Starry night sky over mountains

The 4 Search Intent Types in SEO (And What They Need)

The 4 Search Intent Types in SEO (And What They Need)

The 4 Search Intent Types in SEO (And What They Need)

Search intent types in SEO explained with the page format, proof and action each one needs. See how to match intent before you write a word.

Search intent types in SEO explained with the page format, proof and action each one needs. See how to match intent before you write a word.

10 min read

Steel leaflet rack on a dark wall, one blank sheet forced into a pocket too small so it buckles, with the words "Built For The Wrong Slot" set on it

Imagine a Bristol plumbing firm publishing a long guide called "Power Flush Cost Bristol". It is well written. It has a diagram, a cost table, a section on radiator sludge. And it still gets almost no clicks, because the top of that result page is a local pack and five service pages with a phone number above the fold. The article was never going to win that slot. The four search intent types in SEO get taught as a vocabulary test, and that is the problem: the labels are the easy part, and almost nobody tells you what each type actually needs built onto the page.

So this piece treats intent as a build specification. For each of the four types: the signals that confirm it, the format Google is already rewarding, what goes on the page, what stays off, and how to tell within a few weeks whether you called it right.

Intent is the job the searcher wants done, not a tag in a tool

Every keyword tool will hand you an intent label next to the volume figure. Those labels are pattern guesses. They are generated from the words in the query, not from anything Google publishes, and they fall over on exactly the queries where the money is: short head terms, product category names, anything with a location in it. A tool will cheerfully mark "content marketing platform"as informational because it reads like a definition. The live results say otherwise.

Worth knowing: the four-category model is an industry convention, not an algorithm output. Google's own Search Quality Rater Guidelines, which are published openly and updated periodically, describe user intent in different terms again: Know, Know Simple, Do, Website and Visit-in-person. The four labels stuck because they map neatly onto the page formats that rank, which is the only reason they are useful.

Read the top ten as your primary evidence. Open the query in a fresh window with UK results, then do two things. Count the page types out of ten: blog guide, product page, service page, listicle, forum thread, video, documentation. Note the SERP features: People Also Ask, featured snippet, AI Overview, shopping unit, local pack, sitelinks. Whatever that count tells you overrides the tool label, every time. If seven of the ten are product pages, your 2,000-word guide is not competing with those pages. It is competing with nothing, because it is not in the format being ranked.

Informational intent: what "know"queries need on the page

Signals: how, what, why, "guide", "examples", "checklist". The results fill with People Also Ask boxes, a featured snippet, often an AI Overview sitting above everything.

What the page needs is a direct answer to the exact question in the first 80 to 120 words, before any throat-clearing about why the topic matters. Subheads should match the sub-questions a reader actually has, in the order they would ask them. Then original examples: a real scenario, a real number with its source, a named tool, something that could only come from having done the work.

What to leave off: a sales pitch above the fold, a gated download, and padding written to hit a word count somebody invented. Long does not mean thorough. Long usually means the writer was paid by volume.

AI Overviews have changed what these pages have to earn. The shallow version of most know queries is now answered on the results page itself, so a post that restates common knowledge competently loses the click to a summary that does the same thing without making anyone scroll. We cover the broader shift in what AI Overviews mean for an SME blog plan, but the short version for intent purposes is narrow: informational pages survive on first-hand detail a summary cannot reproduce. The quote you were actually given. The thing that broke in month four. The bit almost nobody writes about.

Internal linking job for an informational page: pass the reader onward to the commercial comparison page, not straight to pricing. Someone who has just learned what a power flush is does not want a booking form. They want to know what the options cost and which suits their boiler.

Navigational intent: the category SMEs leave on the table

Signals: brand names, product names, "login", "pricing", "contact". The result page is usually one dominant domain plus sitelinks, with very little else worth having.

For your own brand terms, the build is unglamorous and mostly technical. Clean, unambiguous titles. One canonical URL per destination. A logical path (/plans, not /landing-page-v3-final). Above all, one obvious destination per query rather than three near-duplicate pages quietly competing with each other, which is how a lot of small sites end up ranking their weakest page for their own name.

Then the part that gets ignored: brand plus modifier. "[Your brand] reviews". "[Your brand] alternatives". "[Your brand] pricing". That is navigational traffic from people already thinking about you, and if you do not publish an honest answer, an affiliate site with a comparison table will. We publish our own plans and pricing as a crawlable page for exactly this reason: the alternative is somebody guessing at our prices in public.

Chasing a competitor's branded terms, on the other hand, rarely pays. You will rank below their own site, the clicks you do get arrive already committed, and you spend budget renting attention on somebody else's name. Build the honest "alternatives"page that includes them instead. Name the cases where they are the better choice. That page converts better than the one that pretends there is no comparison to make.

Commercial investigation: what "best"and "vs"queries need

Signals: "best", "vs", "alternatives", "review", "for small business", "cheapest". The results fill with listicles, comparison tables and review sites, usually with at least one page old enough that nobody has updated the prices.

The page needs a comparison table with criteria a buyer would actually use, named alternatives including ones you do not sell, stated prices or honest ranges, and a plain verdict on who each option suits. Every figure carries its source or an admission that it is a typical range at the time of writing. That last habit is what separates a useful comparison from a brochure.

The commonest mistake we see on small business sites: a "best X"page that lists only the author's own product in five flattering sections. It matches neither the format nor the expectation, so people bounce back to the results within seconds, and that behaviour is not invisible.

Commercial pages are where genuine expertise shows, because you can name trade-offs honestly. An agency, a freelance writer, an in-house hire, doing it yourself on WordPress, and content software are five genuinely different purchases with different failure modes, and a comparison that says so is more persuasive than one that does not. Our own write-up of what hiring a content writer in the UK actually costs takes that approach: ranges, caveats, and the cases where the cheaper option is the right one.

Transactional intent: what "do"and buy queries need

Signals: "buy", "price", "book", "quote", "near me", a town name. The results show shopping units, a local pack, product and service pages. Rarely a blog post, and when there is one it is usually sitting at position nine getting nothing.

The page needs a price or a defensible range, availability, one primary action, trust proof within sight of the button (reviews, accreditations, a named person), and structured data that matches what the page actually is. Not three competing calls to action. One.

Local modifiers change the build completely. A location page wants the address, the service area, opening hours and alignment with the Google Business Profile, not a 1,500-word essay about the history of underfloor heating. Back to the plumbing firm: "what is a power flush"is a guide with a diagram and a cost range; "power flush cost Bristol"is a service page with a price band, the postcodes covered and a booking action. Same business, same service, two completely different builds. Writing one when you needed the other is the most expensive mistake available in a lean content plan, because the slot was never winnable and you spent a month of capacity finding out.

Mixed and shifting intent: where SME content actually fails

Plenty of result pages are fractured. "Content marketing platform"returns product pages, "best of"listicles and definition articles together, because Google is hedging across three different searchers. Pick one format and hope, and you satisfy a third of the demand.

The decision rule we use: if two formats each hold three or more of the top ten, build a hybrid, define the category in two short paragraphs, compare named options in a table, and link clearly to pricing. If one format holds seven or more, build that single format and put the other job on a separate page with deliberate anchor text between them. Two focused pages linked properly beat one page trying to be both.

Intent also moves. A query you ranked for eighteen months ago may now show a different mix of page types after a core update, a new SERP feature, a product launch or a news story that hijacked the term. Before refreshing an old post, re-check the live results rather than the word count. Most refresh projects we see skip that step and spend a day adding paragraphs to a page whose format stopped matching a year ago. If your rankings dropped across a whole section at once, our one-week core update audit plan runs the same check at scale.

Turning intent into a plan you can check

Before anything gets written, five lines per page is enough:

  • Query: the exact phrase, not a theme.

  • Top ten page types observed: counted today, with the SERP features noted.

  • Chosen format: guide, comparison, service page, location page.

  • Primary action: one, named.

  • The single question this page must answer: written as a question.

Grounding briefs in current results is the whole point of our AI business profile and keyword research step: the profile decides which queries belong to your business, and the live results decide what shape each page takes. A tool label alone will wave a mismatch straight through.

Then map the chain: informational articles link to the commercial comparison, the comparison links to the transactional page, and the anchor text says what the destination is. Not "click here". Not "learn more".

Measure it honestly in Search Console, by query group rather than by page. High impressions with a low click-through rate is the signal that matters, and it is almost always a format problem rather than a title problem. Take a page averaging position 7 with roughly 4,000 impressions and 30 clicks over three months, where the top ten is comparison tables and your page is a narrative essay. A new title tag will not fix that. The page type is wrong, so the fix is a rebuild. We go further into reading the numbers without panicking in rank tracking made easy.

Keeping this consistent across thirty pages is a workflow problem more than a skill problem, which is why we built research, planning, writing, publishing and measurement into one connected process rather than five tools and a spreadsheet that nobody updates after week three. Use the check on us if you like: look at the query this page is built for, count the page types in the results, and decide whether the format matches. That is the test, and it applies to everybody.

Frequently Asked Questions

What are the four search intent types in SEO?

Informational (the searcher wants to know something), navigational (they want to reach a specific site or page), commercial investigation (they are comparing options before buying) and transactional (they are ready to act, buy, book or request a quote). The four search intent types in SEO are an industry convention rather than a Google output: they are useful because they map onto the page formats that tend to rank for each kind of query.

How do I work out the search intent of a keyword?

Search it in a fresh window with UK results and count the page types in the top ten, then note the SERP features present. If seven results are service pages and there is a local pack, the intent is transactional and local whatever your keyword tool says. Tool labels are pattern guesses based on the wording of the query, so treat them as orientation, not evidence.

Can one keyword have more than one search intent?

Yes, and the fractured results are the giveaway. When the top ten contains three different page types, Google is hedging across several searcher types. Either build a hybrid page that handles the dominant two jobs in order, or build two focused pages and link them with clear anchor text.

What happens if my page format does not match search intent?

Usually the page ranks on the fringes, picks up impressions and very few clicks, and never improves regardless of how often it is edited. High impressions with a low click-through rate in Search Console is the classic symptom. On-page tweaks do not repair a format mismatch: the page needs rebuilding as the type the results are already rewarding.

Have AI Overviews changed how informational intent works?

For know queries specifically, yes. The generic version of the answer is now frequently resolved on the results page, so pages that only restate widely available information lose the click. What still earns a visit is first-hand detail: original examples, named tools, real figures with sources, and the parts of the job a summary cannot reproduce.

Imagine a Bristol plumbing firm publishing a long guide called "Power Flush Cost Bristol". It is well written. It has a diagram, a cost table, a section on radiator sludge. And it still gets almost no clicks, because the top of that result page is a local pack and five service pages with a phone number above the fold. The article was never going to win that slot. The four search intent types in SEO get taught as a vocabulary test, and that is the problem: the labels are the easy part, and almost nobody tells you what each type actually needs built onto the page.

So this piece treats intent as a build specification. For each of the four types: the signals that confirm it, the format Google is already rewarding, what goes on the page, what stays off, and how to tell within a few weeks whether you called it right.

Intent is the job the searcher wants done, not a tag in a tool

Every keyword tool will hand you an intent label next to the volume figure. Those labels are pattern guesses. They are generated from the words in the query, not from anything Google publishes, and they fall over on exactly the queries where the money is: short head terms, product category names, anything with a location in it. A tool will cheerfully mark "content marketing platform"as informational because it reads like a definition. The live results say otherwise.

Worth knowing: the four-category model is an industry convention, not an algorithm output. Google's own Search Quality Rater Guidelines, which are published openly and updated periodically, describe user intent in different terms again: Know, Know Simple, Do, Website and Visit-in-person. The four labels stuck because they map neatly onto the page formats that rank, which is the only reason they are useful.

Read the top ten as your primary evidence. Open the query in a fresh window with UK results, then do two things. Count the page types out of ten: blog guide, product page, service page, listicle, forum thread, video, documentation. Note the SERP features: People Also Ask, featured snippet, AI Overview, shopping unit, local pack, sitelinks. Whatever that count tells you overrides the tool label, every time. If seven of the ten are product pages, your 2,000-word guide is not competing with those pages. It is competing with nothing, because it is not in the format being ranked.

Informational intent: what "know"queries need on the page

Signals: how, what, why, "guide", "examples", "checklist". The results fill with People Also Ask boxes, a featured snippet, often an AI Overview sitting above everything.

What the page needs is a direct answer to the exact question in the first 80 to 120 words, before any throat-clearing about why the topic matters. Subheads should match the sub-questions a reader actually has, in the order they would ask them. Then original examples: a real scenario, a real number with its source, a named tool, something that could only come from having done the work.

What to leave off: a sales pitch above the fold, a gated download, and padding written to hit a word count somebody invented. Long does not mean thorough. Long usually means the writer was paid by volume.

AI Overviews have changed what these pages have to earn. The shallow version of most know queries is now answered on the results page itself, so a post that restates common knowledge competently loses the click to a summary that does the same thing without making anyone scroll. We cover the broader shift in what AI Overviews mean for an SME blog plan, but the short version for intent purposes is narrow: informational pages survive on first-hand detail a summary cannot reproduce. The quote you were actually given. The thing that broke in month four. The bit almost nobody writes about.

Internal linking job for an informational page: pass the reader onward to the commercial comparison page, not straight to pricing. Someone who has just learned what a power flush is does not want a booking form. They want to know what the options cost and which suits their boiler.

Navigational intent: the category SMEs leave on the table

Signals: brand names, product names, "login", "pricing", "contact". The result page is usually one dominant domain plus sitelinks, with very little else worth having.

For your own brand terms, the build is unglamorous and mostly technical. Clean, unambiguous titles. One canonical URL per destination. A logical path (/plans, not /landing-page-v3-final). Above all, one obvious destination per query rather than three near-duplicate pages quietly competing with each other, which is how a lot of small sites end up ranking their weakest page for their own name.

Then the part that gets ignored: brand plus modifier. "[Your brand] reviews". "[Your brand] alternatives". "[Your brand] pricing". That is navigational traffic from people already thinking about you, and if you do not publish an honest answer, an affiliate site with a comparison table will. We publish our own plans and pricing as a crawlable page for exactly this reason: the alternative is somebody guessing at our prices in public.

Chasing a competitor's branded terms, on the other hand, rarely pays. You will rank below their own site, the clicks you do get arrive already committed, and you spend budget renting attention on somebody else's name. Build the honest "alternatives"page that includes them instead. Name the cases where they are the better choice. That page converts better than the one that pretends there is no comparison to make.

Commercial investigation: what "best"and "vs"queries need

Signals: "best", "vs", "alternatives", "review", "for small business", "cheapest". The results fill with listicles, comparison tables and review sites, usually with at least one page old enough that nobody has updated the prices.

The page needs a comparison table with criteria a buyer would actually use, named alternatives including ones you do not sell, stated prices or honest ranges, and a plain verdict on who each option suits. Every figure carries its source or an admission that it is a typical range at the time of writing. That last habit is what separates a useful comparison from a brochure.

The commonest mistake we see on small business sites: a "best X"page that lists only the author's own product in five flattering sections. It matches neither the format nor the expectation, so people bounce back to the results within seconds, and that behaviour is not invisible.

Commercial pages are where genuine expertise shows, because you can name trade-offs honestly. An agency, a freelance writer, an in-house hire, doing it yourself on WordPress, and content software are five genuinely different purchases with different failure modes, and a comparison that says so is more persuasive than one that does not. Our own write-up of what hiring a content writer in the UK actually costs takes that approach: ranges, caveats, and the cases where the cheaper option is the right one.

Transactional intent: what "do"and buy queries need

Signals: "buy", "price", "book", "quote", "near me", a town name. The results show shopping units, a local pack, product and service pages. Rarely a blog post, and when there is one it is usually sitting at position nine getting nothing.

The page needs a price or a defensible range, availability, one primary action, trust proof within sight of the button (reviews, accreditations, a named person), and structured data that matches what the page actually is. Not three competing calls to action. One.

Local modifiers change the build completely. A location page wants the address, the service area, opening hours and alignment with the Google Business Profile, not a 1,500-word essay about the history of underfloor heating. Back to the plumbing firm: "what is a power flush"is a guide with a diagram and a cost range; "power flush cost Bristol"is a service page with a price band, the postcodes covered and a booking action. Same business, same service, two completely different builds. Writing one when you needed the other is the most expensive mistake available in a lean content plan, because the slot was never winnable and you spent a month of capacity finding out.

Mixed and shifting intent: where SME content actually fails

Plenty of result pages are fractured. "Content marketing platform"returns product pages, "best of"listicles and definition articles together, because Google is hedging across three different searchers. Pick one format and hope, and you satisfy a third of the demand.

The decision rule we use: if two formats each hold three or more of the top ten, build a hybrid, define the category in two short paragraphs, compare named options in a table, and link clearly to pricing. If one format holds seven or more, build that single format and put the other job on a separate page with deliberate anchor text between them. Two focused pages linked properly beat one page trying to be both.

Intent also moves. A query you ranked for eighteen months ago may now show a different mix of page types after a core update, a new SERP feature, a product launch or a news story that hijacked the term. Before refreshing an old post, re-check the live results rather than the word count. Most refresh projects we see skip that step and spend a day adding paragraphs to a page whose format stopped matching a year ago. If your rankings dropped across a whole section at once, our one-week core update audit plan runs the same check at scale.

Turning intent into a plan you can check

Before anything gets written, five lines per page is enough:

  • Query: the exact phrase, not a theme.

  • Top ten page types observed: counted today, with the SERP features noted.

  • Chosen format: guide, comparison, service page, location page.

  • Primary action: one, named.

  • The single question this page must answer: written as a question.

Grounding briefs in current results is the whole point of our AI business profile and keyword research step: the profile decides which queries belong to your business, and the live results decide what shape each page takes. A tool label alone will wave a mismatch straight through.

Then map the chain: informational articles link to the commercial comparison, the comparison links to the transactional page, and the anchor text says what the destination is. Not "click here". Not "learn more".

Measure it honestly in Search Console, by query group rather than by page. High impressions with a low click-through rate is the signal that matters, and it is almost always a format problem rather than a title problem. Take a page averaging position 7 with roughly 4,000 impressions and 30 clicks over three months, where the top ten is comparison tables and your page is a narrative essay. A new title tag will not fix that. The page type is wrong, so the fix is a rebuild. We go further into reading the numbers without panicking in rank tracking made easy.

Keeping this consistent across thirty pages is a workflow problem more than a skill problem, which is why we built research, planning, writing, publishing and measurement into one connected process rather than five tools and a spreadsheet that nobody updates after week three. Use the check on us if you like: look at the query this page is built for, count the page types in the results, and decide whether the format matches. That is the test, and it applies to everybody.

Frequently Asked Questions

What are the four search intent types in SEO?

Informational (the searcher wants to know something), navigational (they want to reach a specific site or page), commercial investigation (they are comparing options before buying) and transactional (they are ready to act, buy, book or request a quote). The four search intent types in SEO are an industry convention rather than a Google output: they are useful because they map onto the page formats that tend to rank for each kind of query.

How do I work out the search intent of a keyword?

Search it in a fresh window with UK results and count the page types in the top ten, then note the SERP features present. If seven results are service pages and there is a local pack, the intent is transactional and local whatever your keyword tool says. Tool labels are pattern guesses based on the wording of the query, so treat them as orientation, not evidence.

Can one keyword have more than one search intent?

Yes, and the fractured results are the giveaway. When the top ten contains three different page types, Google is hedging across several searcher types. Either build a hybrid page that handles the dominant two jobs in order, or build two focused pages and link them with clear anchor text.

What happens if my page format does not match search intent?

Usually the page ranks on the fringes, picks up impressions and very few clicks, and never improves regardless of how often it is edited. High impressions with a low click-through rate in Search Console is the classic symptom. On-page tweaks do not repair a format mismatch: the page needs rebuilding as the type the results are already rewarding.

Have AI Overviews changed how informational intent works?

For know queries specifically, yes. The generic version of the answer is now frequently resolved on the results page, so pages that only restate widely available information lose the click. What still earns a visit is first-hand detail: original examples, named tools, real figures with sources, and the parts of the job a summary cannot reproduce.

Start Your Free Seven-Day Trial

Try Awareness Marketing free for seven days and see the blogs and social posts it creates for your business. No card needed.

Enter your website and see the blogs and social posts Awareness Marketing would create for your business. Free, no account needed.

Enter your website and see the blogs and social posts Awareness Marketing would create for your business. Free, no account needed.

Start for Free

In-House vs Agency Content Marketing: The Real Costs

The 4 Search Intent Types in SEO (And What They Need)

Content Marketing Retainer Cost UK: What You Actually Get

How to Choose AI Blog Writing Software in the UK

Monthly SEO Content Packages: What's Actually Included

Open Graph Image Size: Correct Dimensions for Every Platform

Cost to Hire a Content Writer in the UK: Real Numbers

What Is an Automated SEO Content Platform?

AI Mode in Google Discover: What It Means for Your Small Business Blog Traffic

What Is Genuine Thought Leadership Content?

Google Core Update Recovery: a One-week Audit Plan for Small Business Blogs

Rank Tracking Made Easy: From Numbers to Next Article

AI Overviews Now Hit 40% of Searches: What UK SMEs Should Change in the Blog Plan

Search Console's Multimodal Filter: How to Prove Your Blog Images Earn Traffic

Why Track Keywords If Nothing Changes Afterwards?

AI Mode Data in Search Console: How to Read Your Generative AI Report in 15 Minutes

Tracking Mobile Rankings: Is Your Site Really Slipping?

How Competitor Keyword Tracking Finds Content Gaps

CMA Google Search Choice Screens: What SMEs Should Do

Does AI Content Rank? What the Data Shows

UK Marketing Budget Benchmarks (And What They Buy)

Content Marketing Platform: What Lean Teams Need

Google AI Content Guidelines: A Publisher's Checklist

AI Social Media Post Generator: 7 Real Limits

Copywriting vs Content Writing: Which Do You Need?

Unconfirmed Google Update September 2026: What to Check

Rank Tracking: What the Numbers Won't Tell You

Google Spam Update AI Content: 7 Real Triggers

Site Reputation Abuse: How to Audit Your Own Site

How Much Does a Website Cost UK? (Real Quotes)

SEO Content Writing Service: What to Ask Before You Buy

Google Core Update 2026: What to Do If You Dropped

Is AI Content Penalised by Google? What Gets Flagged