Search intent (also user intent) is the goal a person has when typing a query into a search engine. It determines which kind of result fits, for example an explanation, a product page, a specific website, or a business nearby. Matching search intent is a basic requirement for ranking a page on a keyword.
At a glance
- Type
- Goal behind a search query
- Types per Broder (2002)
- informational, navigational, transactional
- Categories at Google
- Know, Know Simple, Do, Website, Visit-in-Person (rater guidelines, section 12.7)
- Affects
- which kind of result fits a query
Last checked:
What are the types of search intent?
According to reports, the classic model goes back to Andrei Broder, who in 2002 split web searches into three groups. Informational queries look for knowledge, navigational queries aim at a specific site, and transactional queries want to get something done, like buying or downloading. Most SEO tools today add a fourth group:
- Informational: “what is a canonical tag”
- Navigational: “search console login”
- Commercial: research before a purchase, such as “best log file analyzer”
- Transactional: “buy trail running shoes”
The commercial category is an industry convention and does not appear in Broder’s paper.
How does Google describe user intent?
Google’s Search Quality Rater Guidelines (section 12.7) use their own labels. Know queries seek information, and Know Simple queries are a subset that expects a short, specific answer such as a fact. Do queries want to accomplish a goal or engage in an activity. Website queries look for a specific site or page. Visit-in-Person queries look for a business or place, often nearby.
The guidelines state that many queries have more than one likely intent. Section 12.6 adds that the meaning of a query can change over time, for instance after a product launch or a news event. The guidelines are a handbook for human raters. Google says that no single rating directly changes how a page appears in Search. The document shows how Google thinks about queries, but it is not a list of ranking factors.
| Type | Characteristic | Example | Category in Google's rater guidelines | Origin of the label |
|---|---|---|---|---|
| Short tail | short, broad term at the head of the demand curve; intent is often unclear | “coffee machine” | – | counterpart to the long tail; no clear first source |
| Mid tail | between the head and the long end of the curve; more specific than short tail, broader than long tail | “coffee machine with grinder” | – | no clear first source |
| Long tail | specific, rarely searched query, usually several words long | “coffee machine with grinder under 300 euros” | – | term from Chris Anderson, “The Long Tail”, Wired, October 2004 |
| Informational | find information on a topic | “how to descale a coffee machine” | Know, special case Know Simple | Broder 2002, “A taxonomy of web search” |
| Navigational | reach a specific website or page | “reddit login” | Website | Broder 2002, “A taxonomy of web search” |
| Transactional | perform an activity on the web, such as buying or downloading | “buy coffee machine” | Do | Broder 2002, “A taxonomy of web search” |
| Commercial investigation | consider a purchase and compare first | “best coffee machine” | no category of its own | industry usage; neither in Broder nor in the rater guidelines |
| Visit-in-person | go to a nearby place, such as a shop | “gas stations” | Visit-in-Person | Search Quality Rater Guidelines, section 12.7.4 |
No primary source defines the boundaries between short, mid and long tail, whether by word count or by search volume. The table therefore gives no thresholds; the characteristics of the three tail types are based on practical experience. Google said in 2019 that 15 percent of daily queries were ones it had never seen before. Mapping Broder's types to Know, Website and Do is an inference, not a Google statement. According to the rater guidelines, a query can have several intents, and the line between Know and Do is often blurry. Examples are chosen freely; “reddit login” and “gas stations” come from the guidelines. Sources: Google: Search Quality Rater Guidelines, Abschnitt 12.7 Understanding User Intent (Ausgabe 11.09.2025); Broder, A.: A taxonomy of web search. SIGIR Forum 36(2), 2002; Anderson, C.: The Long Tail. Wired, Oktober 2004; Google Blog: Understanding searches better than ever before (2019). Last checked: .
How do you identify search intent?
The current SERP shows how Google reads a query. Three things to look at:
- Page types in the top results: guides, product pages, category pages, tools, videos.
- SERP features: shopping results point to purchase intent, a map pack to local intent, People Also Ask to informational intent.
- Mixed signals: if the top results cover different meanings, the query is ambiguous.
Example
The query “serp” on google.com in October 2026 had two competing intents. Some top results explained search engine results pages, others covered the Supplemental Executive Retirement Plan, a US retirement benefit. People Also Ask contained questions about both. A glossary page on the SEO meaning competes here with finance content and needs to make its meaning clear right away.
Common mistakes
- Guessing intent from the keyword alone. Without checking the SERP it remains an assumption.
- Setting intent once and never checking again. If the interpretation of a query shifts, a page that used to fit may no longer match.
- Serving several intents on one page. A product page that also tries to be a long guide often satisfies neither need well.
FAQ
What are the four main types of search intent? Informational, navigational, commercial, and transactional. The first three go back to Broder’s 2002 taxonomy, the commercial type was added by SEO practitioners.
Can a keyword have more than one search intent? Yes. Google’s rater guidelines describe queries with multiple intents explicitly. The SERP then mixes different page types.
What are the three Cs of search intent? A framework from the SEO industry that looks at content type, content format, and content angle of the top-ranking pages. It is a practical checklist, not a Google concept.
Sources
Last updated: