ModelSaid: Find out if AI recommends your brand, or your rival

Hi Product Hunt 👋 I'm Erik. I built ModelSaid after watching a business lose a customer to a competitor who had never once outranked them on Google. The customer had asked ChatGPT for a recommendation, and my client simply was not in the answer. Nobody was measuring that. You can see every Google rank you hold, and nothing at all about the answer an assistant gives when someone asks it who to hire. So ModelSaid asks. Every scan runs the prompts your buyers actually type across five engines (ChatGPT, Claude, Gemini, Perplexity and Google AI Overview), then reports four things: - whether you were mentioned, and in what position - who got recommended instead, and how often - the sources those answers were built from, so you know what to go fix - a score you can re-run on a schedule and get alerted on when it moves Two things I would genuinely like feedback on: 1. Most tools in this space skip Claude and treat Google AI Overview as a paid add-on. I run all five on every scan. Is that the coverage you want, or is there an engine I am missing that matters to you? 2. The citations view is the part I am least sure about. It tells you which domains the models pulled from, which is the closest thing to an actionable next step I have found. Tell me if it reads that way to you. The scan is free and needs no card. Run it on your own domain and reply with what surprised you. If the result is bad, that is the useful case, and I will happily walk through what is causing it.