When drafting cost collapsed, the brief became the product. A strong brief run through mediocre production still yields a useful page; a weak brief run through the best pipeline yields the eleventh restatement of whatever already ranks. This article is the craft of the strong brief - the demand pricing that justifies it, the SERP anatomy that locates the gap, the falsifiable angle that encodes it, and the specification format that a draft cannot quietly dodge. It is the planning chapter of the content operations pillar, in working detail.
The brief is the product
The economics are worth stating plainly. Production - the expensive stage of the old pipeline - now costs minutes per draft. Planning still costs judgment, and its output quality varies enormously: a brief that encodes a real gap produces a page with a reason to exist; a topic label produces convergence. So the leverage moved wholesale into the brief, and teams that still treat briefing as a calendar chore while polishing their production pipeline have optimized the cheap stage. The pipeline's gates exist to make this impossible to ignore: production refuses briefs that fail the standard below.
Stage one: pricing the opportunity
Every candidate topic gets priced before anyone debates it: search volume across the query cluster (never the single keyword - clusters are what pages actually rank for), difficulty against the domains currently holding the SERP, trend direction (a shrinking topic must be cheap to justify), CPC as a proxy for commercial value, and business relevance - the honest question of whether ranking here produces customers or just traffic. The output is a portfolio table where every row carries its economics, which converts the content meeting into a portfolio review: fund the underpriced opportunities, skip the vanity head terms, schedule the strategic-but-hard ones behind the authority they need. The pricing machinery is the same one marketing budget allocation uses, pointed at editorial.
Stage two: SERP and coverage anatomy
For funded candidates, dissect what currently wins. The SERP read: what ranks (formats, angles, depth), what the People-Also-Ask boxes reveal about adjacent questions, what the AI answers currently say and cite for the cluster's questions - the answer surface is part of the battleground now, per the AEO vs SEO comparison. The coverage read: every ranking page's actual contents, mapped - which produces two lists that become the brief's spine. Table stakes: what every winner covers, which the piece must match or consciously skip. The gap: what none of them covers - the question unanswered, the audience unserved, the evidence nobody brought, the format nobody tried.
Stage three: the falsifiable angle
Angles come from crossing the gap list with what you distinctively have: data nobody else holds (your platform's observations), experience nobody else earned (what your missions actually find), or synthesis nobody else did (two literatures the SERP keeps separate). The gap says where to aim; your assets say what to fire.
Stage four: the specification itself
| Field | Contents |
|---|---|
| Target cluster | Primary query + variants, with the pricing attached |
| Audience and job | Who reads this and what they do next |
| The angle | The falsifiable claim, one sentence |
| Table stakes | What must be covered (from the coverage read) |
| The gap sections | Where the piece goes beyond every ranking page |
| Evidence requirements | What must be sourced, computed or cited |
| Structure | Answer-first sections, table/FAQ requirements per the citable formats |
| Internal links | Named cluster edges in and out |
The structure field deliberately imports the LLM SEO formats - the brief is where answer-readiness gets specified, not retrofitted - and the internal links field is how cluster architecture survives contact with production: every piece knows its edges before it is written.
Brief QA: refusing bad briefs
The gate that keeps the system honest: a brief dies if it fails any of three checks. No real demand - the pricing shows volume too thin for the effort, and no strategic override is claimed in writing. No locatable gap - the coverage read found the topic genuinely saturated, and the piece would be the eleventh restatement. No falsifiable angle - the brief's claim cannot be checked in the finished draft. Killing briefs feels like waste and is the opposite: every killed brief is a thin page that never diluted the library, never consumed a production slot, and never taught the QA gate to rubber-stamp mediocrity. Mature programs kill a third of candidates at this gate and publish stronger for it.
Frequently asked questions
What should a content brief contain?
The priced target cluster, audience and their next action, a falsifiable angle, table stakes from the coverage read, the gap sections, evidence requirements, answer-ready structure, and named internal link edges.
What makes an angle falsifiable?
It is a checkable claim - a reviewer reading the draft can verify whether it was delivered. "The standard advice breaks under condition X" is falsifiable; "a fresh perspective" is a vibe that guarantees convergence.
How is the gap actually found?
By anatomy: map what every ranking page covers (table stakes) and what none covers (the gap), including what AI answers currently say and miss for the cluster’s questions. Intuition guesses; the coverage read knows.
When should a brief be killed?
On any of three failures: no real priced demand, no locatable gap in a saturated SERP, or no falsifiable angle. Mature programs kill roughly a third of candidates at this gate and are stronger for it.
Can agents write the briefs themselves?
Agents assemble the inputs - pricing, SERP anatomy, coverage maps - exactly and cheaply. The angle, the audience call and the kill decision are editorial judgment, which is precisely where the human hour belongs.
Sources
- Google - Keyword Planner documentation (the pricing input)
- Google Search Central - helpful content guidance
Every playbook on this blog ships as a runnable mission.
Open a workspace and the playbook library is waiting - describe the outcome and the agents carry it end to end, on your plan's monthly credits.