Search intent is what a person wants to achieve with a query: learn something, go somewhere, compare options, or act. Matching it is still the first test a page must pass, but AI search changes how intent shows up. Questions are longer, carry conditions, continue across several turns, and often combine several intents in one request. Pages now need to satisfy the main intent and the sub-intents an assistant breaks the question into. Reading the results page and the AI answer for a query is the most reliable way to see what intent a search engine has inferred.
The classic categories, and Google’s own
| Common SEO label | Google’s quality rater terms | Example |
|---|---|---|
| Informational | Know | what is GEO |
| Navigational | Website | brilad contact |
| Commercial investigation | Know and Do combined | best SEO agency for SaaS in India |
| Transactional | Do | book an SEO audit |
| Local | Visit in person | SEO agency near Wagholi |
Google’s Search Quality Rater Guidelines describe queries in terms of Know, Do, Website and Visit-in-person intents, and note that many queries have more than one reasonable interpretation.
What changes with AI search
1. Intent comes with conditions
“Best CRM” becomes “best CRM for a 20-person sales team in India that works with Tally and costs under a set budget”. Each condition is a sub-intent. Pages that state who they suit, prices and integrations match more of them.
2. Intent unfolds across turns
People ask a broad question, then narrow it: “what is GEO”, then “how long does it take”, then “who does it in Pune”. Each follow-up is a new search behind the scenes. Sites that answer the whole sequence, with pages linked in that order, stay in the conversation.
3. Several intents in one request
“Explain schema markup and tell me whether my WordPress site needs a plugin” mixes learning and doing. Assistants split it and may cite different pages for each part.
4. Tasks, not only queries
People ask assistants to do things: draft a plan, compare vendors, check a website. Pages that provide checklists, templates, tools and step-by-step methods serve these task intents.
How to identify intent for a query
- Search it and look at what ranks: guides, product pages, comparisons, local results, videos.
- Note the SERP features: featured snippet, People also ask, local pack, shopping results, AI Overview.
- Ask two or three AI assistants the full question and note what the answer covers and which kinds of pages it cites.
- Check your own Search Console data for the query and the pages it already sends to.
Matching format to intent
| Intent | Best page type | Format |
|---|---|---|
| Learn a concept | Guide or explainer | Definition first, then detail and examples |
| Solve a problem | How-to or checklist | Numbered steps, with the outcome stated first |
| Compare options | Comparison page | Verdict first, then a table on shared criteria |
| Choose a provider | Service page | Who it is for, what is included, price, process, evidence |
| Act now | Contact, booking or product page | Short, with the action visible immediately |
| Visit | Location page and Business Profile | Address, hours, directions, services at that location |
Common mistakes
- Targeting a comparison query with a service page that only praises your offering
- Answering a “how do I” query with a long introduction before the steps
- One page trying to serve learners and buyers at once, and serving neither well
- Ignoring the follow-up questions that come after the first answer
Map the follow-ups with query fan-out keyword research, and give each intent one clear page, as described in building a topic map.
Frequently asked questions
Does intent change over time?
Yes. What ranks for a query can shift as Google’s understanding changes or as new content appears. Re-check important queries every few months.
Can one page rank for informational and transactional queries?
Sometimes, for closely related queries. Usually separate pages, linked together, serve each intent better.
Is keyword research still useful?
Yes, for demand and wording. Intent analysis tells you what kind of page to build for each keyword.
Sources
- Google, Search Quality Rater Guidelines (understanding user intent)
- Google, description of query fan-out in AI Mode, 2025