: queries where the user expects a synthesized answer from an AI engine rather than a list of links (SE Ranking). Optimizing for it has its own name, Answer Engine Optimization, and it overlaps heavily with generative engine optimization. The core skill has not changed. You still start by understanding what the searcher wants. You just have one more surface to satisfy.
Informational intent increasingly resolves inside the SERP: an AI Overview answers the query before the user reaches a single organic link.
Common search intent mistakes to avoid
Most intent problems come down to the same handful of errors:
Forcing transactional keywords into blog posts. If Google ranks product pages, an article will not win the spot.
Ignoring the SERP and guessing. Your assumption about intent loses to the pages already ranking. Check first.
Mixing multiple intents on one page. Trying to inform and sell in the same breath usually does neither well.
Publishing thin content for informational keywords. Informational SERPs reward depth, and a 300-word skim gets buried.
Not updating when the SERP shifts intent. Intent changes over time. A keyword that was informational last year may be commercial now, so re-check your winners.
Get help matching your content to search intent
Aligning your title tags and meta descriptions with intent is where a lot of pages quietly win or lose their click-through. It is a small edit with an outsized effect on how well your snippet matches what the searcher wants.
If you would rather not hand-code any of it, a no-code plugin like Schemafy lets you rewrite meta titles and descriptions and preview the snippet the way Google displays it, so you can tune the wording to intent in a few minutes.
Frequently asked questions
What are the 4 types of search intent?
The four types of search intent are informational (learning something), navigational (finding a specific site), commercial (researching before a purchase), and transactional (ready to buy or act). Each maps to a different content format, so matching the right type is key to ranking.
How do I find the search intent of a keyword?
The most reliable way is to type the keyword into Google and study the top results. If they are guides, the intent is informational; if they are product pages or listicles, it is transactional or commercial. Query modifiers and keyword-tool intent labels help confirm.
What is the most common type of search intent?
Informational intent is the most common, accounting for roughly 70% of all searches. These users want to learn or answer a question, using modifiers like “what,” “how,” and “why.” That is why guides and blog posts dominate informational SERPs.
Can one keyword have multiple search intents?
Yes. Some keywords are ambiguous and show mixed results in the SERP, part informational, part commercial. Google hedges by ranking a blend of formats. In those cases, review the SERP and match the dominant intent while covering secondary angles.
Final thoughts
Search intent, not the keyword itself, is what decides whether you rank. The page that best satisfies the goal behind the query wins, and everything else you do in SEO is downstream of that one match. Backlinks, word count, and clever titles only pay off once the intent is right.
So before you write your next piece, open Google, read the SERP for your target keyword, and commit to a single intent type. Match the format, cover the topic, and align your title and meta description to what the searcher actually wants. Then re-check that same SERP in a few months, because intent shifts over time, and the page that keeps matching it is the page that keeps ranking.
Think of structured data like labels on a shipping container. A machine reads the label and knows what is inside without opening the box. It does the same for your web pages: it tells search engines what your content means, not just what it says.
The term also describes database data, but this guide covers structured data for SEO. Done right, your pages become eligible for rich results, higher click-through, and AI-search citations across every client site you run.
Structured data is a standardized format for labeling the content on a web page so search engines can understand what it means, not just what it says. It uses a shared vocabulary called Schema.org to describe entities like a product, an article, or a business, and it lives in the page’s code where machines read it, invisible to visitors.
That shared vocabulary is the important part. Schema.org gives every entity type an agreed set of properties, so a search engine reading Product knows to look for a name, a price, and a rating in the same place every time.
Back to the shipping analogy. The label does not change what is in the box, it just makes the contents readable without unpacking. On one site, that is a nice-to-have. Across 10 or more client sites, it is the difference between listings that qualify to look better in search and listings that stay plain.
Structured data vs. schema markup vs. rich results
These three terms get blurred constantly, often in the same sentence. Keeping them straight saves you from promising a client something search engines never guaranteed.
Structured data is the concept: labeling content in a machine-readable format. Schema markup is the implementation, the actual code you add using the Schema.org vocabulary. Rich results are the output, the enhanced listing Google may show once it reads valid markup. All schema markup is structured data, but not all structured data is schema markup.
Term
What it is
Example
Structured data
The concept of labeling content so machines understand it
Any machine-readable data format, including database tables
Schema markup
The implementation using the Schema.org vocabulary
A Product JSON-LD block on a product page
Rich result
The enhanced search listing Google may display
A star rating and price shown under a result
How structured data works
The flow is short. You add the markup to a page. Search engines crawl and parse it. They use it to understand the entities on the page and how those entities relate to each other. Pages that qualify may then earn rich results.
There is a bigger payoff behind that. Google uses structured data it finds on the web to understand the content of a page and to gather information about the world, including the people, books, and companies your markup describes. That information feeds the Knowledge Graph, the map of entities Google fills in as your markup confirms who and what your page is about. If you want the full workflow, see our guide on how to use schema markup.
The three formats: JSON-LD, Microdata, RDFa
Schema.org markup comes in three formats. JSON-LD is a <script> block placed in the head or body, kept separate from the visible HTML. Microdata and RDFa are inline attributes woven directly into your HTML tags.
Each has a trade-off. JSON-LD is clean and easy to edit because it sits apart from your layout. Microdata ties the data to the exact HTML element, which some developers like but makes bulk edits fragile. RDFa is the most flexible for combining vocabularies, and also the most verbose. JSON-LD is by far the most common format today, and you can write or check a block in a JSON-LD editor before it goes live.
Why Google recommends JSON-LD
Google’s position is explicit. It recommends JSON-LD because it is the easiest solution for site owners to implement and maintain at scale, and less prone to user errors.
The reason is the separation. Because the JSON-LD block lives apart from your visible content, changing the page layout does not break the data, and updating the data does not touch the layout. You can inject it through a tag manager or a CMS, which is what makes it maintainable across many sites at once.
Structured data examples (with code)
Here is what schema markup actually looks like. This is a minimal, valid Article block. Every field is labeled in plain English so you can see what it does.
You can build either block by hand, or generate one field by field in a schema markup generator. The property names come straight from Schema.org, and Google documents the requirements for each rich result type in its product structured data docs.
The payoff shows up in the search result. Without markup, a product listing is a title, a URL, and a description. With valid Product markup, Google can add the star rating, the price, and the stock status directly to the listing, which takes up more space and gives shoppers a reason to click before they even land.
Common types of schema markup
Schema.org defines hundreds of types, but a handful cover most of what site owners and stores actually need. Here are the high-value ones, each with the rich result it can earn.
Schema type
What it labels
Rich result it can earn
Article
Blog posts and news content
Article and headline enhancements
Product
E-commerce items for sale
Price, availability, star rating
LocalBusiness
Stores and physical locations
Business info, hours, map details
FAQPage
Question-and-answer blocks
Expandable FAQ under the listing
HowTo
Step-by-step tutorials
Step-by-step rich result
Review / AggregateRating
Ratings and opinions
Star ratings
BreadcrumbList
Site hierarchy and navigation
Breadcrumb trail in the result
Organization
Company or brand identity
Knowledge panel and brand details
Event
Concerts, webinars, meetups
Event date and location card
Recipe
Cooking instructions
Recipe card with photo and time
VideoObject
Video content
Video thumbnail and key moments
Two of these are the easiest wins across a whole site. BreadcrumbList and Organization apply to almost every page, take little effort to set up once, and pay off site-wide. FAQPage markup can surface your questions in the People Also Ask box. If you are rolling schema out across many client sites, starting with the types that apply everywhere gives you the most return per hour.
Why structured data matters for SEO and AI search
Here is the honest version. Structured data is not a direct Google ranking factor. Adding a Product block will not, by itself, move you up the results the way meta descriptions are not a direct ranking factor yet still shape performance.
What it does provide is two things: enhanced listings that earn more clicks, and machine-readable context that helps search engines and AI systems confirm what your page is about. Both are covered below.
Rich results and higher CTR
Rich results are enhanced listings: star ratings, FAQ accordions, prices, breadcrumbs, and recipe cards. They occupy more space in the search results and give people more reasons to click, which lifts click-through rate.
The numbers Google publishes are strong. Rotten Tomatoes added structured data to 100,000 unique pages and measured a 25% higher click-through rate for pages with structured data compared to pages without it. Nestlé found that pages shown as rich results have an 82% higher click-through rate than the same pages without a rich result. Rakuten measured that users spend 1.5 times longer on pages with structured data than on pages without.
One caveat matters. Eligibility is not a guarantee. Valid markup makes a page eligible for a rich result, but Google decides when to show one. You can preview how a listing might look in a SERP preview tool before you publish.
Nestlé found that pages shown as rich results have an 82% higher click-through rate than the same pages without one.
AI Overviews, LLM citations, and entity recognition
This is the part most older guides miss. Structured data helps Google’s Knowledge Graph and AI systems, including AI Overviews and assistants like ChatGPT, Gemini, Perplexity, and Claude, verify entities and surface your content in AI-generated answers.
The mechanism is confirmation, not magic. When your markup states clearly that a page is about a specific product, business, or author, it helps machines confirm what the page is about and how its details relate, which makes your content easier to cite accurately. That is the core idea behind Generative Engine Optimization and answer-engine visibility.
Adding structured data is more approachable than the code makes it look. The process is the same for one page or a thousand:
Pick the schema type that matches the page (Product for a product, Article for a post, LocalBusiness for a location).
Generate the JSON-LD for that type, filling in the real values.
Paste the block into the page or push it through your tag manager.
Validate the markup before it goes live.
Request indexing in Search Console so Google recrawls the page.
You can generate the JSON-LD three ways: by hand, with an online generator, or with a WordPress plugin. Plugins are where non-developers stop worrying about code. Tools like Schemafy include an Auto Schema Generator that scans a site and suggests the right types per page, and an AI Schema Generator that produces the JSON-LD for you, so the markup gets created without hand-writing a line of it.
The step that scales badly is doing this page by page. On a single site it is a Saturday afternoon. Across 10 or more client stores, a per-page approach is why schema projects stall, which is exactly why the at-scale route matters.
How to test and validate your structured data
Two canonical tools cover almost everything. Google’s Rich Results Test tells you whether a page is eligible for a rich result. Schema.org’s Schema Markup Validator checks whether your syntax is correct against the vocabulary.
Understand the difference between the two words they return. “Valid” means the syntax is correct. “Eligible for rich results” means the page also meets Google’s specific requirements for a given rich result type. A page can be perfectly valid and still not be eligible, usually because a required property is missing.
For monitoring at scale, the Enhancements reports in Google Search Console track your markup across the whole site and flag errors by type. That is the report agencies live in, because it catches a broken schema on page 400 that no one would test by hand.
Common structured data mistakes to avoid
Most schema problems come from a short list of avoidable errors:
Marking up content that is not visible on the page. The markup must describe what the visitor actually sees.
Missing or mismatched required properties, so the markup validates loosely but fails eligibility.
Forgetting to update markup when the content changes, so the price or availability in your schema no longer matches the page.
Stacking multiple conflicting types on one page, which confuses parsers about what the page is really about.
That fourth one bites agencies hardest. When you manage many sites, stale markup piles up quietly. Build the habit of revalidating after any content change, and the problem never compounds.
Get rich-result-ready schema without the code
For teams that do not want to hand-write and validate JSON-LD across a whole site, a WordPress schema plugin can generate and maintain valid, rich-result-ready structured data automatically. That is the practical route when you are keeping schema correct across 10 or more client stores at once, without turning anyone on the team into a JSON-LD engineer.
Caption: Schemafy scans a site and suggests the right schema type per page, so structured data gets applied across many products without hand-coding.
Frequently asked questions
What is structured data in SEO?
In SEO, structured data is code (usually JSON-LD using the Schema.org vocabulary) added to a page so search engines understand its content and context. It does not directly boost rankings, but it makes pages eligible for rich results and helps them appear in AI-generated answers.
What is an example of structured data?
A common example is Product schema, which labels a page’s product name, price, availability, and review rating. When the markup is valid, Google can show those details, like a star rating and price, directly in the search result, making the listing more prominent and clickable.
What’s the difference between structured data and schema markup?
Structured data is the general concept of labeling content in a machine-readable format. Schema markup is the specific implementation that uses the Schema.org vocabulary. In short: all schema markup is structured data, but not all structured data uses Schema.org.
Does structured data help rankings?
Structured data is not a direct Google ranking factor. It can help indirectly by earning rich results that raise click-through rates, and by helping search engines and AI systems confirm what your page is about, which improves visibility in rich results and AI Overviews.
What is JSON-LD?
JSON-LD (JavaScript Object Notation for Linked Data) is a lightweight code format for adding structured data to a web page. It sits in a script block separate from visible content and is Google’s recommended format because it is clean, flexible, and easy to maintain.
How do I check if my structured data is working?
Use Google’s Rich Results Test to see if a page is eligible for rich results, and the Schema.org Markup Validator to check syntax. For ongoing monitoring across your whole site, watch the Enhancements reports in Google Search Console.
Final thoughts
Structured data stopped being a technical problem and became an operational one. On a single site, it is a nice-to-have. Across 10 or more client sites or stores, it decides whether your listings, and your clients’ listings, are eligible to look better in search and get confirmed by the AI systems that now answer questions directly.
Start where the return is highest. Pick one high-value type, apply it, validate it, and watch Search Console for the change. Organization and BreadcrumbList apply to almost every page, take little effort to set up once, and pay off site-wide, which makes them the right first move before you touch anything product-specific.
Most guides still tell you FAQ schema earns those dropdown questions under your result in Google. As of May 2026, it doesn’t. Google removed FAQ rich results from Search entirely. But the markup matters more than ever now, for a different reason: it is how machines and AI answer engines read your questions and answers. This guide covers what FAQ schema is, whether it still earns its keep, the JSON-LD to copy, and how to add and validate it.
What Is FAQ Schema Markup?
FAQ schema markup is structured data, in the FAQPage format from Schema.org, that labels a list of questions and their answers so search engines and AI systems can read them as discrete Q&A pairs. You add it as JSON-LD in the page’s code. It doesn’t change what visitors see.
That last point trips people up. FAQ schema doesn’t rewrite your content or redesign your page. It sits in a <script> tag and tells a machine “this string is a question, and this string is its answer.” Nothing more.
It is also not a ranking factor. Google has never treated FAQ markup as a signal that lifts your position. What it does is remove ambiguity about what your content means, which is exactly what both search crawlers and AI answer engines need.
JSON-LD is the format to use. It has been Google’s recommended way to add structured data since 2015, and it is a W3C recommendation, so you are not betting on a proprietary syntax. FAQ (FAQPage) is one of the schema types you can generate with a plugin instead of hand-coding, the same operational shift covered in how to use schema markup more broadly. That matters once you have more than a page or two to mark up.
FAQPage vs QAPage: Which One Do You Need?
These two types look similar and get mixed up constantly. The difference is who writes the answers.
Use FAQPage when your site publishes both the question and the single, authoritative answer. Think a support page, a product FAQ, a policy explainer. Use QAPage when users can submit their own answers to a question, like a forum thread or a community Q&A. Google is explicit: if there is one answer and users can’t add alternatives, it is FAQPage.
Type
Use it when
Who writes the answers
Example
FAQPage
One authoritative answer per question
You (the site)
Product FAQ, support page, policy page
QAPage
Users can submit alternative answers
Your users
Forum thread, community Q&A
Pick wrong and your markup describes something the page isn’t, which is one of the fastest ways to get structured data ignored.
Does FAQ Schema Still Work in 2026?
Yes and no, and the distinction is the whole point of this section.
The rich result is gone. According to Google’s own documentation, as of May 7, 2026, FAQ rich results no longer appear in Google Search. Google is dropping the FAQ search appearance, the rich result report, and support in the Rich Results Test in June 2026, with the Search Console API following in August 2026.
As of May 7, 2026, FAQ rich results are no longer appearing in Google Search. Support in the Rich Results Test is being dropped in June 2026. Source: Google Search Central documentation
So if your reason for adding FAQ schema was the expandable questions under your listing, that reason expired. Don’t build a page around a feature that no longer renders.
The markup itself still does real work. The FAQPage format turns a wall of text into clean question-answer pairs that machines can lift without guessing. That is valuable to a different audience than it used to be.
The impact was immediate. After the change, roughly 82.96% of sites with FAQ schema stopped earning FAQ results, up from 46.06% in February of that year, per Search Engine Land. A separate study measured a 37% drop in FAQs appearing in results.
So who still gets the dropdowns today? Nobody. The 2023 change narrowed eligibility to gov and health sites; the May 2026 change removed the appearance entirely. There is no tier of site that earns the classic FAQ rich result anymore.
The Real Payoff Now: AI Overviews, ChatGPT & Perplexity
Here is where the value moved. AI answer engines read structured Q&A pairs and reuse them.
When Google AI Overviews, ChatGPT, or Perplexity assemble an answer, pre-formatted question-answer pairs are easy to extract and attribute. Your FAQPage markup hands them the exact text you want associated with each question, instead of making them infer it from prose. Google now publishes a dedicated guide to optimizing for generative AI search, which tells you where its own priorities sit.
Practitioners tracking answer engine visibility report that FAQ-structured content gets pulled into AI answers more often than the same content left unstructured, a pattern the AEO and GEO literature keeps returning to. Treat the specific percentages floating around with caution, but the direction is consistent. If you care about being cited in AI answers, this is the discipline: it is part of generative engine optimization, and clean FAQ markup is one of its cheapest inputs.
The goal changed from “earn the dropdown” to “be the source the machine quotes.” FAQ schema still serves that goal.
How to Add FAQ Schema Markup (Step by Step)
The process is four steps whether you hand-code the JSON-LD or generate it. Write the FAQ, produce the markup, place it on the page, then validate.
Write a real, visible FAQ on the page.
Generate the FAQPage JSON-LD.
Paste it into the page’s code.
Validate the syntax and monitor it after edits.
You have two honest options at step 2: write the JSON-LD by hand (the code is below) or generate it. If you have one page, hand-coding is fine. If you have dozens or hundreds, generating and maintaining it is the only sane path.
Step 1: Write a Real, Visible FAQ First
Google’s guidelines are strict on one thing: every question and answer in your markup must be visible to the user on the page. The answer can sit behind an expandable accordion, but it has to be reachable. You cannot mark up content that only exists in the JSON-LD block.
That rule is not bureaucratic. The markup is supposed to describe the page, so write the FAQ as real page content first. Use questions people actually ask, and answer each one directly in one or two sentences.
Get the human-readable version right before you touch any code. The markup mirrors it.
Step 2: Generate the JSON-LD
Hand-writing FAQPage markup for one page takes a few minutes. Doing it for 400 product pages does not scale, and every manual edit is a chance to desync the markup from the visible text.
To generate it for a single page in WordPress, open WP Admin → Schemafy → AI Schema Generator, search for the page or paste its URL, choose FAQ under Schema Type, and click Generate Schema with AI. Open Review technical JSON-LD to read the output, check the validation indicator, and click Save to Website. If the draft isn’t right, Regenerate.
To do it across a whole site, open WP Admin → Schemafy → Auto Schema Generator and click Scan Site. Filter Post Type and set Status to Needs Schema, then review the suggested schemas and match percentages before applying in bulk. This is how you mark up hundreds of pages without opening each one.
Schemafy generating FAQPage JSON-LD for a single page from the AI Schema Generator, with the output validated before it goes live.
If you prefer to fill fields yourself, Open Schemafy and navigate to Smart Builder. Select the FAQ schema template, enter each question and answer into the corresponding FAQ fields, and review the generated FAQPage JSON-LD output. Before publishing, verify that all questions, answers, and schema properties match the current Schemafy interface and your page content, then save the schema. You can also generate a block in the browser with a standalone schema markup generator if you just want the code.
Step 3: Paste It Into Your Page
The JSON-LD goes inside a <script type="application/ld+json"> tag. Google accepts it in either the <head> or the <body>, so placement is flexible.
In WordPress, a schema plugin applies the markup to the page for you, so you never edit the theme. To paste your own block instead, Open Schemafy and navigate to Manual JSON. Paste your FAQPage JSON-LD markup into the JSON editor, select the page or URL where the schema should be applied, and review the configuration before saving. Confirm that all field names, menu paths, and save actions match the current Schemafy interface before publishing the schema.
Once schemas are live, you can view and edit everything applied across the site from the Rich Snippets screen, which keeps you from losing track of which page has which markup.
Step 4: Validate and Monitor
Validate the syntax before you publish. A single missing bracket or misnamed property makes the whole block unreadable to a crawler.
One honest caveat about tooling: Google is retiring the FAQ report and FAQ support in its Rich Results Test during 2026, so don’t rely on that specific tool to confirm FAQ markup going forward. Use a general JSON-LD validator instead. A JSON-LD editor and validator checks your block against the Schema.org and Google structured data specs in real time.
Then monitor. The most common way FAQ markup breaks is drift: someone edits the visible answer and the JSON-LD keeps the old text. Recheck after any content change so the markup still matches what a visitor reads.
FAQ Schema JSON-LD Code (Copy-Paste)
Here is a minimal, valid FAQPage block based on Google’s official example. Replace the questions and answers with your own, keep the structure, and paste it into your page.
Basic FAQPage Example
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "FAQPage",
"mainEntity": [{
"@type": "Question",
"name": "Does FAQ schema still earn rich results in Google?",
"acceptedAnswer": {
"@type": "Answer",
"text": "No. As of May 2026, Google no longer shows FAQ rich results in Search. The FAQPage markup still helps search engines and AI systems understand your questions and answers."
}
}, {
"@type": "Question",
"name": "Which JSON-LD format should I use for FAQs?",
"acceptedAnswer": {
"@type": "Answer",
"text": "Use the FAQPage type with a mainEntity array of Question items, each containing a name and an acceptedAnswer of type Answer with a text property."
}
}]
}
</script>
The required properties are small in number. FAQPage needs mainEntity, an array of Question items. Each Question needs a name (the full question text) and an acceptedAnswer. Each Answer needs a text property with the full answer. Everything else is optional.
WordPress, Webflow & Next.js/React
The JSON-LD is identical across platforms. Only the placement differs.
On WordPress, a schema plugin like the Schemafy plugin injects the block for you, so you don’t paste anything into the theme or risk an update wiping it out.
On Webflow, paste the <script> block into the page’s custom code (Page settings → Custom code, before </body>) or into an Embed element on the page.
On Next.js or React, render the <script type="application/ld+json"> tag in the head. In the App Router you can return it from a component or the metadata export; in the Pages Router use next/head or next/script. Serialize the object to a JSON string with JSON.stringify and set it as the script’s content rather than as children, so it isn’t escaped.
[SCREENSHOT: The Schemafy Rich Snippets management screen showing a list of applied schemas with a FAQPage entry attached to a WordPress page.]
A FAQPage schema applied to a WordPress page, managed alongside every other schema from the Rich Snippets screen.
Google’s FAQ Schema Guidelines & Character Limits
Google’s FAQ documentation sets a short list of required properties. Get these right and your markup is technically valid.
Type
Required property
What it holds
FAQPage
mainEntity
An array of Question items (at least one)
Question
name, acceptedAnswer
The full question text, and its answer
Answer
text
The full answer text
Inside the Answertext, Google supports a limited set of HTML tags: <h1> through <h6>, <br>, <ol>, <ul>, <li>, <a>, <p>, <div>, <b>, <strong>, <i>, and <em>. Any other tags are ignored, so don’t put custom markup or scripts in an answer.
The content rules matter as much as the syntax. Every question and answer must be visible on the page. Use one answer per question, or switch to QAPage. Don’t use FAQPage for advertising copy. And if the same FAQ appears on many pages, mark up only one instance for the whole site, per Google’s structured data guidelines.
On character limits: Google does not define one in the FAQPage spec. There is no maximum length for a question or an answer. The real constraint is that the markup must contain the complete text of each question and answer, and that text must match what is visible on the page. Concise answers are still a good idea, since a tight one or two sentences is easier for an AI engine to quote cleanly, but that is best practice, not a hard rule.
Common FAQ Schema Mistakes to Avoid
Most FAQ markup problems come down to a handful of repeat offenders. Google’s guidelines rule out several of these directly.
Marking up content that isn’t visible on the page. The JSON-LD has to describe text a visitor can actually see.
Using FAQPage where QAPage belongs, on pages where users submit their own answers.
Using FAQ markup for promotional or advertising copy rather than genuine questions.
Duplicating the same FAQ block across many URLs. Mark up one instance for the site.
Letting the JSON-LD drift out of sync with the visible question and answer text after an edit.
Building the page around the old dropdown rich result that no longer exists.
The drift problem gets worse with scale. If you maintain FAQ markup across hundreds of pages by hand, some copy will fall out of sync eventually. Generating and managing it from one place, then reviewing coverage from the Auto Schema Generator and Rich Snippets screens, keeps markup and visible text aligned without a manual audit.
FAQ Schema Best Practices for 2026
The best practices follow directly from where the value now sits.
Write real, concise, visible questions and answers, one clear idea per answer. Keep the markup identical to the visible text after every edit. Generate and maintain FAQPage at scale instead of typing it, using the Auto Schema Generator for bulk coverage and the AI Schema Generator for individual pages. Validate the syntax before publishing.
Then think one step further than search. When you write each answer, picture an AI engine lifting that single question-answer pair into a response. If the answer stands on its own in two sentences, it travels well. If it only makes sense in the surrounding paragraph, tighten it.
That mindset, clean pairs that survive being quoted out of context, is what separates FAQ markup that still earns attention in 2026 from markup that just sits in the page unused.
Schemafy’s Auto Schema Generator flagging pages that need schema and suggesting FAQPage across a whole site in one scan.
Ready to Win AI Citations With Structured Data?
The dropdown rich result is gone, so the job changed. The work now is making sure search engines and AI answer engines can read your questions and answers as clean, quotable pairs, which comes down to correct FAQPage markup that mirrors your visible content.
Write a real FAQ, mark it up properly, validate the syntax, and then generate and maintain the FAQPage (and the other schema types your pages need) instead of hand-coding each one.
[CTA_DOWNLOAD]
Frequently Asked Questions
Does FAQ schema help SEO?
FAQ schema is not a ranking factor, and as of May 2026 it no longer produces a rich result in Google Search. It still helps search engines and AI systems understand your questions and answers as structured pairs, per Google’s documentation.
Is FAQ schema still worth it in 2026?
Yes, but for structural clarity and AI-search citation rather than for the old expandable questions. If you want your Q&A content to be readable and quotable by AI answer engines, FAQPage markup is a cheap, low-risk input.
What’s the difference between FAQPage and QAPage?
FAQPage is for one authoritative answer per question, written by the site. QAPage is for pages where users can submit their own answers, like a forum or community thread.
How many FAQs should I add per page?
Google doesn’t set a number. Include only the questions that are genuinely relevant to the page, keep every one visible to the reader, and keep the answers concise. Padding a page with questions no one asks helps no one.
Where do I put FAQ schema code?
Inside a <script type="application/ld+json"> tag, in either the <head> or the <body> of the page. On WordPress, a schema plugin injects it for you so you don’t edit the theme.
