Reviews

We tested ChatGPT, Gemini & Perplexity on the questions that matter

AI models comparison

The short version: we ask ChatGPT, Gemini, Perplexity, Google AI and Copilot the questions your customers actually ask before they buy, and we track who gets named and who gets buried. The pattern is remarkably consistent, and it is rarely the brand you would expect.

What we do. We take a category - say, "best off-grid solar kit for a campervan" - and put it to every major AI assistant, the way a real buyer would. Then we log which brands get recommended, in what order, and why.

Three things it shows, almost every time:

  1. The biggest brand often is not the one recommended. Budget and ad spend do not buy you a place in the answer. The brands the models name are the ones the internet talks about, in the right places.

  2. The models mostly agree with each other. If ChatGPT does not rate you, Perplexity and Gemini usually do not either, because they are reading the same sources. Being out of the answer tends to mean being out everywhere at once.

  3. It is driven by sources, not slogans. Reviews, Reddit threads, YouTube explainers, buying guides, editorial round-ups. The models trust those, and repeat them.

A real example. Ask for the best campervan solar kit and you get five confident recommendations, with reasons. In one test, a genuine specialist - great products, spends real money on Google - was not named once. Not top five. Not anywhere. The customer with £500 to £5,000 to spend was quietly pointed elsewhere, before that brand got a look in. No lost click to spot in the analytics, because there was never a click.

What it means for your brand. If the models do not name you, you are not in the running, and it is invisible: you cannot see the sale you never had a shot at. The first job is simply finding out where you stand.