Off-Page SEO

On-page work caps out around DR + 20 in achievable KD targets. Off-page work expands that ceiling. This skill generates a prioritized backlink target list across three lanes — directories, listicles, guest posts — and a starter-stack checklist of universal SaaS submission flows. It does not submit anything. The user (or the user's outreach team) does the actual submission.

The off-page half of the SEO operation. Pairs with seo-machine (on-page programmatic sprint) — ship pages with seo-machine, queue outreach with this skill, run them in parallel.

On Activation

1. Load brand context: positioning.md, competitors.md, audience.md, voice-profile.md. Missing files are fine.
2. Surface what loaded:

   Brand context loaded:
   ├── Positioning    ✓/✗  (drives outreach angle + value-add framing)
   ├── Competitors    ✓/✗  (target list source — Recipe F runs on them)
   ├── Audience       ✓/✗  (selects niche-specific directories + communities)
   └── Voice Profile  ✓/✗  (shapes outreach email tone if user chains into cold-email)
   

3. If competitors.md is missing, prompt the user for 3-7 direct competitors before running Recipe F (the researched-target-list recipe needs competitor domains as input).
4. Check for .seo/backlink-targets.json in the current project. If present, this is a re-run — show the user the existing target counts and ask whether to refresh or just present the existing list.

When to use vs adjacent skills

The Starter Stack (universal SaaS submission flow)

Every new site should be submitted to these by week 4. Most are free; all are worth a one-time submission. This is the mechanical work — checklist, not creative.

Directories (submit and check off)

Community surfaces (selective)

Founder outreach (slow-burn channel)

• Founder link trades with complementary (non-competitive) products. Reach out to 5-10 founders of products your audience also uses. Offer mutual blog mentions, "tools we love" pages, or guest posts.
• Guest posts on DR 30+ niche blogs. Find via Recipe F (competitor referring-domains). Pitch 1-2 strong posts per month — not spray-and-pray.

Recipe F — Researched Target List

Use Exa MCP (mktg default) or Ahrefs MCP (if connected) to surface high-value link opportunities your competitors already have. The target list is built FROM your competitors' existing backlinks.

Exa path (mktg default)

For each top-3 competitor:

mcp__exa__web_search_advanced_exa
query: "[competitor-domain] -site:[competitor-domain]"
filter: linksTo:[competitor-domain]
limit: 50

Then for each result, scrape the page (Exa crawling_exa or firecrawl) and extract: domain, page title, anchor text. Filter to DR≥30 sites via a separate exa__company_research_exa lookup, OR by manual triage of well-known authority domains.

Ahrefs appendix (only if you have a paid sub)

If mcp__ahrefs__subscription-info-limits-and-usage returns data AND you want numeric precision on referring-domain DR, run:

mcp__ahrefs__site-explorer-referring-domains
target: <competitor-domain>
where: domain_rating >= 30 AND traffic_dofollow >= 100
order_by: domain_rating:desc
limit: 100

This replaces only the DR-proxy step in the Exa path above. Ahrefs MCP is not in mktg's chained-in ecosystem — keep this as a footnote, not a default.

Then for each candidate domain (Exa-path or Ahrefs-path), filter manually into buckets:

Output schema (.seo/backlink-targets.json)

{
  "generated_at": "YYYY-MM-DD",
  "research_backend": "exa" | "ahrefs" | "manual",
  "directories": [
    { "domain": "g2.com", "dr": 88, "url_to_submit": "https://www.g2.com/products/new", "competitor_uses": ["hootsuite", "buffer"] }
  ],
  "listicles": [
    { "domain": "thedigitalprojectmanager.com", "dr": 65, "url": "https://...", "topic": "Best Twitter monitoring tools", "competitor_listed": "hootsuite" }
  ],
  "guest_posts": [
    { "domain": "indiehackers.com/blog", "dr": 81, "topic_fit": "founder-led marketing" }
  ]
}

This file lives at the project root under .seo/ and is read by the mktg-backlink-prospector agent on future runs to avoid duplicate prospecting.

Parallel research via mktg-backlink-prospector agent

When the user wants a fresh target list and there are 3+ competitors to research, /cmo should spawn the mktg-backlink-prospector agent rather than running Recipe F inline. The agent fans out competitor-by-competitor research in parallel, returns the consolidated .seo/backlink-targets.json, and frees /cmo's context for the next decision.

See agents/research/backlink-prospector.md for the spawn contract.

Phase template (when seo-machine sprint integration is wanted)

When the user is also running a seo-machine sprint, off-page phases land near the end of the tracker:

### Phase N — Starter-stack directory submissions

**Why:** baseline backlinks from high-DR directories. Cap at 2-3 hours of submission work.

**Scope:**
- [ ] Product Hunt — coordinate launch with startup-launcher
- [ ] G2 — submit product page
- [ ] Capterra — submit product page
- [ ] SaaSHub
- [ ] AlternativeTo — list as alternative to: <competitor-1>, <competitor-2>, ...
- [ ] Indie Hackers — claim product page
- [ ] Crunchbase — basic profile

**Files modified:** none (external work). Track status in `.seo/off-page-status.md`.

### Phase N+1 — Listic