Distribution plan for a free dev tool. Roast it.
I'm building lighthouse-md.com · a free tool that turns a PageSpeed Insights audit into a CLAUDE.md fix brief for Claude Code. No signup, no paywall.
Free tool, single-person team, no revenue. Distribution is the entire game.
Here's the 8-week playbook I'm running. Open to roasting.
Already shipped
- 16 audit explainer pages at /audits, each with Quick Answer block, question-format H2s, and FAQ schema, tuned specifically for AI engine extraction (ChatGPT, Perplexity, Google AI Overviews)
- Chrome extension that does the same thing on the active tab
- /about as the entity anchor for Organization + Person schema
- Homepage at 90 Lighthouse perf, 100 across the other three categories (recursion as marketing)
Trying for distribution
Reddit, manual replies only. Hourly /new fetcher across 11 dev subs writes JSONs to a local folder. I read them and write 2 or 3 thoughtful replies a day. Karma first, link only when the asker would actually click.
dev.to drip, 2 posts per week for 8 weeks. All 16 audit .md files pre-generated with canonical_url back to mine. SEO credit stays here. Starts next Tuesday.
Awesome-list PRs. 3 of 4 submitted today (awesome-wpo, awesome-seo, best-chrome-extensions). Dofollow backlinks from DA-90 repos compound forever.
Show HN next Tuesday. Single biggest expected spike of the 8 weeks. Title is "Show HN: Turn a Lighthouse audit into a CLAUDE.md brief for Claude Code." Will live or die on the first 90 minutes.
Chrome Web Store SEO refresh. Title bumped to include "Lighthouse" (the highest-search keyword). 5 screenshots not 2. Version bump to surface in "recently updated" sort.
AI-tool directories. 18 ranked, 2 per week, tier 1 first (the awesome-list PRs above) then tier 2 (DevHunt, Product Hunt, TheresAnAIForThat).
Newsletter pitches. Smashing, JS Weekly, Frontend Focus, Bytes. Expect 1 or 2 of 5 to land.
Explicitly not doing
- No paid ads
- No Twitter/X grinding (no audience there, time sink)
- No cold outreach to agencies
- One announcement per channel, then content cadence (not "I built this" blasting everywhere)
Open questions for the room
-
Anyone landed Show HN with a free dev tool and seen sustained traffic 3 months later? The spike I get. I want the tail.
-
dev.to canonical credit: does Google actually respect the canonical when dev.to has higher DA than your own site, or am I deluding myself?
-
CWS listing rewrites: anyone tracked install-rate delta from a listing rewrite vs the baseline? Hard to find clean data.
The thesis under all of this: owning the AI context layer for one domain (web perf audits) becomes a defensible brand before anyone else figures out it's a category. Free tool today, brand play long-term.
What would you do that I am not doing?
This is a sharp distribution plan because it matches the product’s real buyer path: developers searching with a specific performance problem, not people casually browsing another dev tool.
The strongest part is the “AI context layer for web perf audits” thesis. That is much bigger than a free Lighthouse-to-CLAUDE.md utility. If you own that frame, the tool can become the default bridge between audit diagnosis and AI-assisted implementation.
One thing I’d pressure-test before Show HN is the brand frame. lighthouse-md.com explains the current utility, but it also ties the product very tightly to Lighthouse and one output format. That may be fine for search, but if the bigger play is owning AI-readable performance remediation, the name may start feeling smaller than the category.
Davoq .com would fit better as a hard technical brand shell if this grows into a serious dev workflow layer for audits, fixes, and AI coding handoff. Same product, same distribution plan, but with a name that can carry more than one tool page.
Since Show HN and directory listings can create early memory fast, I’d decide whether you want people remembering a narrow utility or a broader developer infrastructure brand before those assets compound.