Search intent (or user intent) is the goal behind a query. It's the most important concept in keyword research: search engines have spent two decades getting better at one thing above all, which is detecting what searchers actually want and ranking the pages that deliver it. A page that mismatches intent will not rank, no matter how well optimized it is otherwise.
The four classic intent types#
| Intent | The user wants to… | Example queries | Winning content |
|---|---|---|---|
| Informational | Learn something | "what is crawl budget", "how to tie a tie" | Guides, tutorials, explainers |
| Navigational | Reach a specific site | "github login", "next.js docs" | The actual destination (you can't outrank it) |
| Commercial | Research before buying | "best seo tools", "ahrefs vs semrush" | Comparisons, reviews, listicles |
| Transactional | Act now | "buy domain name", "screaming frog license" | Product, pricing, signup pages |
Intent maps roughly to funnel position: informational at the top, transactional at the bottom. Conversion rates rise as you descend; search volume usually falls.
How to read intent from the SERP#
Never guess intent, because Google has already told you. Search your target query (logged out, or in a private window, set to the right country) and observe:
- What ranks? If the top 10 are all listicles, Google has decided the query is commercial investigation. Your product page will not crack it.
- What features appear? Shopping ads and product grids point to transactional intent. A featured snippet and People Also Ask point to informational. A local pack means Visit-in-person. A "Discussions and forums" block means people want first-hand opinions, not another brand article.
- Is there an AI Overview? It tells you Google thinks a synthesized answer adds value. Google says AI Overviews "are only shown when our systems determine that it is additive to classic Search" (AI features and your website). Note which pages it cites: they are your real competition for that query.
- What formats dominate? Video carousels mean users want to watch. Tables and side-by-side comparisons mean they want structured, scannable information.
Mixed and fractured intent#
Many queries carry multiple intents. "wordpress" could be navigational (wordpress.com), informational (what is it), or transactional (start a site). Google handles this by fracturing the SERP: it mixes result types and lets click behavior settle the ratio.
For you, fractured SERPs mean:
- Lower expected click-through even if you rank #1 for one interpretation
- An opportunity if you can cover the dominant interpretation better than anyone
Intent also drifts. "Masks" meant something different in 2019, 2020 and 2023. Re-check the SERP for your most important queries every quarter instead of trusting a label you assigned once.
Intent should decide page type#
A practical mapping you'll reuse constantly when building topic clusters:
informational → blog post / docs page / free tool
commercial → comparison page / category hub / review
transactional → product page / pricing page / landing page
navigational → your homepage & brand pages (defend them)
local → location page + Google Business ProfileOne page should target one dominant intent. When a single page tries to be a tutorial and a sales pitch, it usually loses to specialized competitors on both fronts. For the local row, see Local SEO.
Intent in the AI era#
AI Overviews and AI Mode make intent matching more important, not less, but the old rule of thumb ("AI only eats informational queries") is out of date.
- AI Overviews spread down the funnel in 2025. In Semrush's tracking of 10M+ US keywords, the informational share of keywords that triggered an AI Overview fell from 91.3% in January 2025 to 57.1% in October 2025, while commercial rose from 8.15% to 18.57% and transactional from 1.98% to 13.94% (Semrush, updated with November 2025 data). Expect AI summaries on "best X" and even "buy X" queries.
- Clicks fall when an AI summary appears. Pew Research Center's panel (March 2025 browsing data, published 2025-07-22) found users clicked a traditional result on 8% of visits that showed an AI summary, against 15% without one. Informational queries that a summary fully answers are the most exposed: to win value there you need to be a cited source (see GEO and SEO for AI Mode & AI Overviews).
- AI Mode splits one question into many intents. Google's docs say AI Overviews and AI Mode "may use a 'query fan-out' technique", issuing related sub-queries to gather sources. A long conversational question can therefore pull in pages that each answer one sub-intent well. That rewards clear, specific pages, not one page stuffed with every angle.
- Commercial and transactional intent still converts on site. People still want to see prices, stock, proof and checkout on a real site before spending money, even if an AI summary did the shortlisting.
Plan portfolios accordingly: informational content builds authority and citations; commercial and transactional content captures revenue; and every page needs a reason to click that a summary can't provide (a tool, original data, a decision you can make on the page).
Checking the intents you already serve#
Your own data shows what people actually want from you. In Search Console, the Insights report has a Query groups card (launched 2025-10-27) that clusters similar queries, typos and translations into groups with total clicks, plus "Trending up" and "Trending down" views. Click a group to drill into its individual queries in the Performance report. Google says the groups are AI-computed, may change over time, don't affect ranking, and only appear for properties with a large volume of queries (Google).
For smaller sites, do it by hand: export the Queries table from Performance → Search results, label each query with an intent, and look for groups that land on a page built for a different intent. Those mismatches are your fastest wins.
Next: The Research Process, turning intent theory into a prioritized keyword map.