SEO · AEO · GEO Toolkit: how it came together.
How it came together.
- 01
Separate the signals: SEO, AEO, GEO
A page can pass classic SEO checks and still fail answer-engine readiness, so the toolkit scores and reports the two separately rather than folding them into one number.
Marketing & BD - 02
Audit what the engines actually read
The scanner reads the live page the way a crawler does: structure, structured data, and the lever most sites miss: which AI crawlers the site allows. The recurring surprise is that a brand blocks the search crawler, OAI-SearchBot, thinking it blocked a training crawler, and drops out of the answers it wanted to be in.
Research - 03
Simulate visibility, honestly
Live tracking of who an assistant cites isn't a solved, stable API, so the toolkit asks the question the way a buyer would, across variations, and checks whether, where, and how the brand shows up. A directional instrument, not a guarantee, and the UI says so.
AI & Machine Learning
Decisions on the record.
The few calls worth defending. Each one is a fork; the other branch would have been a different project.
Three separate scores: SEO, AEO, and GEO
Google ranking and answer-engine citation run on different criteria. A page can pass every classic SEO check and still never be cited in a ChatGPT answer because it lacks schema clarity, source attribution, or blocks OAI-SearchBot, the crawler behind ChatGPT's search results. That is a different crawler from GPTBot, which collects training data and has no bearing on whether you appear in an answer. Scoring SEO and AEO/GEO separately keeps that gap visible instead of folding it into one number.
A transparent, checkable readiness score
A marketing lead has to defend recommendations to a stakeholder, so each point of the readiness score maps to a concrete, checkable signal rather than a model's guess: every number traces back to something visible on the page.
First-party Search Console data, kept local
The common shortcut is to give a hosted SEO tool read access to Search Console and Analytics, handing your query history to a third party for a dashboard you could build yourself. The first-party layer talks to Google's own APIs locally instead, so the data stays put. Built and public as seo-gsc, MIT: quick wins, query clustering, content gaps, title and CTR opportunities, and a week-over-week report.