TemplatesContentPostWyse

AI SEO Brief Template

If your AI-drafted SEO content reads generic, the problem is the brief, not the model. This is the eight-section brief PostWyse uses internally.

10 to 25 minutes per briefDifficulty: easyFree, no signup

Why this matters

The template, step by step

  1. 01

    Lock the search intent in one sentence

    Before keywords: who is searching this, and what do they want to know after reading? Example: 'Marketing operators at 20-50 person SaaS companies evaluating AI content tools want a non-marketing-speak comparison.' This sentence drives every other decision.

  2. 02

    Pick the primary keyword and three secondary clusters

    Primary keyword goes in the H1, URL, first 100 words, and meta. Secondary clusters (3-4 related queries) get H2s. Don't stuff. If you can't naturally fit a secondary into an H2, drop it.

  3. 03

    List the top 5 ranking pages and what they get wrong

    Paste the URLs. For each, note the format (listicle, deep-dive, comparison), the word count, and the one weakness. Your post needs to do whatever they do plus the thing they all miss. AI is good at copying. The brief is where you tell it not to.

  4. 04

    Name the proof you want to include

    Numbers, case studies, customer quotes, screenshots, citations. List 3-5 specific proof points the AI should weave in. Without this, AI defaults to generic claims like 'studies show.' With it, the post reads like you ran the studies.

  5. 05

    Set the voice constraint and the must-avoid list

    Voice: brand voice descriptor (e.g. 'direct, no hype, occasional dry humor'). Must-avoid: words that would tell a reader this was AI. Common offenders: 'leverage', 'unlock', 'in today's fast-paced world', em-dashes, every paragraph starting with 'Furthermore'. List 8-12.

  6. 06

    Define the structural targets

    Word count band (e.g. 1,400 to 1,800). H2 count (5-7). Bullet list count (no more than 3). Internal links (5-8 to existing posts). External links (2-3 to authoritative sources). FAQ section (4-6 Q&As at the end). These are guardrails, not handcuffs.

  7. 07

    Specify the CTA and the conversion path

    One CTA per post. Where does it go? What does the reader get? If the post is for top-of-funnel awareness, the CTA is a free resource. If it's bottom-of-funnel comparison, the CTA is a demo or signup. Don't mix.

  8. 08

    Hand it to the AI and edit, don't rewrite

    If the brief is good, the AI gets you to 85 percent. Your job is the final 15 percent: tighten openers, swap one or two generic sentences for a specific one, add the one zinger paragraph only you would write. Editing takes 30-45 minutes. Rewriting takes 2+ hours and means the brief was bad.

Eight-section AI SEO brief
1. SEARCH INTENT (one sentence)
   Who is searching, and what do they want after reading?

2. KEYWORDS
   Primary: [exact keyword]
   Secondary clusters (3-4):
     - [cluster 1]
     - [cluster 2]
     - [cluster 3]

3. TOP 5 RANKING PAGES + THEIR WEAKNESSES
   [URL 1] — format, words, weakness
   [URL 2] — format, words, weakness
   [URL 3] — format, words, weakness
   [URL 4] — format, words, weakness
   [URL 5] — format, words, weakness

4. PROOF TO INCLUDE
   - Numbers / stats: ___
   - Case studies / quotes: ___
   - Screenshots / citations: ___

5. VOICE + MUST-AVOID
   Voice: [3-5 word descriptor]
   Must-avoid words: [8-12 words]

6. STRUCTURAL TARGETS
   Word count: [low-high band]
   H2 count: [number]
   Bullet lists: [max]
   Internal links: [count]
   External links: [count]
   FAQ section: [Y/N, count]

7. CTA + CONVERSION PATH
   Single CTA: [name]
   Destination: [URL]
   What reader gets: [one line]

8. EDIT DIRECTIVE
   "Get me to 85%. I'll handle the 15."

Common mistakes

Skip the manual work

Let PostWyse run this template on autopilot.

Wyse drafts every input, every personalization, every follow-up in your brand voice. You approve before anything goes live.

Questions

How is this different from a normal SEO brief?

A normal SEO brief tells a writer what to write about. An AI SEO brief tells an AI what to be specific about. Sections 4 (proof) and 5 (must-avoid) are AI-specific because AI defaults to generic without explicit constraints.

Which AI model handles this brief best?

We've tested with Claude Opus, GPT-4o, and our own Wyse agent. The brief works across all three. Claude tends to follow voice constraints best, GPT-4o tends to hit structural targets best, Wyse is tuned specifically for this brief format and produces output that needs the least editing.

Can I skip sections to save time?

Sections 1, 4, and 5 are non-skippable. Without intent (1), proof (4), and voice constraints (5) the AI output will read generic regardless of how good the model is. Sections 6 and 7 can be templated and reused across posts.

How long does the brief take to write?

8 to 12 minutes once you've done it five times. The first one or two take 25-30 minutes because you're figuring out your voice descriptor and must-avoid list. Those carry over to every brief after.

Does this work for non-English content?

Yes, with one tweak: in section 5, you also specify language-specific tells (e.g. for German, 'avoid Anglizismen'; for Spanish, 'avoid neutral textbook Spanish, use the regional voice'). Otherwise the structure transfers.

Related templates