search intent · definition
Search intent: the job the searcher wants done
The four labels are the easy part. What matters is reading intent off a live results page, picking the page type that fits, and knowing why a well-written article can rank for nothing when the format is wrong.
Search intent is the goal behind a query — what the person actually wants to do or know — which search engines infer from behaviour and express through the kinds of pages they choose to rank for it.
reviewed 2026-09-02 · by the IT Master editorial team · how we check facts
What search intent is in practice
Search intent is the job a person wants done when they type a query. The standard four labels are informational (learn something), navigational (reach a specific site), commercial (compare options before buying) and transactional (buy, book or download now).
The labels are shorthand. The useful work is reading intent off a live results page, because the search engine has already decided what intent it thinks the query carries and has ranked pages accordingly. Look at what the top ten actually are. Ten how-to guides means informational. Ten category pages with prices and filters means transactional. A mix of buying guides and product roundups means commercial.
That reading tells you the format before it tells you the words. A query answered by product listings will not accept a 2,000-word essay, however good the essay is. A query answered by tutorials will not accept a category page. Classification tools hand you a label; the results page hands you evidence, and where the two disagree, the evidence wins.
Search intent is the goal behind a query — what the person actually wants to do or know — which search engines infer from behaviour and express through the kinds of pages they choose to rank for it.
Why it matters in 2026
Two things sharpened the cost of getting intent wrong.
The first is that a mismatch behaves like an exclusion rather than a penalty. Search engines appear to draw their candidates from the page type they judge the query wants, so a page of the wrong type usually does not rank badly — it never enters the running. On-page tuning does not move it, because the problem is the format and not the copy.
The second is that assistants split queries. Asked a question, ChatGPT or Perplexity typically rewrites it into several narrower searches and reads what those return. The rewrites carry intent of their own, usually informational or comparative, so a site holding only transactional pages can be missing from the retrieval step even when it sells the exact product under discussion. AI visibility is how you find out whether you are being read at all.
The practical shift is one of order: decide the page type from the evidence, then write. Doing it the other way round is how good articles end up ranking for nothing.
An intent mismatch behaves like an exclusion, not a penalty: a page of the wrong type usually does not rank badly, it never enters the running, and no amount of on-page tuning changes that.
How we handle intent in the pipeline
Topic discovery starts from demand rather than from a hunch: keyword and SERP data from DataForSEO, plus the queries a site already earns impressions for in Search Console. Research then reads the pages currently ranking for the term, alongside forum threads and the customer's own first-party data — product records, datasheets, support Q&A — which is the same evidence a human editor would use to decide what shape the page has to be.
The honest limit: the validation gauntlet checks facts, novelty against competitors, E-E-A-T signals and AI tells, using models from a different vendor than the writer, across up to three revision rounds. It is not an intent oracle. A page that answers the wrong question well can still clear every check.
Intent is settled when the topic and its target query are chosen, which is why that choice earns the two minutes of looking at a live results page. How it works sets out the full sequence.
Common misunderstandings
- That every query has exactly one intent. Plenty are mixed. A results page holding three guides, three category pages and a video is telling you the engine is hedging, and the page that wins usually serves the dominant intent while covering the second one in a section.
- That intent is stable. It drifts. A query that returned reviews last year can return listings this year once the product becomes a commodity. Re-check the results page before refreshing an old article, not afterwards.
- That transactional queries are the only ones worth having. Informational pages are what assistants retrieve and what earns links, and they feed the topic cluster that makes the commercial pages credible.
- That matching intent means copying page one. Matching the format is required. Copying the substance is what a novelty score rejects, and it also leaves the page with no reason to be chosen over the nine already ranking.
- That a keyword tool settles it. Its labels are inferred from wording patterns. Two minutes looking at the live results beats any classifier.
questions people ask
What are the four types of search intent?
Four labels cover most cases. Informational queries want to learn something, such as how PoE cameras are powered. Navigational queries want one specific site or page, usually signalled by a brand name. Commercial queries are comparing before buying, and show up as best, versus and review phrasing. Transactional queries are ready to act now: buy, price, book, download. Treat the labels as a starting point rather than a rule, because plenty of queries sit between two of them, and the live results page is the better authority on which one the engine believes applies.
How do you work out the search intent of a keyword?
Search it and look at what ranks. The top ten are the search engine's own answer to the question, so count the page types you see: guides, category pages, product pages, tools, videos. Whichever dominates is the intent you have to serve. Then read the titles and the People Also Ask box, which expose the sub-questions attached to the query. Keyword tools attach intent labels by pattern matching, which is useful for sorting a long list quickly, but where a label and a live results page disagree, believe the results page.
What happens if a page does not match search intent?
Usually it does not rank at all, rather than ranking poorly. Search engines pick candidates that fit the format they judge the query wants, and a page of the wrong type tends not to enter that set. The symptom is a page that looks well optimised, has no technical fault, and collects impressions for nothing close to its target term. The fix is rarely more words or better headings. It is either changing the page type or pointing that page at a different query it genuinely answers.
Does search intent matter for AI assistants and AI Overviews?
It matters more, because assistants typically rewrite a question into several narrower searches before fetching anything. Those rewrites tend to be informational or comparative even when the person asking is ready to buy, so a site holding only product and category pages is missing from the pages an assistant reads. Answering the underlying question plainly, on a page whose format matches the rewritten query, is what puts you in the retrieved set. Citation comes after retrieval, and retrieval depends on having a page shaped like the answer.
See what search engines and AI assistants find on your site
Free, no account. Type your address and we show you what is missing and what we would write first.