Search Intent: Classifying Keywords Before You Write a Word
A team spends three weeks producing a comprehensive buyer’s guide for a high-volume commercial keyword. The domain is authoritative, the content is thorough, the on-page optimization is textbook. It ranks at position 11 for six months and never moves. The problem is not the content quality — it is that Google consistently ranks product category pages and comparison grids for that query, because the SERP tells anyone willing to look that the keyword is transactional, not informational. The guide was the wrong asset for the job.
This is the #1 reason well-optimized pages fail to rank: intent mismatch. The keyword is there, the links are there, but the page type is wrong for what the searcher actually wants.
The Four Intent Classes
The taxonomy that has stuck — and that Google effectively encodes in its ranking systems — has four categories. They are not mutually exclusive, and the interesting work happens at the boundaries.
Informational intent covers queries where the user wants to understand something. “How does a CDN work,” “what is anchor text,” “symptoms of iron deficiency” — these queries are served by guides, articles, explainers, and reference pages. The user has a question; they want an answer or an explanation. Ad conversion rates are low on these queries; that’s not the point. The point is establishing topical authority and capturing top-of-funnel traffic that may convert later.
Navigational intent is the simplest category to understand and the hardest to compete for. “Gmail login,” “Stripe dashboard,” “New York Times” — the user already knows where they want to go. They are using Google as a URL bar shortcut. You will not outrank the target brand on its own navigational queries. The practical implication is that navigational queries in a keyword set are either about your own brand (in which case you should be ranking by default) or they indicate a user loyalty signal about a competitor that is worth tracking.
Commercial investigation intent — sometimes called commercial or research intent — is where most B2B and considered-purchase buying cycles live. “Best project management software,” “Ahrefs vs SEMrush,” “enterprise SEO platform reviews” — these queries come from people who intend to buy something but have not yet decided what. They want comparison, evaluation, and social proof. The winning content types here are comparison pages, roundups, review posts, and “X alternatives” content. A product page ranks poorly for these queries; a feature-comparison article ranks well.
Transactional intent is the bottom of the funnel. “Buy noise-canceling headphones,” “SEMrush coupon code,” “sign up for Shopify” — the user has made a decision and is ready to act. Product pages, category pages, landing pages, and signup flows are the content types Google rewards here. Long-form educational content is the wrong answer.
Reading Intent from the SERP
The fastest way to classify intent for a specific keyword is to look at what Google already ranks for it. This is not a shortcut — it is the ground truth. Google’s ranking systems have been trained on user behavior at enormous scale; the SERP composition is the most direct signal available about what users actually want when they type that query.
When you search a keyword and see a featured snippet answering a specific question above a list of how-to guides and Wikipedia articles, the dominant intent is informational. When you see a product grid (the Shopping unit) above organic results, the dominant intent is transactional. When you see comparison posts and roundups from review sites like G2, Capterra, or Wirecutter occupying the top positions, the dominant intent is commercial investigation. When you see the target brand’s own pages dominating — especially their homepage or login page — it is navigational.
SERP features are intent signals too. A “People Also Ask” box suggests informational intent with multiple sub-questions. A local pack indicates local transactional intent. Video results in organic position suggest how-to or entertainment intent. Rich results for product reviews point to commercial investigation or transactional intent depending on the overall result composition.
The practical exercise: before finalizing the content type for any target keyword, run the query, note the format and content type of the top five organic results, and note which SERP features appear. If four of the top five are product category pages and a Shopping unit is present, building an informational article is not a viable ranking strategy regardless of how good the article is.
Intent Signals in the Query Itself
SERP analysis is the authoritative method, but at the keyword-research stage — before you have a ranked SERP in front of you for every candidate keyword — query modifiers are useful heuristics for pre-classification.
Modifiers that strongly suggest informational intent: “how to,” “what is,” “why does,” “guide to,” “tutorial,” “explained,” “definition of,” “examples of,” “history of,” “can you.” These words appear in queries where the user is in learning mode.
Modifiers that suggest commercial investigation: “best,” “top,” “vs,” “versus,” “alternatives to,” “review,” “reviews,” “comparison,” “ranked,” “compared,” “pros and cons,” “worth it.” These words appear when someone is evaluating options.
Modifiers that suggest transactional intent: “buy,” “cheap,” “discount,” “coupon,” “deal,” “order,” “price,” “pricing,” “near me,” “for sale,” “in stock,” “download,” “free trial,” “sign up,” “get started.” Action words that indicate readiness to convert.
Brand name queries without additional context are typically navigational. “Slack” is navigational. “Slack pricing” slides toward commercial investigation. “Slack enterprise plan” is commercial investigation with transactional overtones. The modifiers shift the classification along the funnel.
At scale, these heuristics can be encoded as a bitmask classifier — which is how tools like DataForSEO and SEMrush represent intent in their API responses. The Fk field in SEMrush’s keyword data uses bit flags: 1 for informational, 2 for navigational, 4 for commercial, 8 for transactional. A value of 5 (1+4) indicates a keyword classified as both informational and commercial. That bitmask approach is useful for bulk classification but should be validated against SERP observation for any keyword that will drive a content investment. Keyword research and cluster strategy covers how to organize these classifications into workable content plans.
Mixed Intent and Blended SERPs
Not all queries have a clean dominant intent, and the SERPs for mixed-intent queries are instructive. “Best laptop” produces a blend: some roundup articles from tech publications, some product category pages from retailers, and sometimes a featured snippet with a quick recommendation. Google is not sure whether the user wants an editorial recommendation or a shopping experience, so it serves both.
Blended SERPs are both an opportunity and a diagnostic signal. They indicate that a range of content types has a realistic shot at ranking, and that Google will surface the content that best satisfies whichever intent segment the individual user represents. A comparison guide can rank on the same SERP as a product category page because different users clicking the same query have slightly different objectives.
The error is treating a blended SERP as license to ignore intent. If the top three positions are all roundup articles and you build a product page, you are still competing against the wrong format for the majority of clicks. Blend means “multiple formats can rank,” not “any format can rank equally.”
When you see a blended SERP, the strategy is to identify which intent segment represents the higher business value and build the strongest possible content for that intent, while structuring it in a way that can plausibly serve the adjacent intent. A comparison guide that leads with a clear recommendation (satisfying commercial investigation intent) and includes links to product pages (acknowledging transactional intent) can perform across the blend.
Mapping Intent to Page Type and Funnel Stage
The practical output of intent classification is a page-type decision: what kind of page should exist to serve this keyword?
Informational → blog post, guide, tutorial, reference page, FAQ. These feed top-of-funnel awareness and can be linked internally to commercial or transactional pages to support conversion paths.
Commercial investigation → comparison page, feature breakdown, “X vs Y” page, category-level buyer’s guide, review aggregation page. These pages exist to help users decide; they should link directly to the pages where users can take action.
Transactional → product page, category page, landing page, pricing page, signup flow. These should minimize friction; every element should support the conversion action.
Navigational → your brand pages. Dominating your own navigational queries is about brand health: having a strong homepage, clear sitelinks, and consistent entity signals across GSC and structured data.
Funnel mapping matters because it determines internal linking architecture. Informational pages should link to commercial investigation pages for users who want to go deeper. Commercial investigation pages should link to transactional pages for users who are ready to act. Internal linking and site structure covers how to build this path systematically, but the foundation is intent classification: you cannot build the right link paths if you have not classified your pages by the stage they serve.
Classifying at Scale
Individual keyword classification by SERP analysis is the right method for high-investment decisions — tier-1 keywords driving major content builds. But keyword sets often number in the thousands, and manual SERP analysis at that scale is not practical.
Scalable classification uses a layered approach. Start with the API-level intent bitmask from your data provider for initial sorting and triage. Apply modifier-based heuristics to identify obvious informational and transactional buckets. Manually validate SERP composition for every keyword above your traffic volume threshold — whatever that is for your domain, whether 500 monthly searches or 5,000. For keywords below that threshold, the API classification plus modifier heuristics is usually sufficient because the content investment decision is smaller.
Content decay detection and refresh strategy is the downstream problem: pages that ranked well for a keyword whose dominant intent has since shifted, or pages that were built for the wrong intent and have been losing ground. Intent classification at the outset is far less expensive than diagnosing and fixing intent mismatches on pages that have already accumulated link equity.
VisibilityIQ’s keyword workbench surfaces intent classification alongside keyword metrics, so content decisions are tied to intent data rather than volume alone. When a page’s ranking for a keyword starts declining, the first diagnostic question is whether the SERP intent composition has shifted — Google updates its ranking systems continuously, and queries that once rewarded informational content sometimes shift toward transactional as commercial activity around a topic grows.
Intent as a Constraint, Not a Factor
The framing that most teams miss: search intent is not a ranking factor to optimize for — it is a constraint on what can rank. You do not optimize for intent the way you optimize for page speed or heading structure. You either match the dominant intent or you do not. If you match it, you are eligible to compete on the other factors. If you do not match it, the other factors are largely irrelevant.
This reframes the content strategy decision. The question before writing is not “how do I optimize this page for the keyword?” The question is “what type of page does Google reward for this keyword, and do I have that type of page?” If the answer is no, the keyword either needs to be paired with the right asset type or it needs to be deprioritized until the right asset is built.
Anchor text distribution, by the way, reflects intent indirectly: when a page earns natural editorial links, the anchor text people choose tends to reflect what they think the page is about. A page genuinely serving informational intent earns anchors like “guide to,” “how to,” and the topic phrase. A product page earns anchors like the brand name and product category. Anchor text distribution and intent alignment are two faces of the same question: is this page clearly what it claims to be?
The discipline of classifying intent before producing content sounds obvious written out. In practice it is bypassed constantly — by teams chasing volume without checking the SERP, by content calendars built around keyword lists rather than keyword lists grouped by page type, by marketers who produce informational content for transactional queries because informational content is easier to write. Intent classification is the prerequisite that makes everything downstream tractable.