AEO / GEO scoring | Ora

What Discovery measures

Discovery is weighted 20 of 100 in the overall ora score. Every check is computed from a public URL with no authentication, no self-reporting, and no panel data.

CheckPoints

• Category share of voice (beta) AEO
  Run a category prompt across multiple AI search engines in parallel; rank where the product appears against competitors and report a recommended % score.
  Score: 6

• Developer resource discoverability AEO
  Brand-name search returns the product's developer resources (API docs, OpenAPI spec, dev portal). The cleanest single AEO signal.
  Score: 3

• Brand name discoverability AEO
  Clean brand-name search (no developer keywords): does the product's domain surface in the top results? The baseline for any implicit AI discovery.
  Score: 3

• Wikipedia / Wikidata entity presence (bonus) GEO
  Verified Wikipedia article + Wikidata entity with P856 back-reference to the domain. Wikipedia is the single largest source of citations in AI answers.
  Score: 4

• Listed in MCP registries
  Smithery, mcp.so, Glama, PulseMCP.
  Score: 1

• NPM/PyPI SDK package
  Published SDK or CLI package matching domain or product name.
  Score: 1

• Agent platform configs
  Published agent rules/configs (.claude/, .cursor/, .windsurf/) on GitHub.
  Score: 1

• Listed on skills.sh
  Official skills published on the skills.sh agent skills directory.
  Score: 1

How ora is different

The enterprise answer-engine analytics category - tools that track brand visibility inside ChatGPT, Claude, Perplexity, and Gemini - has emerged in the last eighteen months. ora covers the same AEO/GEO measurement surface and four additional layers, with a few deliberate differences.

• Public URL only
  No login. No tracker. No browser-extension panels. No CDN log integration. Every check is reproducible from the open web, which is exactly the surface answer engines see.

• Every domain in our index, scored automatically
  Enterprise answer-engine analytics tools price by tracked-prompt count and gate the interesting features (full engine coverage, prompt-volume datasets, agent analytics) behind custom plans. ora gives every domain a public AEO/GEO score and a public history.

• AEO/GEO is one of four layers, not the whole product
  Visibility tells you whether agents find you. ora also scores whether agents can access your data and understand you (Access), whether they can authenticate, integrate, and complete a task with you (Usability), and whether they can transact with you (Payments). Visibility alone is a vanity metric if the auth flow blocks the agent at minute two.

• Open methodology
  Every check is documented, every weight is in source, the scoring formula is in the repo. No black-box visibility score - you can see why your number is what it is and what changes will move it.

How it relates to Access

Discovery answers can the answer engine reach you. Access answers once it does, can it read and understand you. The two layers are coupled - a great llms.txt with a hostile robots policy still loses Discovery, and great Discovery with no JSON-LD still loses Access. ora ranks them separately so you can fix the layer that's actually limiting you.

See the full methodology for check-by-check scoring and per-layer weights, or run a scan from the homepage to see your AEO/GEO breakdown.