Search Intent: Matching a Page to What People Want

A page covers the keyword thoroughly, reads well, loads fast, and never ranks. The usual reason is not quality. It is that the page answers a different question from the one people were asking when they typed those words. Search engines are very good at recognising what a query is for, and a page that does not match that purpose competes for a position it was never eligible for.

The four kinds of query

Nearly every search falls into one of four groups, and each wants a different kind of page.

• Informational. Someone wants to understand or do something: "how to clean a cast iron pan". The winning page explains, in the order the task happens.
• Navigational. Someone wants a specific site or page: a brand name, a login. You cannot rank for someone else's navigational query, and you should not try.
• Commercial investigation. Someone is deciding: "best accounting software for freelancers", "X versus Y". The winning page compares, with criteria and honest trade-offs.
• Transactional. Someone is ready to act: "buy", "book", "sign up", often with a product name. The winning page lets them do it immediately.

The same subject supports all four. "Running shoes" as a guide, as a comparison, and as a product page are three different pages, and trying to serve all three on one page usually serves none of them.

Read the results page as a brief

You do not have to guess what a search engine thinks a query means. The results page shows you, because it is the outcome of a great deal of measurement. Search the query, ignore the ads, and note four things.

What format dominates the first page: step-by-step guides, product pages, comparison lists, videos, forum threads. What angle the titles take, which tells you what the audience is worried about. Which features appear: a video carousel means people expect to watch, a local pack means the intent has a geographic component, product listings mean the query is commercial whatever it looks like. And how deep the results go: if the top pages are short answers, a four-thousand-word essay is not a competitive advantage, it is a mismatch.

If the first page is entirely forum threads and personal accounts, the query wants experience, not a company explainer. That is a signal to write from real practice or to pick a different query.

Match the format before you polish the words

A comparison query needs a comparison: named alternatives, the criteria that separate them, and a recommendation you are willing to defend. An informational query needs the steps in order, with the prerequisites stated. A transactional query needs price, availability and a button, not three paragraphs of introduction.

Word count is not a ranking factor, and treating it as one is how thin pages get padded into long thin pages. Length should be whatever the question needs. The useful discipline is the opposite of padding: can a reader get the answer on the first screen, with the detail available below for those who want it?

One page per intent

When two of your pages target the same intent, they compete with each other. Search engines pick one, often not the one you would have chosen, and both pages end up weaker than a single merged page would have been.

Find these by listing your pages against the queries they rank for, in whatever search console your site has. Pages that rank for overlapping queries and rank poorly for all of them are candidates to merge: keep the stronger URL, fold in what the other one did better, and redirect.

The opposite problem is one page trying to serve a guide, a comparison and a product page at once, usually because it grew. Splitting it into three pages, each with a clear job, and linking them in the order a reader moves through them, tends to lift all three.

Signals that you missed

• Impressions without clicks. The page is eligible but the title and description do not match what the searcher wanted. Sometimes that is a title fix; often it is an intent mismatch the title is papering over.
• Clicks that bounce straight back. People arrive, see the page is not what they needed, and return to the results.
• A ranking stuck around the bottom of page one. The page is relevant but competing against a different format. Look at what the pages above it do differently in shape, not in length.
• Ranking for the wrong queries. A useful hint: the page is well built for an intent you did not plan for. Sometimes the right move is to lean into that.

A workflow you can repeat

Before writing, search the query and write one sentence describing what the results page says the searcher wants. Decide which of the four intents it is. Choose the format that matches, and check that you do not already have a page serving that same intent. Write the answer first and the context second. After publishing, come back in a couple of months and compare the queries the page actually earns against the one you aimed at.

None of this replaces writing something worth reading. It decides whether what you wrote gets the chance.