Blog Content

The SEO content brief that doubles as the acceptance test

What an SEO content brief should include and leave out, written as an acceptance test rather than a list of ingredients.

Alexander González October 21, 2026 1,928 words

Most content briefs are a specification for filler. They hand the writer a keyword, a word count, a list of terms to include a certain number of times, and a due date. Every one of those four instructions tells the writer to hit a target instead of answering a question, and the article that comes back does exactly that. A brief is the one document written before the work, so it is the cheapest place to decide whether the work will be any good. The version that works specifies the question and how the answer will be checked, and nothing that can be gamed.

What should an SEO content brief include?

An SEO content brief should name the target query, the intent read off the results page, the questions the piece must answer, the sources to cite, and the one-line answer up top. A thin brief lists a keyword and a word count; a working one lists what makes the draft complete, which is also how it is accepted or rejected.

Quick answer

Brief saysWriter doesResult
"2,000 words"Pads to lengthBuried answer
"Use the keyword 8 times"StuffsReads as written for a machine
"Cover this query and these 5 questions"Answers themComplete when the list is done
"Intent: transactional, top 10 are product pages"Writes the matching page typeEnters the contest it can win
"Answer in 40-60 words up top"Leads with the answerSurvives the AI summary

What kind of document this is

The dominant view treats the brief as a set of ingredients: keyword, secondary terms, length, headings. Under that view a good brief is a complete recipe, and the writer's job is to follow it. It produces uniform articles that answer nothing in particular, which is the most common defect in outsourced content.

The brief is better understood as the acceptance test written in advance. Everything in it should be something you can check when the draft returns, and check without opinion. "2,000 words" is checkable and worthless. "Answers what the service costs, what changes the price, and how to compare two quotes" is checkable and decides whether the piece is finished. When the brief is written as a test, the review stops being a matter of taste and becomes a matter of fact, which is exactly what you need when someone else is writing, and the wider version of that control is in how to tell if your SEO agency is working.

What goes in the brief, in order

1. The target query, and only one

One piece answers one query and its follow-ups. Two target queries in one brief is the instruction that produces internal competition later, because the writer tries to serve both and serves neither. If two queries are genuinely distinct, they are two briefs.

2. The intent, read off the results page

The single most useful line in a brief, and the one almost never present. Search the target query, look at which type of page holds the top 10, and write it down: informational, commercial, transactional, with the evidence. "Top 10 are product pages, so this is transactional, write the page accordingly" saves the writer from producing a long guide for a query Google answers with product pages. How to read that reliably is in search intent.

3. The questions the piece must answer

This is the real specification, and it replaces the word count. List the target query plus the three to five questions the reader asks immediately after. The free sources for that list are the people-also-ask block on the results page and, if the piece already exists, the Search Console queries report, both covered in free keyword research.

The list does two jobs. It tells the writer what complete means, and it gives the reviewer a checklist: every question answered, the piece is done; a question missing, it needs a section, not more words. That is why length is an output of the brief, not an input, and why the whole word-count argument is in its own place, how long a blog post should be.

4. The one-line answer, up top

The brief should require the answer to the query in the first forty to sixty words, written to stand on its own. This is not a stylistic preference. It is what survives when an AI summary sits above the results and answers before anyone reaches the page: with zero-click searches up from 56 % to 69 % between May 2024 and May 2025 according to Similarweb, the self-contained block is the part a generative engine can lift and cite, as set out in how to get cited by ChatGPT.

5. The sources, named before writing

A brief that says "cite reputable sources" gets vague attribution back. A brief that names the sources, or requires that every figure carry who measured it and when, gets checkable claims. This is the difference between a piece that can be verified and one that asks the reader for faith, and it is the single requirement that most raises the credibility of outsourced work.

6. What NOT to put in the brief

Word count, keyword density, a number of headings, a list of terms to repeat. Each of these is an instruction to hit a target rather than answer a question, and each is exactly the thing the writer will optimize for at the expense of the reader. The craft the writer still owns, the sentence-level work a brief cannot specify, is in how to write SEO content.

Who, how and why, borrowed from Google

Google's guidance on creating helpful content publishes a framework worth copying into the brief: who made the content, how it was produced, and why it exists. The brief is where those are decided. Who signs it, so there is a real byline. How it is produced, including whether AI generation is involved and disclosed. And why the piece exists at all: to help the reader, or to attract a visit. A brief that cannot answer the last one honestly is commissioning the second kind of content, and no amount of specification fixes that.

When a brief is overkill

For a single factual update to an existing page. Rewriting one paragraph does not need a test document.

When the writer knows the topic better than the briefer. Then the brief becomes a constraint on someone who would make better calls without it, and the useful version is a short list of the questions to cover, nothing more.

When the piece is a service or sales page rather than an answer to a query. Its job is to convert, not to rank for a question, and the brief for it looks different.

Mistakes that repeat

Data and transparency

The framework of who, how and why, and the self-assessment approach to helpful content, come from Google's public guidance on creating helpful content, checked on 5 September 2026. The rest of this article is operating method, not a published standard: the structure of the brief as an acceptance test, the order of its parts, and the exclusion of word count and keyword density are judgment from content work across a portfolio recording more than 300 million impressions a year in Search Console. The zero-click figure, 56 % to 69 % between May 2024 and May 2025, is Similarweb's, reused from this blog's verified data. No statistic on brief effectiveness or content ROI appears here: figures of that kind circulate widely and none I checked traced to a primary study. The forty-to-sixty-word answer block and the reading of intent from the SERP are this blog's own editorial standard, applied to its own articles. Verified as of September 2026.

What this changes

A brief is usually treated as the moment you tell the writer what to produce. It is more useful as the moment you decide how you will judge what comes back.

That flips what belongs in it. The gameable instructions, word count and density and heading counts, come out, because they only measure compliance. The checkable ones go in: the query, the intent, the questions, the sources, the answer up top. Written that way, the brief stops being a wish handed to a writer and becomes the test the draft has to pass, which is the only version that survives contact with work done by someone else.

Frequently asked questions

What is an SEO content brief?

It is the document given to a writer before they start, setting out what a piece has to accomplish to rank and to be accepted. A working brief names the target query, the intent read from the results page, the specific questions the piece must answer, and the sources to cite. It leaves out word count and keyword density, which instruct the writer to pad and to stuff rather than answer.

Should a content brief include a word count?

No. A word count tells the writer to hit a length rather than cover a question, and it produces filler that buries the answer. Replace it with the list of questions the piece must answer: when they are all covered, the piece is complete, and the length comes out on its own, an argument this blog makes in full elsewhere.

How do I put search intent in a brief?

Search the target query, look at which type of page holds the top 10, and write it down with the evidence: ten product pages means transactional, ten guides means informational, comparisons mean commercial. Telling the writer the page type to produce keeps them from writing a guide for a query Google answers with product pages.

Who should write the content brief?

Whoever can read the results page and knows the business goal, which is usually not the same person who writes the draft. The briefer decides the query, the intent and the questions to answer; the writer owns the craft. When the writer knows the topic better than the briefer, the brief shrinks to a list of questions and gets out of the way.

How is a brief different from an outline?

An outline is a proposed structure for one piece; a brief is the specification and the acceptance test. The outline usually comes from the writer as a first step of the work, and it can be checked against the brief. A brief that is just an outline with a keyword on top has skipped the part that decides quality: what the piece must answer and how that will be verified.

Most sites do not have a ranking problem

They have a what-happens-next problem. You can rank first and still sell nothing. The diagnostic looks at both and tells you which one is costing you money.

See the diagnostic