Most SEO advice comes wrapped in vocabulary that makes it feel like a specialist trade. Some of it is. But one idea does more work than the rest put together, and you do not need a tool subscription to use it: search intent.
Search intent is what the person typing a query is trying to get done. Not the words they typed, the job behind them. Search engines spend enormous effort guessing that job, and they rank pages that do it. A beautifully written post that does a different job will sit on page four no matter how many times the keyword appears in it.
If you only learn one SEO concept, learn this one.
The four jobs a searcher is doing
The classic split has four buckets. They blur at the edges, but they are good enough to make decisions with.
Learn something (informational)
"What is a content calendar." "How does compound interest work." "Why is my sourdough flat."
The reader wants an answer or an understanding. They are not shopping yet. The pages that win are explainers, how-to guides and tutorials, and the best of them answer the literal question in the first few lines before going deeper.
This is where most blog posts live, and where most blogs should spend most of their effort.
Go somewhere (navigational)
"Notion login." "Stripe docs." "IKEA returns."
The reader already knows where they are going and is using the search box as an address bar. You will not win these for someone else's brand, and you should not try. The only navigational queries worth caring about are your own brand name and your own product names.
Compare options (commercial investigation)
"Best project management tool for agencies." "Mailchimp vs ConvertKit." "Is a standing desk worth it."
The reader has a need and is weighing ways to meet it. The pages that win are comparisons, roundups, reviews and honest "who this is for" breakdowns. These posts are harder to write well, because readers can smell a comparison where the author's product always wins.
Do something (transactional)
"Buy running shoes size 10." "Canva pricing." "Book a dentist near me."
The reader is ready to act. The pages that win are product pages, pricing pages and booking flows. A blog post is almost never the right answer here, and trying to force one in usually means writing a thin page that ranks for nothing.
How to read intent without a tool
You do not need to guess. The search results page tells you what the search engine has concluded about intent, and it has far more data than you do.
Search the query in a private window and look at the first page:
- What kind of pages are they? Ten guides means informational. Ten product pages means transactional. A mix of "best X" lists and review sites means commercial investigation.
- What format do they share? If every result is a numbered list, a long essay is swimming upstream. If every result is a step-by-step tutorial, a list of tips will look thin next to them.
- How fresh are they? If the top results are all from this year, the query rewards recency and you are signing up to update the post.
- What else is on the page? A video carousel, a map, a shopping row or a "People also ask" box are all hints about what searchers click.
If the first page is mixed, with guides next to product pages, the intent is split. That is not a reason to avoid the query, but it is a reason to be clear about which half you are serving.
The three mistakes intent explains
Once you look at posts through intent, a lot of mysterious underperformance stops being mysterious.
Right topic, wrong format
You wrote a two thousand word essay on "how to choose a CRM." Every page ranking for it is a comparison table with short verdicts. Your post is good. It just does not look like the answer to the question searchers are asking, so it does not get clicked, and posts that do not get clicked do not climb.
The fix is usually not a rewrite from scratch. Keep the substance and change the shape: lead with the comparison, push the essay underneath it.
A blog post for a buying query
Someone on the team wants to rank for "[your product category] pricing." A blog post goes up. It never ranks, because the searcher wanted a pricing page and the search engine knows it.
If a query is transactional, the answer is a page on your site that lets people act, not a post about the topic. Hand it to whoever owns the product pages.
One post, two jobs
The post tries to explain what a thing is for beginners and also compare the five best vendors for experts. Neither reader gets what they came for in the first screen, and both leave.
Split it. One explainer, one comparison, linked to each other. Two focused posts beat one confused one almost every time.
A simple intent check before you write
Before anything goes into a draft, answer four questions in a sentence each. If you cannot, the post is not ready to write.
- What did the reader type? The actual query, not a paraphrase.
- What do they want to be able to do after reading? "Understand X", "choose between A and B", "fix Y".
- What does the current first page look like? Format, length, freshness.
- Is a blog post the right kind of page for this at all? If the honest answer is no, stop here.
That last question saves more wasted posts than any keyword tool.
Matching the post to the intent
Once you know the job, a few structural choices follow almost automatically.
Informational posts should answer the question early. Put a plain two or three sentence answer near the top, then earn the rest of the reader's time with depth, examples and the follow-up questions they will have next.
Comparison posts should make the comparison scannable. Readers want to see the options side by side before they read your reasoning. Be explicit about who each option suits, including when it is not you.
Posts adjacent to a transactional query should point somewhere useful. If a reader lands on your "how to budget for accounting software" guide, a link to the pricing page is a service, not a sales tactic.
Intent shifts, so check again
Intent is not fixed. A query that used to return guides can start returning product pages when a category matures. A seasonal query can swing between informational in the off season and transactional in the buying season. News can turn a sleepy query into a fresh one overnight.
When an old post that used to perform starts sliding, search the query again before touching the prose. Very often the words on your page are fine and the results page has simply changed what it thinks the job is.
What this means if an agent writes your drafts
Everything above applies whether a person or an AI drafts the post. An agent that writes fluently in the wrong format is still writing in the wrong format.
In RankMill, the topic finder proposes each topic with a title and a short outline, and nothing is written until the topic is approved, by a person or by auto-approve if the team has switched it on. The manual approval step is the natural place for the intent check. Read the outline and ask whether it describes the kind of page the first page of results rewards. If it does not, leave the topic unapproved, or write your own brief through "Write from an idea" with the format spelled out.
It also helps to write your own rule of thumb into the website profile, which the topic finder, the writer, the editor and the auditor all read: which kinds of query deserve a blog post and which belong on a product or pricing page.
The short version
- Search intent is the job behind the query, not the words in it
- Four jobs: learn, go somewhere, compare, act
- The first page of results is the best free intent tool there is
- Match the format of the page to the format that is already winning
- Do not write blog posts for queries where people want to buy
- Recheck the results when an old post starts slipping
Get the intent right and ordinary writing has a chance. Get it wrong and excellent writing does not.