Most SEO content fails before a word is written. The writer was given a keyword and a word count, searched the keyword, read the top results, and produced a competent average of them. The page is fine. It is also indistinguishable from the pages it was built from, and Google has no reason to prefer it.
The brief is the fix. It is the one document that carries everything learned in research, intent reading and keyword mapping to the person producing the page. This sub-guide, part of the content SEO guide, covers what a brief must contain, what mine look like, and how to use them with writers and with AI tools.
What a good brief contains
A brief is not an outline with a keyword on top. It is a set of decisions, each made by the strategist so the writer does not have to guess it.
Intent and SERP shape
The query cluster the page owns, the dominant intent, and what the SERP currently rewards: page type, typical length, features present. This is the most important line in the brief, and it comes from reading the SERP as described in search intent. "Commercial investigation; top ten are comparisons with pricing tables" tells a writer more than any keyword list.
Audience and situation
Who is searching and what is going on in their work when they do. "A founder of a 20-person D2C brand in Jaipur who has outgrown spreadsheets and is choosing inventory software this month" is a brief. "SMB owners" is not. The situation decides the tone, the assumed knowledge and the examples.
Angle
What this page will say that the current top results do not: a stronger opinion, first-hand data, an Indian-market treatment of a topic covered only for the US, a clearer structure, a decision tool. If the brief cannot name an angle, the page should not be written yet.
Outline
The H2 and H3 structure, with a sentence under each heading saying what it must establish. The outline encodes the secondary terms and sub-questions from the keyword map, so the writer covers the cluster without a keyword list to sprinkle.
Entities and concepts to cover
The things, names, terms and relationships a credible page on this topic mentions: regulations, product categories, standards, named alternatives. This is not keyword density; it is completeness. A page on GST invoicing that never mentions e-invoicing thresholds or HSN codes is not complete, and both the reader and the ranking systems notice. The semantic SEO guide explains why.
Internal links
Which hub this page sits under and must link to early, which siblings it references, and which pages should be updated to link to it once published. Specified in the brief, this gets done. Left to "add links later", it does not.
Sources and evidence
Primary sources the writer should use and cite: official documentation, regulator pages, the product itself, internal data. Also what is unacceptable: other blog posts on the same query, unverified statistics, anything without a traceable origin.
What not to do
The explicit constraints. No claims we cannot back. No repeating the primary term in every heading. No generic introduction when the audience already knows the topic. No sections the SERP does not call for. This section saves more revision cycles than any other.
A sample brief
A condensed version of the template I use, filled for a fictional page. Real briefs carry more detail under the outline.
| Field | Content |
|---|---|
| Page | /blog/gst-invoice-format/ (new); hub: /blog/gst-for-small-business/ |
| Primary term | gst invoice format |
| Secondary terms | gst invoice format in excel, gst bill format, mandatory fields in gst invoice, e-invoice vs gst invoice |
| Intent and SERP | Informational with a strong download sub-intent. Top results: template downloads and field-by-field explainers; PAA on mandatory fields; one government page. Winning pages combine a template with an explanation. |
| Audience | Owner or accountant at a newly GST-registered small business, issuing invoices manually or in Excel, worried about compliance. Assumes basic GST knowledge; does not assume accounting software. |
| Angle | Field-by-field walkthrough tied to the actual rule each field satisfies, plus a downloadable template. Competing pages give the template without the why, or the rules without the template. |
| Outline | H2 What a GST invoice must contain (mandatory fields, grouped). H2 Template (download, annotated). H2 Tax invoice vs bill of supply. H2 E-invoicing: when the format changes (threshold, IRN, QR). H2 Common mistakes (five, each with the consequence). FAQ: four questions from PAA. |
| Entities | GSTIN, HSN and SAC codes, place of supply, reverse charge, IRN, bill of supply, composition scheme, CGST, SGST, IGST |
| Internal links | Hub in paragraph one or two; link to /blog/hsn-code-lookup/ in the fields section; update /blog/gst-registration/ to link here after publication |
| Sources | CBIC invoice rules, GSTN e-invoice portal documentation, the product's own invoicing feature. Do not cite other blogs. |
| Do not | Do not open with "GST was introduced in 2017". Do not give legal advice beyond the rules. Do not state thresholds without a date, as they change. Do not pad with history. |
| Length | Range from SERP: [DATA] words; completeness of the outline is the real target |
| Reviewer | Accounting lead, for rule accuracy before publication |
Briefing writers
A brief is a conversation starter, not a contract thrown over a wall. When I hand one to a writer I walk through the angle and the audience on a call, because those are the parts easiest to misread and hardest to fix after a draft exists. I ask for pushback on the outline before writing, not after; good writers often know something about the audience the research missed.
I also make clear what each side owns. The brief owns intent, structure and constraints. The writer owns voice, examples, clarity and the judgement to deviate when the outline does not fit what they discover. Followed slavishly, a brief produces stiff pages; ignored, it produces the generic average.
Briefing AI assistants responsibly
AI writing tools are now part of most content workflows, including some of mine, and the brief is what separates useful output from plausible filler. The same brief that guides a human writer is the right input for a model, with a few differences in use.
Paste the brief in full as the primary context, not a one-line prompt. Models produce the average of their training data when given nothing specific; the angle, the audience and the "do not" list pull them off that average. Supply sources rather than requesting them: a model asked to find statistics will produce confident numbers that may not exist, so I give it the source material and ask it to work from that only.
What a model should not do is make the claims. Every fact in a draft, especially numbers, legal thresholds, product capabilities and quotations, is verified by a person against the sources before publication. Anything presented as first-hand experience must be first-hand; the brief marks where it is needed and a human adds it. The AI SEO guide covers the wider picture, and the AI search citation test looks at what AI search systems actually cite, which rewards the same specificity.
The review checklist
Every draft is reviewed against the brief before anything else. This is the list.
Does it answer the intent in the first screen?
A reader arriving from the primary query should know within a few sentences that this page is for them. No throat-clearing introductions.
Is the angle visible?
Could a reader say what this page does that the others do not? If the angle is in the brief but not in the draft, the draft is not done.
Is the outline covered, and nothing else?
Every heading and every listed entity is present. Extra sections are challenged: did the writer find something the research missed, or did they pad?
Is every claim traceable?
Numbers, rules, product facts and quotes each have a source from the brief's allowed list. Anything else is verified or removed.
Are the internal links in place and natural?
Hub linked early, siblings linked where referenced, anchors descriptive. Then the pages that should link here are updated.
Does it break any of the constraints?
Run the "do not" list as literal checks. These are the fastest items to verify and the most often violated.
Would the named reader trust it?
Reread as the person in the audience line: tone, assumed knowledge, examples. This catches generic content when everything else passes.
A brief is the difference between content that targets a keyword and content that answers a person.
Frequently asked questions
How long should a content brief be?
Long enough to carry every decision the writer would otherwise guess: usually one to two pages. A keyword and a word count is too short; a brief that drafts the article is too long and removes the judgement you hired the writer for.
Who should write the content brief?
The person who did the research and read the SERP, usually the SEO strategist, with input from whoever knows the audience best. Writers should challenge it, but the intent and angle decisions belong to whoever owns the content strategy.
Should I include a target word count in the brief?
Include a range derived from what the SERP rewards, as a floor for completeness. Do not set a target to pad towards. If the outline is fully covered in fewer words, the page is finished.
Can I use AI to write SEO content from a brief?
You can use it to draft from a full brief with supplied sources, provided a person verifies every claim, adds first-hand experience and edits for the audience. Content generated from a keyword alone is the generic average that ranking systems are designed to filter out.
Do briefs work for product and category pages, not just articles?
Yes, and they matter more there because the template constrains the writer. A product page brief specifies the intent, the comparison points the SERP rewards, the entities such as specifications and certifications, and the internal links, and leaves the copy to the writer within that frame.
Where this connects
This sub-guide sits under the content SEO guide, and the brief template is the central artefact of my SEO content framework. The AI search citation test examines which qualities of a page lead AI search systems to cite it, which is the same specificity a good brief enforces. The SaaS organic growth case study shows a content programme run entirely on briefs of this shape. If you want briefs written for your team, or the whole programme run, see my content SEO service.