Reading search intent: why good content still does not rank
The most common reason a thorough piece does not rank is neither a technical weakness nor a missing link. It is that it answers a different question from the one that was asked — and that can be checked in two minutes before writing.
In short
- The results page shows which intent the search engine assigns to a term — the most reliable information available.
- Four intents: knowledge, method, comparison, purchase. A text aimed at the wrong one does not rank even if it is better.
- Five signals on the results page reveal the intent: type of result, format of the top entries, length, extra elements, and the kinds of provider listed.
- Some terms carry mixed intent. Then whoever serves both wins — answer at the top, depth below.
A text can be professionally superior, well structured and technically flawless and still sit at position 30. In most such cases it is not the text but the fact that it was written for a different intent from the one behind the search term.
The four intents
| Intent | The person wants to … | Expected format | Typical phrasing |
|---|---|---|---|
| Knowledge | understand something | explanation, definition | what is, why, meaning |
| Method | do something | guide, template, steps | how, guide, do it yourself |
| Comparison | decide | side-by-side, table | or, vs, alternative, best |
| Purchase | commission something | offering page, price | price, cost, provider, buy |
Reading the intent — five signals
Open the results page in a private window without signing in and read off five things. Two minutes per term.
1. What kind of pages sit at the top?
Written pieces, vendor pages, comparison sites, forums or videos? The dominant kind is the answer. If not a single article is there, it is not an article term.
2. What format do the top three have?
Prose, numbered steps, a table or a price? The expected format is a strong signal — and it costs nothing to follow.
3. How long are they?
Not as a target but as an indication of expectation. If the top three run to 800 words each, the question is answerable in 800 — and 3,000 then reads as missing the point.
4. Which extra elements does the page show?
An answer box points to knowledge, expandable questions to mixed intent, product ads to purchase, a map extract to local intent.
5. Which kinds of provider appear?
Large portals and trade media alone mean: out of reach for years. If at least one provider your size is there, the term is realistic.
Worth knowing
The results page is the only source revealing how a search engine actually classifies a term — no tool metric comes close. It shows not what the search engine thinks about the term, but what has proved itself in practice.
That is why the two-minute look before writing is worth more than any volume research: it prevents the one kind of mistake that cannot be fixed afterwards by better quality.
When the intent is mixed
For some terms the results page shows several kinds at once — two guides, a comparison and a vendor page, say. That means the search engine is not sure either, because different people want different things from the same term.
That is an opportunity, not a problem. Serving both intents in one piece covers more than the specialised results:
- The knowledge question at the very top, short and complete — two or three sentences.
- The method below it, as numbered steps.
- The decision aid at the end, as a table or list of criteria.
A pattern that shows up regularly: a piece sits at position 12 to 18 for months and does not move, even though it is substantively stronger than the results ahead of it. A look at the results page then reveals that all the leading entries are numbered guides — and your own piece is continuous prose.
Switching to the same form, with no change to the content, moves such pieces noticeably in experience. That is unsatisfying, because it has nothing to do with quality — but it is the cheapest improvement available.
The check belongs before the writing
Two minutes before writing replaces two hours of rework afterwards. Three questions are enough:
- Which intent does the results page show? If it does not match your planned text, change the text — not the term.
- What format do the top three have? Adopt it, even if you would have written it differently.
- Is at least one provider your size there? If not, choose a more specific term.
Help me classify the search intent behind a term. Search term: [term] What I want to achieve with it: [goal] I have looked at the results page. Here is what sits in the top ten (kind of page, title, rough length, kind of provider): [paste list] Extra elements on the results page: [answer box / expandable questions / ads / map / videos] Tasks: 1. Assign the term one or more of the four intents: knowledge, method, comparison, purchase. Argue from the results, not from the wording of the term. 2. Tell me which format I have to write to be in contention at all – and whether my stated goal fits it. 3. If the intent is mixed: propose a structure serving both. 4. Judge from the kinds of provider whether the term is reachable for a small company. If not, suggest three more specific terms. Do not invent search volumes or difficulty scores.
In closing
Search intent is the one factor quality cannot compensate for. An excellent explanatory text does not rank on a comparison term, and an offering page does not rank on a method term.
The check takes two minutes and happens before writing. Skip it and you will occasionally write a very good text past the question — and you will never find the fault in the text, because it is not there.
Common questions
What is search intent?
The purpose behind a query — whether someone wants to understand something, do something, decide, or commission something. It determines which kind of page is in contention at all: a text aimed at the wrong intent does not rank even when it is substantively better than the leading results.
How do you recognise search intent?
From the results page, opened in a private window without signing in. Five signals: which kind of pages sit at the top, what format the top three have, how long they are, which extra elements the page shows, and which kinds of provider appear.
Why does a good piece sometimes not rank?
Often because it has a different format from the one the results page expects — continuous prose, say, where all the leading results are numbered guides. Switching to the same form with no change to the content moves such pieces noticeably in experience.
What do you do with mixed search intent?
Serve both intents in one piece: the knowledge question short and complete at the very top, the method below it as numbered steps, the decision aid at the end as a table. That covers more than the specialised results do.
When should you check the intent?
Before writing. Two minutes looking at the results page saves two hours of rework — and prevents the one mistake that cannot be fixed later through better quality.
Marketing that sets itself up
The Studio Engine beta is live. Claim your spot and help shape it from the start.
Join the beta →