Method
How Epovest measures
A measurement you can audit: real questions, official APIs, sampled rates, and every full answer kept.
-
1
You write the questions
Freely, in the language of your customers, with optional AI-assisted suggestions. One line = one question, asked exactly as written, without rewriting.
-
2
We ask the engines
At your cadence (monthly, weekly or daily), Epovest puts each question to each selected engine through its official API, several times per check depending on the resolution: 3 samples in HD, up to 12 in 8K.
-
3
We read the answers
The detectors look for your keywords and their aliases in every answer, read the tone for reputation trackers, and extract the sources the engines cite. Every figure is computed from the archived text.
-
4
You verify
The scoreboard, the tones, the share of voice: every number links back to full answers you can read yourself, engine by engine, check after check.
AI answers are stochastic: the same question can get different answers minutes apart. Epovest therefore never concludes from a single run: each check samples every question several times and reads a rate.
Collection and reading are separate
A tracker archives raw answers. Keywords and aliases are a reading applied on top: you can change the reading without touching the record.
The record
Collection, versioned
What defines the measurement itself. Changing it on a measuring tracker starts a new version of the series, so the points stay comparable; the past series is kept, never purged.
- The questions, asked verbatim
- The engines, the cadence, the resolution
- The full answers and their sources, archived
The reading
Interpretation, editable in place
What you read on top of the record. Change it at any time: the history is recomputed on the spot, at no extra cost.
- The keywords and their aliases
- The tracker name
- The discovery suggestions you accept or dismiss
Recomputing a reading is free because the raw answers are already archived: nothing is asked to the engines again.
The engines
Epovest asks ChatGPT, Claude, Gemini and Perplexity through their official APIs, with your exact wording, in the language you chose.
Each tracker picks its own set of engines; every engine multiplies the measurement, and the price.
What Epovest does not do
No promised rankings
Nobody can guarantee that an AI will cite you, and Epovest does not sell that. We sell the measurement: what the engines answer, tracked rigorously over time.
No black-box score
No opaque grade out of 100. Rates, shares of voice and tones are computed from archived answers you can read yourself.
No hand on your site
Epovest is a measuring instrument. It does not rewrite your pages and does not publish in your name: it tells you, factually, where you stand.
Common questions
How do I know if ChatGPT recommends my business?
You measure it. Create a Ranking tracker with the questions your customers actually ask (for example "best accounting firm in Geneva"), pick the engines and the cadence: Epovest asks them, counts who is mentioned, and shows your citation rate next to the names you compare yourself with. First checks within 24 to 48 hours.
What is GEO (generative engine optimization)?
GEO is the practice of improving how often, and how favourably, AI engines cite you in their answers, the way SEO does for search rankings. It also goes by AEO (answer engine optimization) or LLM SEO. Epovest is the measurement layer of GEO: it does not rewrite your pages, it tracks what the engines answer so you can see whether your work pays off.
Is Epovest a GEO tool?
It is the measuring instrument of the category: an AI visibility tracker. Optimization (content, sources, press) stays in your hands or your agency's; Epovest gives both of you the same verifiable numbers.
How is this different from an SEO rank tracker?
Same job, new terrain. An SEO rank tracker follows your positions on Google keywords; Epovest follows your citations in AI answers. Positions are ordered lists; AI answers are free text and vary from run to run, so Epovest samples each question several times and reads rates instead of ranks.
See what the AIs answer about you
Create a tracker in minutes: your questions, your engines, your cadence. First checks within 24 to 48 hours.
Sign-ups are paused: we're refining the offering.
No subscription. Credits from $10, valid 12 months. Pause anytime: the history stays yours.