Briefs
A brief is the plan for an article, built automatically from a gap’s evidence and yours to change before a single word gets written. This page walks through what’s in it, where each field comes from, and how to move it from draft to ready-to-generate.
What is a brief?
When you click Create brief on a content gap, OptimizeCamp turns that gap’s evidence into a structured writing plan: a title, an outline, the questions the piece needs to answer, the competitor pages it needs to beat, and more. It lands on its own page as a draft with nothing generated yet.
Brief
A structured plan for one article, generated from a content gap and reviewed by you before any drafting happens. Most fields — title, target queries, questions, sources, tone — can be changed, added to, or removed; the outline is built by the AI and shown for review.
How a brief is built
Building a brief starts from the same evidence you saw on the gap card: its losing prompts, the competitor URLs cited instead, and any uncovered subtopics. From that evidence, OptimizeCamp deterministically derives a format, a title, a target-query list, and a section-by-section outline — every outline section carries a short note explaining exactly which prompt, competitor, or subtopic it exists to address.
That deterministic brief is genuinely usable on its own. When an AI model is configured, an additional refinement pass then rewrites the title, outline, questions, and entities for sharper language, working from the same evidence — and if that pass fails for any reason, the brief silently falls back to the deterministic version instead of erroring out. Either way, only open gaps can be briefed, and briefing one flips its status to briefed so it can’t generate a second, duplicate brief.
What a brief contains
Each field below shows up as its own section on the brief page.
| Field | What it is |
|---|---|
| Working title | A format-appropriate headline — e.g. a “vs” headline for a comparison brief, or “The complete guide to …” for a guide. |
| Intent | Informational, comparison, or commercial. |
| Format | Guide, comparison, FAQ, or listicle. |
| Target prompts | The gap’s losing prompts, plus up to a few related active tracked prompts on the same topic — the queries this article exists to win. |
| Questions | The 5–7 questions the article must answer, phrased as questions from the losing prompts plus a couple of standard ones for the topic. |
| Entities | The 5–7 terms, competitor names, and category concepts the article should cover so it reads as complete to an AI answer engine. |
| Outline | Section-by-section H2/H3 headings, each with a note on what evidence or angle that section must use. The outline is built by the AI from the gap evidence — it’s shown for review, not hand-edited. |
| Citations to beat | The specific competitor URLs earning citations on this topic, with how often each is cited and an editable note on how to out-answer it. |
| Sources | Up to 3 neutral (non-brand, non-competitor) pages the writer reads and fact-checks the draft against. |
| Internal links | Your own existing pages that already relate to this topic, suggested as a “Related reading” list at the end of the draft. |
| Voice notes | Free-text tone and style guidance for this article, on top of your project’s brand voice settings. |
Note
The format decides the default title pattern and which outline sections get added — a comparison brief gets a head-to-head section with pricing, features, adoption, and support subsections against the strongest competitor; an FAQ-leaning brief gets a dedicated FAQ block; every brief closes with a “The bottom line” section that names the citations it has to beat. Article length isn’t something you set: every article targets about 1,800 words.
Editing the brief
Most fields are editable while the brief is a draft, with a few sensible guardrails on what’s free-form versus derived from evidence:
- Title, intent, format, voice notes — edit freely.
- Target prompts — remove any, or add your own; prompts pulled from a tracked prompt carry a “tracked” chip, prompts you add by hand show “manual.”
- Questions and entities — simple add/remove chip lists, capped at 7 each.
- Outline — read-only. It’s built by the AI from the gap evidence, and the writer follows it section by section.
- Citations to beat — the URL, domain, and citation count are evidence and stay fixed; only the note is editable.
- Sources — add any http(s) URL (up to 3 total), or remove ones you don’t want the writer to read.
- Internal links — remove suggestions you don’t want; this list is auto-suggested only, not hand-added.
Every saved change is recorded as a version, and the brief page shows how many times it’s been edited — useful if you want to see, at a glance, how much you reshaped the automatic first draft.
From draft to approved to generated
A brief moves through four statuses, and the available actions change with each one:
Draft
The brief is freshly created and fully editable. Save your changes as you go, or use Approve & generate to save, approve, and kick off drafting in one step.
Approved
You’ve signed off on the plan. Approving auto-saves any unsaved edits first. From here, Generate article starts the writing pipeline — see Writing articles.
Generating
The article is being written in the background. The brief can’t be edited while this is running — the page switches to a live progress view instead.
Generated
A draft article exists in the content editor, pre-audited and ready for your review. If generation fails instead, the brief reverts to approved so you can retry.
Tip
Nothing about approving a brief is irreversible on its own — it just means you’re ready to spend the writing step. Generating the article is the point where an article credit gets used.
Key takeaways
- A brief is built automatically from a gap's evidence — a deterministic base, optionally sharpened by an AI refinement pass that falls back safely on any error.
- It contains a title, intent, format, target queries, 5–7 questions and entities, an evidence-linked outline, citations to beat, up to 3 sources, internal links, and voice notes.
- Most fields are editable before generation; the outline is AI-built and read-only, and every article targets about 1,800 words — length isn't configurable.
- A brief moves draft → approved → generating → generated, and only an approved brief can start generating an article.
Next step
See exactly what happens once you hit generate — Writing articles walks through the six-step pipeline and what lands in your editor.