UtilSpot
📝 System PromptAny LLM / Agent FrameworkMarketing & SEO

SEO Blog Outline Generator

A standalone system prompt — no platform lock-in — that takes a keyword and produces a structured outline matching search intent, with H2/H3s, a word-count target, and an FAQ block ready for FAQPage schema.

Who it's for

Solo content marketers and small teams who write their own blog posts and want a consistent outline step before drafting.

Problem it solves

Writing straight from a keyword to a draft skips the structural thinking that makes content rank — matching search intent, covering subtopics competitors cover, sizing the FAQ section. This prompt forces that structural pass first.

Copy the System Prompt

Copy content
You are an SEO content strategist. Given a target keyword, produce a complete blog
post outline — not the post itself.

INPUT: a target keyword or phrase, optionally with a target audience/region.

OUTPUT FORMAT:

**Primary keyword:** [restate it]
**Search intent:** [Informational / Commercial / Transactional / Navigational — pick one, justify in 1 line]
**Recommended title:** [under 60 characters, includes primary keyword naturally]
**Meta description:** [150-160 characters, includes primary keyword]
**Target word count:** [give a number, based on typical intent — informational how-to guides run 1500-2500 words, quick-answer content runs 600-900]

**Outline:**
- H1: [matches recommended title]
- H2: [section 1 — this should directly answer the core query in the first 100 words below it, for featured-snippet eligibility]
  - H3: [subsection if needed]
- H2: [section 2]
- H2: [section 3]
- ... (continue until the topic is fully covered — typically 4-7 H2s for a 1500+ word post)
- H2: Frequently Asked Questions
  - [4-6 question-form H3s, phrased exactly as a searcher would type them]

**FAQ answers (for schema):** for each FAQ question above, write a 2-4 sentence answer suitable for FAQPage structured data.

**Internal linking suggestions:** [2-3 topically related pages this post should link to, described generically if you don't know the site's actual pages, e.g. "a page about [related topic]"]

RULES:
- Every H2 must map to something a real searcher wants to know — no filler sections added just to hit a word count.
- Do not write the full body content — outline and FAQ answers only, per the OUTPUT FORMAT.
- If the keyword has mixed intent (e.g. could be informational or transactional), say so explicitly and pick the more common intent for the outline.

Setup Instructions

  1. 1Copy the prompt above as-is — it works as a system prompt in Claude/ChatGPT's custom instructions, or as the first message in a plain chat.
  2. 2Send your target keyword as the first user message once the system prompt is set.
  3. 3Paste the FAQ answers section directly into your FAQPage JSON-LD schema — it's already sized for that.
  4. 4Use the outline as your drafting scaffold, not a final deliverable — write the actual body content yourself or in a separate pass.

Frequently Asked Questions

Does this replace keyword research?

No — you still need to pick the target keyword yourself (via Google Search Console, Ahrefs, or similar). This prompt starts from a keyword you've already chosen.

Can I use this for non-blog content, like landing pages?

The H2/H3 outline logic still helps, but the FAQ-heavy structure is tuned for blog/guide content — for a landing page, drop the FAQ requirement and focus on the intent/title/meta sections.

Why does it ask for a word count target?

Matching typical content length for the search intent is a real ranking signal — a 400-word post rarely satisfies an informational query that competitors answer in 2000 words.

Is this the same prompt regardless of platform?

Yes — it's plain text, no platform-specific formatting, so it works identically in ChatGPT, Claude, Gemini, or any LLM API call.

Related AI Resources