Search Intent Mapping: Match Your Page Type to the SERP

Search intent mapping means labeling each query by what the searcher is trying to do, then checking what the results page already rewards and building the page type that fits. Google's own Search Quality Rater Guidelines (the September 11, 2025 edition, the file at the public URL when I fetched it on 13 November 2026) sort intent into four groups: Know, Do, Website and Visit-in-person. This guide turns that into a repeatable process.

Key Takeaways

• Google's rater guidelines describe four intents: Know (with a Know Simple subtype), Do, Website and Visit-in-person. The same document says many queries fit more than one and the Do/Know line is "sometimes blurry".
• Rater scores do not rank pages. The guidelines state that no single rating can directly change how a page appears. Use them as a statement of what Google is trying to measure, not as a ranking formula.
• Google's how Search works page says results depend on the query and the searcher's context, including location, language and device. Map intent per market, not once globally.
• The SERP itself is the best evidence. Google's visual elements gallery lists the result types (text, rich, image, video, related questions), and which ones appear varies by query.
• A page that is well written but built as the wrong type (a long guide for a Do query, a product page for a Know query) is a mismatch no on-page tweak fixes.
• Google's helpful content guidance asks whether a reader will "learn enough to achieve their goal". That is the intent test in plain words.

What Google's own documents say about intent

Google's public documentation does not publish a ranking intent taxonomy for site owners. The closest primary source is the Search Quality Rater Guidelines, which humans use to evaluate whether a result meets the need behind a query. Section 12.7 of the September 11, 2025 edition lists the following.

Two caveats from the same document matter for practice. First, the guidelines say the Do and Know categories overlap and that they are "big-picture patterns", not a strict sort. Second, they say a query can have several interpretations and that the dominant interpretation can change over time. Their own example is the query "iphone", where most searchers once wanted the first model and now want the newest.

The ranking side is documented separately. Google's guide to ranking systems describes BERT as helping Google understand how combinations of words express different meanings and intent, and lists freshness systems that show newer content when the query calls for it. So intent is handled by named systems, and you cannot toggle it. You can only match it.

The five-step method

Step 1: Pull the real queries

Export queries from the Search Console Performance report for the page, or for the topic if you are planning new content. Keep impressions and clicks. If you have no data, write the five phrasings a customer would use, including the clumsy ones. Do not rely on a keyword tool's intent label alone; treat it as a first guess.

Step 2: Label each query with one primary and one secondary intent

Use the four groups above. Add a fifth working label for yourself, "Compare", for queries like "X vs Y" or "best X", because they behave like Know queries with a decision attached. Mark queries where you are unsure. Those get the most SERP scrutiny.

Step 3: Read the live SERP, not a tool's summary

Search the query in a clean browser session for the market you care about. Record what you see, using the names from Google's visual elements gallery:

• How many text results are guides, product pages, category pages, tools, forum threads, or videos
• Whether image results, video results, or a related questions group are present
• Whether rich results (for example product, recipe or review snippets) appear, since Google says these typically rely on structured data on the page
• Whether any AI-generated answer sits above the results, and which kinds of pages it links to

Do this for each market. Google's own example is that a bicycle repair query returns different results in Paris than in Hong Kong.

Step 4: Decide the page type

Match the dominant result type, then look for the gap you can fill. The table below is the decision aid I use.

If the page types on the SERP are all of one kind, a different type will rarely be shown. That is a pattern, not a rule, so test it with one page before reorganizing a section.

Step 5: Choose the action

For each query group there are four options: create a new page, rewrite the existing page to fit, merge two pages that serve the same intent, or leave it alone. Leave it alone when the page already matches the SERP type and the issue is elsewhere, such as indexing. Merge when two of your pages target the same intent; the canonical tag guide covers how to consolidate them without losing signals.

A hypothetical worked example

The example below is invented to show the method. It is not a client and the SERP descriptions are assumptions for illustration. Check the live results before you copy any decision.

Suppose a small software company sells invoicing software and has one page, a 3,000-word blog post, trying to rank for four queries.

The point is that one long post cannot serve four intents. The honest result is one page kept, one page created, and two queries moved to new pages. Each new page links to the others, which also supports the topic structure described in Topical Authority Strategy Without a Content Team.

What this means for marketers and founders

• Start with pages that already get impressions but few clicks. Google's traffic debugging guide says that when impressions hold and clicks fall, your title link, snippet or rich results may look less appealing than competitors'. The title test workflow applies there.
• Run the SERP read before writing a brief. The content brief template for AI search works better when the page type is already decided.
• Treat AI answers as another SERP element, not a separate world. Google says AI features may issue multiple related searches across subtopics. That is the idea covered in Query Fan-Out Explained, and it makes sub-question coverage part of intent matching.
• Do not rewrite a page that matches the SERP. If the type fits and the page is indexed, changes should be small and measured.

Common mistakes

• Using one intent label from a keyword tool and never opening the SERP.
• Treating the rater guidelines as a ranking checklist. They are an evaluation document and say ratings do not move individual pages.
• Building a bigger page when the SERP wants a different type of page.
• Ignoring market differences. A query can have a Visit-in-person reading in one country and a Know reading in another.
• Merging pages by intent without redirects or canonical tags, which leaves two URLs fighting each other.

Caveats

I could not find a Google document that publishes a mapping from intent to page type, so the page-type table above is my working method built from the rater guidelines and the visual elements gallery. It is a heuristic. The rater guidelines are edited periodically and the date I cite may have been superseded by the time you read this; check the current edition before quoting section numbers.

FAQ

What is search intent in SEO?

Search intent is the goal behind a query: to learn something, to do something, to reach a particular site, or to visit a place. Google's rater guidelines describe four groups (Know, Do, Website, Visit-in-person). Matching your page to the goal is usually more important than matching the exact words.

How do I find the search intent of a keyword?

Search the keyword in a clean browser for your target market and note the result types that appear. Add the queries people actually use from Search Console. If the SERP shows mostly guides, treat it as Know; mostly product pages, as Do; mostly one brand, as Website.

Does Google use the Know, Do, Website labels in ranking?

Not as published ranking factors. These labels appear in the rater guidelines, and Google states that no single rater rating can directly change how a page appears. They explain what Google wants results to achieve, not how the system scores pages.

Can one page target more than one intent?

It can, when the intents are close, for example a how-to guide with a template download. When the intents need different page types, such as a comparison and a tutorial, separate pages usually work better and each can link to the other.

How often should I recheck intent?

Check on any page whose traffic changes sharply, and review priority pages each quarter. Google's guidelines note that the dominant meaning of a query can change over time, so a stale label can lead to a wrong rewrite.

Do AI answers change how I should map intent?

They add another place where your content might be shown, but the mapping steps are the same. Google says AI features can run related searches across subtopics, so cover the sub-questions inside a Know page and keep the answer easy to extract.

What is the difference between Know and Know Simple?

Know Simple queries ask for a specific short answer that fits in a small space, such as a fact. Know queries are broader and need explanation or several viewpoints. The guidelines treat them as a spectrum, so use the label as a guide, not a rule.

Want help with this?

I'm Younus Fardeen, an organic growth, SEO and AEO strategist with 4+ years of marketing experience across edtech and startup teams. If you want help mapping search intent and matching page types for your site, see my work and reach me through the contact form at younusfardeen.com.