Get Found.

The product loop

Goose should feel simple even though the evidence machinery underneath it is deep.

Tell Goose the goal → Inspect → Decide → Change → Prove → Watch / revise.

• Inspect: observe the real site, intent ownership, entities/data, crawl/index state, repository ownership and applicable machine surfaces.
• Decide: return the top one to three moves that actually apply, with source/evidence and explicit unknowns.
• Change: prepare deterministic, reviewable mutations where source ownership and policy boundaries are known.
• Prove: distinguish implementation verification, deployment parity and later Search/AI outcomes.
• Watch: re-review recommendations when upstream guidance, site state or observed outcomes change.

Why no Goose score

Search and AI providers expose different evidence. Goose keeps stages separate instead of flattening them into a universal 0–100 number.

access → index/eligibility → exposure → citation/reference → brand mention → visit → useful action

Google generative-AI impressions are not Bing citations. Crawler access is not a citation result. Citation is not traffic or conversion. Missing evidence remains unknown.

Current priority: proof before more product

The immediate product milestone is real site evidence, not another module. Goose is prioritizing owned-site experiments with exact deployment parity, provider-native evidence and explicit keep/revise/continue/revert/retire decisions. Neutral and negative outcomes remain visible.

Evidence Lab · Portfolio Proof Board

What is under the hood

Advanced users can drill into Site Focus, Intent Ownership, Growth/Recommendations, Resolver, Repository Mapper, BraidGraph, Change Receipts, Controlled Cohorts, Winner Observatory, Recommendation Review and Project Maturity Surfaces. Those are internal engines and evidence layers, not required first-run choices.

Product line · Roadmap

Search / GEO boundary

Goose does not claim that special AI markup, schema volume, llms.txt, policy-page count or any one technical artifact guarantees ranking or citation. Current Google guidance says foundational SEO remains applicable to AI Overviews and AI Mode and that no additional special AI optimization is required for eligibility.

The practical focus is useful non-commodity content, intent/page ownership, crawl/index eligibility, truthful identity/data, evidence/provenance and real provider-native outcome measurement after deployment.

Google AI features guidance

Availability

The software is available without payment from the public GitHub repository and npm package under Apache-2.0.

GitHub repository · npm package

Why there is no rating in the schema

The page uses truthful SoftwareApplication/WebApplication semantics. Goose does not currently have grounded public rating or review data, so it does not fabricate either. Ratings, reviews, awards, customer counts and other fields are not added merely to make the product look mature or to imply rich-result eligibility.