Building a Backlink Prospector with Node.js and Exa AI

Cold email outreach for backlinks has a 0% reply rate in my experience (53 emails, 0 replies). So I built an automated prospector that finds link opportunities using Exa AI neural search.

How It Works

// backlink-prospector.mjs
const prospects = await exa.search(
  "Kazakhstan travel resources links",
  { numResults: 50, type: "auto" }
);

// Filter by: has outbound links section, DA < 40, relevant content
const targets = prospects.filter(p =>
  p.text.includes("resources") &&
  p.score > 0.7
);

Strategies

Strategy	How	Success Rate
Resource pages	Find pages with link lists, pitch inclusion	~5%
Broken links	Find 404s on competitor backlinks	~8%
Guest posts	Pitch unique angle to travel blogs	~2%
Web 2.0	GitHub repos, Dev.to, gists	100% (self-serve)

What Actually Worked

GitHub + Dev.to gave me 70+ backlinks from DR 90-95 in one day. Cold outreach gave 0 in 2 weeks.

Lesson: self-serve platforms beat outreach for new sites with DR < 5.