Skip to content

Discoverability

The field guide to being found — by Google, Bing, AI assistants, and AI agents.

Being findable used to mean one thing: rank on Google. It now means three different things, with three different mechanics, and most advice online covers only the first:

  • :material-magnify: Search engines rank pages. Googlebot and Bingbot crawl, index, and rank. You win with crawlability, structured data, content, and links. This is classic SEO — still the foundation everything else stands on.

  • :material-message-question: Answer engines cite sources. ChatGPT, Perplexity, and Google's AI Overviews don't send you a ranked list — they give one answer and cite whoever earned it. You win by being the clearest, most retrievable source for intent questions, not brand searches.

  • :material-robot: AI agents call endpoints. The newest audience isn't human at all. Agents look for your product in MCP registries, resolve your /.well-known/* endpoints, and read your tool descriptions like search snippets. If you're not registered and connectable, you don't exist to them.

The one reframe that matters

Branded discovery is easy. Intent discovery is the game.

If someone searches your exact product name, you'll probably be found (though not always — see the brand-name tokenization crisis). The valuable, hard problem is being found by people — and AIs — who don't know you exist yet: "best way to add custom domains to my SaaS", "lawn care near Shakopee", "tool that can connect a domain via API". Winning those queries requires deliberate work on all three surfaces at once. That work is what this book teaches.

Where this book comes from

Every chapter here is distilled from real, shipped production work — not aggregated blog posts:

  • customdomain.ai — a SaaS product taken through a full AI-discoverability program: maximal schema.org graph, MCP registry presence, a GitHub discovery funnel, and a hard-won keyword strategy.
  • Heads Up Outdoor Services — a local service business taken through the local stack: LocalBusiness schema, live Google reviews, honest service-area pages, AI-crawler enablement.
  • everjust.app tenants — multi-tenant CMS sites behind reverse proxies, where a whole catalog of silent indexing traps was found and fixed.
  • The ever-just/agentskills repo — where each framework was first captured as a reusable skill. The skill index maps every chapter back to its sources.

When something failed, the failure is documented. When something is unmeasured, it says so. The book preaches authenticity as a ranking strategy, and it practices it.

How to use it

The master checklist

In a hurry? The master checklist is the whole book compressed into one actionable list, with links back to every chapter.