Ask ChatGPT, Claude, Gemini or Perplexity where to buy or sell and it picks one destination. We measure whether that's you, find out who it picks instead, and change it.
We measure what AI says about marketplaces, and how to change it.
Ask by name and every engine knows you. Leave the name out, the way a buyer does, and the engine chooses. If it doesn't choose you: no session, no seller lead, no renewal.
60 buyer questions you approve, asked of all four engines, every answer read by a person: were you named, were you the pick, and if not, who was.
Which questions send buyers to a bank, a rival or straight to your sellers. And which of your pages the engines actually read, which they skip, and which third parties they trust over you.
A ranked list each month, each move marked us, you, or both. Typical: one template change on your listing pages that credits every seller at once, the third-party pages engines read before yours, the app-store copy engines quote as your product description.
Six months, priced once, in writing. It opens with a baseline wave and an executive readout, so every move that follows is built on your own numbers.
We build the 60-question panel with your team, you approve every question, we run one wave across all four engines and read every answer. You get an executive readout: who's named, who's the pick, who took the buyer instead, and which of your pages the engines read.
Outcome: your own numbers on your own market, and a ranked list of what to change.
Get a proposalThe same fixed questions every month, so movement is real. Each month a ranked list of moves, marked us, you, or both, and a readout with the person who scored the answers.
Outcome: movement you can prove, wave over wave.
Your dealers or agents measured inside the answers, by region. More markets on the same basis. Weekly waves, a dedicated analyst.
Outcome: the whole network on one basis. Quoted per property, once, in writing.
Why marketplaces are scoped, not priced off a page. A marketplace has two sides, so the panel carries seller questions as well as buyer questions. Every answer is read twice: was the buyer sent to you, or to a dealer, a bank, an app store or a rival. And the seller network underneath you can be measured too, by region. Those three decide the scope, so we settle them with you in one 30-minute scoping call and put a single figure in writing.
We reply within one business day to book a 30-minute scoping call. The call settles the seller side, the network and the markets. The written proposal, with one figure and the 60 draft buyer questions, follows it. You approve the questions, then the baseline wave runs.
One figure for the programme, in writing: a baseline build, then a monthly fee for six months. Three things set it: how much of the panel is seller-side, the second read on every answer for where the buyer was sent instead of you, and whether your seller network is measured underneath. Network and portfolio work is quoted per property.
No. Everything is measured from the outside. We only need your approval of the 60 questions.
No, and we won't show you theirs. What you see about rivals is what the engines said in answers to your questions.
What sits in a model's training data. We measure what engines answer today, and we don't claim a single answer caused a single lead. One wave is a point in time, which is why the programme runs six months.
One 30-minute scoping call, then a written proposal with one figure.