Marketing Content

Content Writing Workflow

One skill, two modes: build a multi-channel content strategy, or write one evergreen post really well.

Pick your mode

• ❌ Writing API / project documentation → use /docs-sync
• ❌ Designing a logo or video → use /marketing-brand
• ❌ Full go-to-market plan (brand + content) → use /marketing command

Mode STRATEGY

Create content strategy: audience research, content pillars, SEO blog posts, social media plans, newsletters.

When to use

• ✅ Building a content strategy from scratch (pillars, calendar, channels)
• ✅ Writing one specific piece (blog post, newsletter, social thread)
• ✅ Repurposing one piece into multi-channel content
• ❌ Writing one evergreen SEO/AI post deeply → Mode POST

Workflow

RESEARCH → PLAN → WRITE → OPTIMIZE → DISTRIBUTE

Phase 1: RESEARCH

Gather brief (skip what's already known)

Audience & SEO research

• Search intent for the topic (informational / navigational / transactional / commercial)
• Top 10 SERP for 3–5 target keywords — what content ranks?
• Gaps competitors miss (more depth, better examples, fresher data)
• Keyword set — primary + 2–5 long-tails per piece

Gate

• Brief captured · Target keywords with search intent · Competitor gap identified

Phase 2: PLAN

Content pillars

Pick 3–5 themes that ladder up to the brand's positioning. Each pillar gets a mix of formats.

Calendar template (4–6 weeks ahead, consistent cadence)

Content mix rule (per channel)

• 80/20: 80% value (educate / entertain), 20% promote
• Don't post when you have nothing to say — skip a slot rather than post filler

Gate

• 3–5 pillars with rationale · Calendar ≥4 weeks · Channels matched to audience habits

Phase 3: WRITE

One format at a time.

Blog post structure (long-form, SEO)

TITLE (≤60 chars, primary keyword)  ·  META DESCRIPTION (≤155 chars, keyword + CTA)
H1 (matches title)
HOOK (2–3 sentences: the reader's problem)
TL;DR (3 bullets: the answer up front)
H2: Problem context — why this matters now
H2: The solution — main argument  (H3: Step 1 / Step 2 / Step 3)
H2: Common pitfalls — what NOT to do
H2: Real example — actual code / screenshot / case
H2: Next steps — what to do today
CTA (1 line: try / subscribe / share)

Target length: 1200–2500 words for SEO; 500–1000 for opinion.

Social thread (X / LinkedIn)

HOOK (1–2 lines — must work alone in the feed)
↓ 3–7 posts each ≤280 chars (each stands alone, one idea per post, line breaks)
↓ CTA (link to blog / signup)

Newsletter

SUBJECT (≤50 chars, curiosity > clickbait)  ·  PREVIEW TEXT (≤90 chars)
GREETING (personal, 1–2 lines)
MAIN STORY (1 idea, 200–400 words)
LINKS (3–5 curated, 1-line why-it-matters)
CTA (one ask only)  ·  SIGNATURE

Writing rules

• One idea per paragraph (≤4 lines) · Active voice · Concrete > abstract (numbers, code, names)
• Cut filler: "in order to"→"to", "due to the fact that"→"because", "very important"→"critical"
• Read aloud before publishing — if you stumble, rewrite

Gate

• Draft matches the format · Primary keyword in title + H1 + first 100 words (blog) · Read aloud, no stumbles · ≥1 concrete example

Phase 4: OPTIMIZE

Combined checklist (SEO + quality)

• Title ≤60 chars + primary keyword · Meta description ≤155 chars · H1 matches search intent
• H2/H3 use related keywords naturally · Alt text on all images
• Internal links ≥2 related posts · External links ≥1 authoritative source per claim
• No keyword stuffing · Mobile-friendly (short paragraphs, clear hierarchy)
• Reading level appropriate (Hemingway grade ≤10 for general) · No typos

Phase 5: DISTRIBUTE

Per-channel adaptation

Tracking

• UTM tags per channel: ?utm_source=X&utm_medium=social&utm_campaign={pillar}
• Track per piece: views, time-on-page, scroll depth, conversions
• Iterate: double down on what works, kill what doesn't after 3 attempts

Gate

• Adapted for each channel · UTM tags applied · Tracking in place

Mode POST

Write one evergreen blog post that ranks on traditional search AND is easy for AI systems (ChatGPT, Bing/Copilot, Perplexity) to cite or recommend. Intent-first, answer-first, overlap-audited.

When to use

• ✅ Writing a single evergreen blog post (entity / trust / compare / glossary / use-case / FAQ)
• ✅ Updating an existing evergreen post that has drifted
• ✅ Replacing a thin / outdated post with a stronger version on the same intent
• ❌ Multi-post content plan / calendar / strategy → Mode STRATEGY
• ❌ Changelog / feature launch → write manually
• ❌ Time-sensitive news → not evergreen, different rules

Core principle

Don't write a blog "to fill the calendar." Write the answer to a question users actually ask on Search or to an AI. One strong intent = one canonical post. If an existing post is stale → fix it; don't pile a new post on top.

Workflow

1 INTENT → 1.5 OVERLAP AUDIT → 2 RESEARCH → 3 FILE → 4 FRONTMATTER → 5 BODY → 6 PATTERN → 7 RULES → 8 LINKS → 9 VERIFY → 10 REPORT

Step 1: Classify intent

Pick one category (don't write a hybrid — split into two posts if needed):

Step 1.5: Overlap audit (mandatory)

Before writing anything new, inventory existing posts:

find {blog_dir} -name "*.md" -maxdepth 2 | sort   # common: content/blog/, app/blog/, docs/blog/

For each existing post, extract: slug, primary intent, funnel stage, evergreen/seasonal, stale/off-brand claims.

Decision rules

Gate

• Inventory read · No >60% overlap (or plan to update existing) · Stale-claim audit done (or N/A)

Step 2: Research (short but solid)

If the post mentions current facts, third parties, or specific numbers — verify before writing: competitor/partner facts, stats/dates/policies, app-store claims, pricing/fees.

Source priority: 1) official project source 2) official third-party source 3) project marketing docs 4) reputable recent third-party (≤1 year for pricing/stats).

Hard rules: no invented benchmarks · no invented comparisons · prefer durable claims over snapshot stats.

Gate

• Every non-trivial factual claim has a source · Sources noted for reviewer

Step 3: File output

{blog_dir}/{slug}.md

Slug rules: lowercase, hyphen-separated, language convention for non-English projects, maps to intent not campaign. Examples: what-is-{brand}.md, is-{brand}-legit.md, {brand}-vs-{competitor}.md.

Step 4: Frontmatter

---
title: "Title that matches the real query"
description: "140-160 chars. What this post does for the reader — not fluff."
date: "YYYY-MM-DD"
author: "{Brand} Team"
image: "/images/blog/{slug}.svg"
category: "<from project's category enum>"
tags: ["<brand>", "<topic>", "<intent>"]
---

Picking category: use the project's existing enum — never invent. Common: guide/tutorial (entity, glossary, use-case, FAQ), tips (optimization), compare/review (comparison).

Step 5: Body structure (answer-first)

{2-4 sentences directly answering the title question — no warm-up, no story}

## TL;DR
- Key takeaway 1 / 2 / 3

## {Primary heading 1 — the main answer expanded}
## {Primary heading 2 — supporting context}
## When to choose / use / consider {Brand}
## Important caveats

## FAQ
### {Question 1 — phrased exactly as users ask}
{Direct answer, 2-4 sentences}

## Conclusion / Next step

Why this works for AI systems: TL;DR + direct answer + FAQ blocks are the highest-cited patterns. AI tools quote concise answers; long intros get skipped.

Step 6: Pattern per intent type

• Entity / Trust: what it is, who it's for, where the core value comes from (sources, business model), who's behind it, support/app-stores/community.
• Compare: real comparison table (criterion × options); fair, no dunking without evidence; state when each option fits better; verify every cell.
• Glossary: define in 1-2 sentences at top, concrete example, end-user language, link to related guide.
• Use case / How-to: numbered steps (≤7), visuals, prerequisites first, expected outcome.
• Problem / FAQ: describe symptom, list causes by likelihood, fix per cause, when to escalate.

Step 7: SEO + AI optimization rules

• H1 = the real query · First paragraph answers directly (no warm-up)
• Headings tell the story (scan H2s → understand the post) · Tables/lists beat paragraphs
• Internal links to related posts + product pages · One consistent brand positioning sentence
• No keyword stuffing · No filler · Sensitive claims need evidence or safe wording ("may", "in some cases")

Step 8: Internal link rules

Minimum 2-4 internal links: 1 to core product/about page, 1 to support/help page, 1-2 to related foundational posts. Anchor text natural + varied (no repeated exact-match).

Step 9: Verify (pre-publish)

• Intent distinct from existing posts · Title = real query · First paragraph answers directly
• Table / FAQ / numbered steps where intent calls for them · ≥2 internal links, natural anchors
• No unverified claims · Frontmatter category matches enum · Sitemap updated (if required) · OG image exists · Reading level reasonable

Step 10: Report

## SEO Blog Post: {title}
### File: `{blog_dir}/{slug}.md`
### Intent: {category} · Primary query: "{search query}"
### Key internal links: {anchor → /path}
### Overlap audit: Checked {N} posts · Decision: {NEW / UPDATED `slug` / REFOCUSED}
### Items needing user review: {claim X — verify source}

Hard Rules (both modes)

• One idea per piece / one canonical post per intent — if you have two intents, write two pieces.
• No filler — every paragraph earns its place. No keyword stuffing — humans first, search engines second.
• Skip a slot rather than post filler. Iterate based on data — kill what doesn't work after 3 attempts.
• POST specifics: overlap audit before writing (Step 1.5 not optional); answer-first (TL;DR + direct answer in first paragraph); no invented facts/benchmarks/comparisons; stale-claim audit fixes outdated posts in the same batch; slug = intent; update sitemap if required.

Related Skills

Recommended Agents

Source: phuthuycoding/moicle — distributed by TomeVault.