012 min
Why search intent matters
Picture a product manager who launches a new page for a time-tracking feature. The page has a clear headline, screenshots and a sign-up button. The team wants it to rank for the query "time tracking".
The page does not rank. When the team reads the results, they see why. Most of the top results are guides that explain what time tracking is and how to do it. A few are comparison lists of tools. Nothing on the first page is a single company's feature page. The people who type "time tracking" mostly want to learn the topic or compare options. They have not decided to try one product.
The page itself was not badly made. It answered a question that the searchers were not asking. This is the cost of ignoring search intent: the page can be accurate, fast and well designed, and still be the wrong kind of page for the query.
The fix was not a better headline. It was a different decision about what to publish. The team could write a guide that ends with the feature, or aim the feature page at a query where people already want to try a product, such as "time tracking app for freelancers". Both choices come from reading intent before writing.
Search intent also affects what happens after the click. A visitor who wanted a quick answer and lands on a sales page leaves. A visitor who wanted to buy and lands on a long essay also leaves. A marketer working on search engine optimisation uses intent to choose which queries are worth a page at all.
022 min
What search intent is
A query is the text a person types or speaks into a search engine. Search intent is the reason behind that query: what the person wants to have done by the time the search ends. Ahrefs, an SEO software company, puts it this way: "Search intent is the reason behind a search query."
The query and the intent are two different things. A query is only a few words. The intent is the goal those words stand for, and different goals can share the same words. Someone who types "python" might want the programming language, a documentation page, a download, or information about the snake. The text is identical. The goal is not.
Three ideas follow from this.
- Intent is a goal, not a topic. "Running shoes" is a topic. "Compare running shoes before buying" and "find the brand's official site" are goals on that topic.
- Intent is inferred, not observed. The search engine and the marketer see only the query, plus clues such as which results people choose. Nobody can read the goal directly. Broder, the researcher who first classified web queries, wrote that there is no assumption that intent can be inferred with any certainty from the query.
- Intent decides what a good result looks like. For a person who wants a fact, a good result is a short, direct answer. For a person who wants to buy, it is a page where they can buy. Search engines try to rank the result that fits the goal, so a page that fits a different goal loses.
Search intent is also called user intent or query intent. This article uses search intent for the goal behind one query, and it explains how marketers sort those goals into types. Intent is a property of a single search at a single moment. It is not a permanent trait of a person, and it is not a stage in a long buying process.
032 min
Where the term search intent comes from
Before the web, search systems were studied as tools for finding information. Classic information retrieval research assumed that every searcher has an information need: a gap in what they know that the search is meant to fill. Under that assumption, a good search system returns the documents that best match the need.
Andrei Broder, then working at the search engine AltaVista, challenged this in a 2002 paper called "A taxonomy of web search", published in the SIGIR Forum. A taxonomy is a classification system. Broder observed that on the web "the need behind the query" is often not informational, and he sorted web queries into three classes by their intent: informational, navigational and transactional. His paper is the most cited starting point for the idea that a query has a goal that is not always a wish for information.
Broder backed the classification with two data sets from AltaVista. One was a survey of search users in 2001 with 3,190 valid returns. The other was a manual reading of 400 queries from the engine's query log. He was careful about their limits. The survey respondents chose themselves, and he called the log data "very soft" because guessing intent from a query is at best inexact.
Later search teams adopted the same idea in their own words. Google's Search Quality Rater Guidelines are the instructions Google gives to the human raters who judge the quality of its search results. The September 2025 edition has a section called "Understanding User Intent" that sorts queries into Know, Do, Website and Visit-in-person. The SEO industry then added a fourth group, commercial investigation, for people who are comparing options before they buy. That label does not appear in Broder's paper. It is a working convention in tools and guides from companies such as Ahrefs, Moz and Semrush.
So there is no single owner of the term. Broder gave the field its first classification. Search engines and SEO teams have refined it since.
043 min
The types of search intent
Broder's three classes remain the base of most schemes. The classes below add the commercial group that practitioners use.
| Type | The searcher wants | Example query | Typical result |
|---|---|---|---|
| Informational | To learn or answer a question | how does compound interest work | Guides, explainers, definitions |
| Navigational | To reach a specific site | linkedin login | The official site or page |
| Transactional | To do something: buy, download, sign up | download invoice template | A page where the action happens |
| Commercial investigation | To compare options before choosing | best invoicing software | Review lists, comparisons |
Navigational. Broder's wording is precise. In it, the immediate intent is to reach a particular site. A person usually types a brand name because it is faster than typing a web address. Only one result is right, which makes these queries hard to win for anyone except the owner of the site.
A navigational query is one where the immediate intent is to reach a particular site.
Informational. Broder describes the goal as acquiring information assumed to be on one or more web pages, with no further interaction beyond reading. Some informational queries are broad, like "cars". Others are narrow, like "normocytic anemia". In almost 15% of searches in his survey, the person wanted a good collection of links on the subject instead of a single document.
Transactional. The intent is to perform some activity that happens through the web. Broder lists shopping, finding services, downloading files and using databases. Price, speed and quality matter to these searchers, and a general search engine could not see those factors in 2002.
Commercial investigation. The person plans to decide soon but has not chosen yet. The query often includes words such as "best", "vs" or "review". Ahrefs notes that a query like "best air fryer" can show informational, transactional and commercial results at once. That mix is a sign that the intent is mixed.
Google's rater guidelines use different names for similar ideas. Google's Search Quality Rater Guidelines define a Know query as one where the intent is to find information or explore a topic. The guidelines separate out Know Simple queries, which ask for one specific fact such as a person's height. The same guidelines define a Do query as one where the intent is to accomplish a goal or engage in an activity. Their examples include to download, to buy, or to interact with a website or app. A Website query is a search for one specific site, and a Visit-in-person query is a search for a place such as a business. In plain terms, Know maps to informational, Do maps to transactional, Website maps to navigational, and Visit-in-person is Google's addition for local searches.
The guidelines also warn that these groups are not strict. They say the line between Do and Know is sometimes blurry and that many queries do not fit neatly into one category. Treat the types as a way to start a conversation about a query, not as a label that is always correct.
052 min
How common each type of search intent is
Broder measured how often each type appeared, and the numbers are the best-known early evidence that informational searches are not the whole picture. He wrote: "In fact as we show later, informational queries constitute less than 50% of web searches."
In Broder's 2002 paper, informational queries made up less than 50% of web searches. The data came from two sources at AltaVista in 2001.
| Type | Share in the user survey | Share in the query log (400 queries) |
|---|---|---|
| Navigational | 24.5% | 20% |
| Informational | about 39% (estimated) | 48% |
| Transactional | about 36% (estimated) | 30% |
In the survey, 3,190 users answered a pop-up question. In the AltaVista survey, 24.5% of respondents said they were trying to reach a specific website they already had in mind. The survey could not separate informational and transactional queries with one question, so Broder estimated those two shares from a sample of 200 answers. The log analysis is more direct but "very soft", in his words, because it is a judgment of intent made by reading each query.
Treat these numbers with care. They are 25 years old, they come from one search engine that no longer exists as a general web search, and the sample sizes are small. They show that all three types are common. They do not show today's split, which differs by industry and by country. Use them to argue that intent varies, not to predict the share of a specific keyword list.
A team that needs real shares for its own market should classify a sample of its own queries.
062 min
How to read search intent from the results page
You cannot ask every searcher what they wanted. You can read the evidence the search engine already collected: the results page. The engine ranks pages that its own measurements suggest satisfy the query, so the top results show what it believes the goal is.
Ahrefs describes a three-part method that it calls the three Cs: content type, content format and content angle. Follow the steps in order, because each one narrows the next.
Search the exact query and read the top results.
Look at the titles first. A page type that appears again and again is the content type: blog post, video, product page, category page or landing page.
Name the format.
If the type is a blog post, decide whether the results are how-to guides, step-by-step tutorials, list posts, opinion pieces, reviews or comparisons. This step only works after step 1, because a format has no meaning until you know the page type.
Find the angle.
Look for words that repeat in the titles, such as a year, "free" or "for beginners". The angle shows what searchers value most. In Ahrefs' example, a year such as 2023 appeared often in the top titles for "best air fryer", which suggests searchers want recent recommendations.
List the subtopics.
Open several top pages and note the sections they share. Those are the questions searchers expect the page to answer.
Decide whether you can meet the intent.
If the type is a product page and you only have a blog, or the reverse, choose a different query.
Ahrefs states the central rule directly. Ahrefs advises that if most of the top pages are how-to guides, you should create a how-to guide. It adds that this does not mean copying the top pages. The goal is to match the type and format the searchers expect, then add something the existing pages lack.
Google's guidance for site owners supports a related habit. Google's own guidance tells site owners that users who know a lot about a topic might use different keywords than someone who is new to it. The example Google gives is "charcuterie" against "cheese board". The two terms point at the same thing, but the second is more likely to come from someone who is new to the topic. Pages should use both words. The point for intent is that vocabulary tells you about the searcher's experience, which can change the best answer.
071 min
A documented case: Ahrefs and its backlink checker page
Ahrefs, the SEO software company, has published its own account of a change it made to match intent. Its page for the query "backlink checker" was not ranking well. The company says the reason was that the page lacked free tool functionality. A backlink checker is a tool that lists the other websites that link to a given page. Searchers for that phrase expected a tool they could use, and the page only described one.
The team added the tool to the page so that the page matched what searchers expected. Ahrefs reported that it got 516% more traffic in less than six months by changing one of its landing pages.
Read this case with its limits in mind. Ahrefs reports its own result, on one page, with no test that isolated the change from other work. The result shows the mechanism clearly: the query implied an action, and the page was changed from a description into the thing the searcher wanted. It does not show that every intent fix produces a gain of that size.
The case also shows what a fix looks like. The team did not rewrite the headline or add keywords. It changed the content type from a page that explains to a page that does.
082 min
How search intent shows up in tech work
Search intent affects decisions in several roles, not only for people with "marketing" in their title.
- A product manager decides which pages the product site needs. Queries with transactional intent, such as "invoice software free trial", need a page with a clear next step. Queries with informational intent need a guide. Putting a sales page against an informational query wastes the page.
- A content designer or writer chooses the format of each article. If results for a query are comparison lists, a long essay is a different format from what the searchers expect.
- A designer shapes the page after the click. A person with a navigational or transactional goal wants the route to the action to be short, so the main call to action should be visible without scrolling. A person with an informational goal first wants the answer.
- An engineer builds the on-site search and the page structure. Internal search has the same kinds of intent as web search. Someone who types a product name into a help centre wants a specific page, and the site's information architecture decides whether they find it.
- A founder decides where to spend limited effort. Transactional queries bring people closer to buying. Informational queries can bring more readers who may never buy.
Teams also use intent when they test pages. If a team runs an A/B test on a page and ignores intent, a test can compare two versions of a page that suits the wrong visitors. A test can improve a page for the visitors it gets, but intent decides which visitors those are.
Intent also helps a team split its audience into groups of searchers with different goals. That is a form of segmentation based on what people want from a visit, which is often more useful than a split based on who they are.
092 min
When search intent is hard to use
Search intent is a good starting question and a poor final answer. These are the common failures, each with a sign you can check.
- The query has more than one intent. The rater guidelines say that many queries have more than one likely user intent. Their example is the query "harvard", where a person may want the official site, directions, or information about the school. A page can only serve one of these well. Check the results: when they are mixed, the engine is hedging, and your page will be one of several kinds.
- The intent shifts over time. The best result for a query can change when news, products or habits change. A guide that matched a query last year may not match it now. Re-read the results page on a schedule instead of once.
- The label is wrong. The rater guidelines say that the line between Do and Know is sometimes blurry. If two people on the team label one query differently, that is information. It means the query is ambiguous, and the results page should settle the question.
- Copying the crowd. Ahrefs notes that matching the top pages does not mean copying them. A page that matches intent and adds nothing new gives the engine no reason to rank it above pages already there.
- Treating the type as the whole story. Matching intent gets a page considered. It does not make it good. A page can match intent and still be inaccurate, slow or hard to read.
Broder warned that the intent behind a query cannot be inferred with certainty from the query alone. That warning still holds. Treat every label as a working guess, and let behaviour on the page, such as how many visitors leave at once and what they click, show whether the guess was right.
101 min
Search intent compared with nearby terms
Keyword. A keyword is the phrase, such as "project management software". Search intent is the goal behind the phrase. One keyword can carry several intents, and several keywords can carry one. Keyword research finds phrases and counts them. Intent analysis asks what the people who type them want.
Levels of awareness. The levels of awareness idea describes how much a person already knows about their problem and the available solutions. Search intent describes what they want from one search. The two are related but separate. A person who knows little about a problem often starts with informational searches. A person with high awareness may search for a specific product. But the same person can run an informational search and a transactional one on the same day.
Search volume. Volume says how many searches happen. It says nothing about what those searchers want. A keyword with high volume and the wrong intent for your page is not an opportunity.
?5 questions
Questions people ask
What is search intent?
What are the four types of search intent?
How do you find the search intent of a keyword?
Does Google use search intent?
Can one query have more than one intent?
§4 sources
Sources
Show all 4 sourcesHide sources
Broder, A. (2002). A taxonomy of web search. SIGIR Forum, 36(2).
Google. Search Quality Evaluator Guidelines (General Guidelines, September 11, 2025).
Google Search Central. Search engine optimization (SEO) starter guide.
Makosiewicz, M. (updated 2024). Search Intent in SEO: What It Is & How to Optimize for It. Ahrefs.





