What are GEO and AEO?
Generative Engine Optimisation and Answer Engine Optimisation. AEO is work on your own pages — structure, schema, answer-first content — so a model can extract you cleanly. GEO is work across the wider web — reference sources, editorial, review platforms, communities — so a model has independent grounds to name you.
Owned content gets you listed. Third-party content gets you chosen.
What's the difference between being cited and being recommended?
A model can read your page, use it to answer a buyer's question, and still not name you. That is a real result from an audit — the assistant cited the vendor's own site as a source, then recommended no vendor at all.
Citation means a model read you. Recommendation means it chose you. They fail independently, which is why they are measured as two separate numbers.
What can the free snapshot actually tell me?
It surfaces one or two concrete gaps — a competitor an AI names when it doesn't name you, or a place where it describes you wrong. It's a scan, not a measurement: enough to see whether there's a real gap worth looking at, not a number you'd report internally.
What's the difference between the snapshot and the paid audit?
The snapshot is illustrative. The audit is a measurement: a frozen, defined prompt set run repeatedly across AI assistants — ChatGPT, Claude, Gemini and Perplexity by default, or a specific platform added on request — with two specific numbers reported. Not a demo of the idea, the actual figure.
How much does an AI recommendation audit cost?
$3,000 for Base, $5,500 for Extended. The price is set by how much of your category gets measured — how many buying questions, how many competitors, how many runs — not by your company size or your revenue.
It is published, it is not negotiated, and it is credited in full toward the first month if you continue. A re-measurement afterwards, re-running the same frozen prompt set, is 60% of the tier price.
What do "unaided recall rate" and "head-to-head win rate" mean?
Unaided recall rate: the share of a defined buying-intent prompt set in which the model names your brand without your brand having been mentioned in the prompt, measured across AI assistants.
Head-to-head win rate: of the prompts where you're named alongside at least one named competitor, the share where the model picks you as the best fit.
The two are never blended into one score.
Why two numbers instead of one visibility score?
Because they answer different questions and sell different work. Recall tells you if the AI mentions you unprompted; win rate tells you if it picks you once it already knows your name. A blended score hides which one is actually broken.
Which AI platforms do you measure?
ChatGPT, Claude, Gemini and Perplexity by default, with the per-platform split reported alongside the blended figure. A brand can sit at 50% on one and 0% on another, and that difference is diagnostic rather than noise.
You can swap one of the four for a platform that matters in your market — DeepSeek, for an APAC-heavy category — at no extra cost. Adding a fifth is priced like any other addition to scope.
How long does it take to see results in AI search?
Measurable citation change usually takes two to four months. Entity-level recognition — a model describing you accurately and consistently — takes three to six.
The first re-measurement often shows entity and sourcing improvements before it shows any movement in recall, which is what the day-90 report leads with.
What do you guarantee?
- No guaranteed placements or analyst inclusion — coverage is earned, not bought.
- No numeric target promised at signature; targets get set as ranges after two real measurements.
- No attribution from a single AI answer to a closed deal — that's not something this measures.
What's measured, always: whether you're named, and whether you're recommended.
How is this different from an SEO tool or a rank tracker?
Search engines rank pages. AI assistants recommend vendors. Monitoring tools count mentions; they don't do the work of earning a citation or fixing what a model believes about you.
Can this be tied to actual pipeline or revenue?
Not directly. What gets measured is whether you're named and whether you're recommended — not what happens after someone reads that answer.
Still deciding?
The snapshot is free and needs no call. If your question isn't here, ask it directly.