Which AI assistant do you trust enough to report on to clients or your boss?

Marketers who report on AI search to a client or your boss: which assistant do you trust enough to put in the report? I ran the same 100 software buying questions through ChatGPT, Claude, Gemini and Perplexity every week for 13 weeks (July 26 to October 18, 2026). Same questions each week, so not 1,300 separate tests. For one brand, answers that named it: Claude 310, Perplexity 174, ChatGPT 14, Gemini 13, out of 1,300 each (1,298 for Perplexity). I don’t know why they differ that much. Do you pick one assistant, or show all of them side by side? (I work on AI visibility measurement, so I look at this a lot.)