SKILL.md
Directory Submissions
Directories are the foundation layer of distribution, never the whole strategy. They pass link equity and put you in front of in-market browsers, but they only pay off when the pages they link to can convert. Build destinations first, then submit.
Three Hard Rules
- Foundation before submission. The landing page a directory will link to must be live, indexed, and carry: a single H1 with clean heading hierarchy, a real pricing page (free-while-in-beta counts), privacy policy and terms, logo assets (PNG, SVG, square, favicon), 5-8 real screenshots, a 60-90 second demo video, FAQ schema, and Organization/Product/SoftwareApplication structured data.
- Destinations before directories. Minimum before submitting anywhere: 3-5 competitor alternative pages, 3-5 use-case pages, a template gallery if the category supports one, and one honest "best of" post about your own category. Alternative pages convert several times better than generic content; that is where the link equity should land.
- Positioning varies by directory type. Never paste one description everywhere. Duplicate descriptions get down-weighted by AI engines and bore every audience. Startup directories get the outcome, SaaS directories get alternative framing, AI directories get the AI-first angle, dev directories get technical depth, review sites get ROI and use cases.
Workflow
- Readiness gate. Check every Rule 1 item plus: product publicly accessible, destination pages live and indexed, and 20+ users who could leave reviews. Any Rule 1 miss is a hard block, build the missing piece first. Thin reviews or no gallery is a soft block: launch, but flag the lost compounding.
- Pick tiers. Read references/tiers.md when selecting: it has the full tier catalog (flagship launch, startup/SaaS, AI, agent/MCP, no-code, listicles, integration marketplaces, and the rest) and the positioning variant per tier. Triage rule: submit only where the product genuinely fits; forced listings get rejected and burn the first-submission advantage.
- Prepare variants. Per tier: tagline under 10 words, 60-char short description, 150-word long description, 5-8 category tags, founder story. Vary the opening sentence, feature emphasis, and audience framing per tier.
- Batch submit and track. Keep a tracker (CSV or sheet) with columns: directory, tier, URL, date submitted, status, listing URL, dofollow (y/n), notes. Work through it in batches, log every submission.
- Run the anchor events. Product Hunt is the highest-leverage single submission and the easiest to waste; G2/Capterra listings are worthless without reviews. Read references/launch-playbooks.md when a PH launch or a review drive is in scope: it has the prep timeline, launch-day execution, and the 10-in-30 review protocol.
- Measure weekly. Domain rating, referring domains, listings live, reviews, signups from directory referrals, and a monthly manual check of where AI assistants mention the product for "best [category]" questions. If a number is flat, investigate; do not just submit more.
Verification
After each batch goes live, verify every listing link: run curl -sIL <listing-url> | grep -i "rel=" and inspect the anchor in the page source. Expect the listing live, pointing at the intended destination page, without rel="nofollow" where the directory claims dofollow. If a link is missing or nofollow on a directory that should pass equity, fix the listing or log it as low-value and deprioritize that directory.
Good vs Bad
Judgment call: product not launched yet, user wants "50 directories this week".
Bad: Blast the homepage link to 50 directories with one pasted description. No destinations, duplicate copy, equity lands on a page that converts nothing.
Good: Block submission, build 4 alternative pages and 3 use-case pages first, then submit in tier order with per-tier positioning, tracking dofollow status per listing.
Footguns
- Paying for submission services. These directories are free; the packages sell you an afternoon of copy-paste. Fix: do it in batches yourself, 2-3 hours each.
- Spam directories. Low-authority junk directories dilute the backlink profile and can trigger spam penalties. Fix: skip anything with no traffic and no editorial standards.
- Asking for upvotes on Product Hunt. Ranking algorithms and moderators punish it. Fix: ask for feedback, reply to every comment fast.
- Zero-review listings on G2/Capterra. A dead listing signals worse than no listing. Fix: run the 10-in-30 protocol or do not submit there yet.
- Dishonest comparison pages. AI engines cross-reference competitor claims and de-rank lies. Fix: honest "when to choose them" sections on every alternative page.
Completion Checklist
- [ ] Readiness gate passed or blockers named to the user
- [ ] Destination pages live and indexed before any submission
- [ ] Per-tier positioning variants written, no duplicates
- [ ] Tracker updated for every submission
- [ ] Live listings verified for dofollow and correct destination
- [ ] Weekly metrics baseline recorded
Any box unchecked: not done. Fix or say so.