An article brief is the shared starting point for one article. It records the question the article answers, why we are the ones to answer it, who it is for, what new information it carries, and how it will be structured, so a writer can draft without guessing and an editor can judge the result against something specific.
This page explains the reasoning behind each part of the brief. It is not the fill-in template. Read it to understand what each section is testing, then fill in the article brief.
An article brief is not a copy brief
The two overlap enough to be worth separating deliberately.
A copy brief governs a piece of persuasion that runs for a while and then stops: an ad, an email, a caption. It is written to a character limit, it is judged on a response rate, and when the campaign ends it is done. An article is a different kind of object. It is a permanent asset that has to stay true for years, gets found by people who were not targeted by anyone, and is judged partly on whether a machine will quote it.
That difference changes what the brief has to capture. A copy brief needs hooks, character counts, and objections. An article brief needs a defensible claim to authority, source material, a heading structure, an update cadence, and the metadata that ships with the page. Neither set fits comfortably inside the other, which is why they stay separate documents that link to each other.
A creative brief sits alongside both. An article that needs a diagram, a chart, or original photography generates a creative brief for that asset, linked from the article brief and not folded into it.
The authority question
The brief opens by asking what question or problem this article answers better than anyone else. That is a deliberately hard first question, and a brief that cannot answer it is doing its job by failing.
Most content briefs open with a keyword instead. A keyword tells you that people are asking something. It does not tell you that you have any business answering it. When the opening question is “what do people search for,” the library fills up with articles that restate what is already available in a hundred places, which is exactly the material an answer engine has least reason to cite.
Answering better usually rests on one of three things: knowing something others do not, having done the thing more times than the people writing about it, or holding a clear position on a question others hedge. Naming which of the three applies is what turns a topic into an article worth writing.
Next to the authority question sits a narrower one: what does a reader get here that they could not get from a model directly.
This is the section most likely to expose an article that should not be written. General knowledge, restated well, is what a language model produces for free. An article competing with that has no angle, and it gives an answer engine nothing to quote, since there is no fact in it that came from anywhere in particular.
What counts is specific: a number you can source, data you collected, a method you named, a decision you made and can explain, a quote from someone with standing. It does not have to be large. A single benchmark from your own operations, stated plainly and attributed, does more for an article than several paragraphs of accurate generality.
Who it is for, and what they are actually asking
An article brief links a customer profile rather than describing a reader fresh. Linking keeps one definition of that reader across every campaign and page, so an article does not quietly invent a slightly different version of the same person.
Separately from who they are, the brief asks what they are asking. Search intent and prompt intent are close cousins: someone typing three words into a search box and someone typing a full sentence into a chat window usually want the same thing, but the second version reveals far more about what they actually need. Writing down the real question, in the reader’s words rather than in keyword form, gives the writer something to answer and gives the article a natural opening line.
The brief asks for both the search terms and the questions, because they serve different purposes. The terms shape the title and the metadata. The questions shape the headings.
Entities
An entity is a thing an engine recognizes as a distinct object with facts attached: a company, a person, a product, a place, a concept. Naming the primary entity an article is about, and the supporting entities it touches, does two things.
It keeps naming consistent, which is the foundation of entity authority. If the same product is called three things across five articles, an engine has three weakly-supported entities instead of one well-supported one.
It also surfaces what the article has to define. An article that mentions an entity without ever explaining it assumes context the reader may not have, and gives a retrieval system nothing to match against.
Structure, before the writing starts
The brief asks for the outline: the working H1 and the H2 sections underneath it.
Agreeing the structure at brief stage is cheaper than agreeing it at draft stage, where a structural disagreement means rewriting rather than rearranging. It also forces the shape of the argument into the open early, where a missing section or a section that does not belong is obvious.
There is a retrieval reason too. Sections are the unit an answer engine pulls out and quotes, so the outline is not a formatting decision but a decision about which self-contained answers this article will contain. A brief with five headings that each pose and answer a real question produces a very different article from one with five headings that only make sense read in order.
Proof
The brief asks for the statistics, quotes, and sources the article will rest on, and it asks for them before writing starts.
Proof gathered after a draft exists tends to be proof found to support sentences already written, which is how an article ends up with a citation that does not quite say what the sentence claims. Proof gathered first shapes what the article can honestly say.
This is also the section with the clearest evidence behind it. In controlled testing, adding credible quotations, relevant statistics, and citations for claims produced the largest gains in whether content was used in generated answers. Listing them in the brief is what makes them likely to actually appear in the piece.
Internal links
An article brief names the pages this article should link to and the pages that should link back to it.
A cluster of connected articles is what topical authority looks like in practice, and links are the connection. Deciding them in the brief means the cluster gets built deliberately rather than assembled later from whatever happened to get published, and it catches the case where a planned article duplicates one that already exists.
The links back matter as much as the links out. A new article with nothing pointing at it is isolated, however good it is.
Maintenance, decided up front
The brief sets a review cadence and a next update date before the article is published.
Content decays quietly. Rankings and citations slide as facts age, intent shifts, and competitors publish something more current, and none of that announces itself. The evidence on AI citations is blunt about freshness: the large majority go to pages updated within the past year. An article with no scheduled review is an article that will be updated when someone notices, which in practice means when someone notices it has stopped working.
Deciding the cadence at brief stage also makes the cost of the article honest. A hero article is not a one-off piece of work, it is a commitment to keep something true indefinitely, and that is worth agreeing to before the article exists rather than discovering later.
Sections that earn their place conditionally
Some sections matter for some articles and not others. Forcing them into every brief adds length without adding direction. The test for each is the same: would answering it change what gets written?
Last modified on August 6, 2026
