AI recommends Zendesk most and puts it first least. Which column would you buy from?

I run an open index of what ChatGPT, Claude, Gemini, Perplexity and Grok recommend when someone asks for software. Yesterday I posted an agency version of this on LinkedIn, and an agency owner replied within fifteen minutes with the number that stuck for him. A tool named in 84% of answers and never first.

So today I ran the category most of you have bought from at least once. AI customer support agents.

63 sampled answers across the five assistants, four weeks to 9 September. The same six to eight neutral buying questions to every engine, no product names in the prompts, one prompt one engine one call, production default settings.

Zendesk is recommended in 45 of the 63 answers, more than any other tool. It is the first name in 11. Intercom and its Fin agent, counted together, are first in 25.

So the category has two leaders depending on which column you read. Read recommendations and Zendesk leads. Read the first name and Intercom does.

The engines do not agree with each other either. Claude recommends Intercom most, ChatGPT and Perplexity recommend Zendesk most, Gemini has five tools level on 14 answers, and Grok, on 7 answers, has Zendesk and Intercom level.

What I take from this as a founder rather than as the person running the index. If you are Freshdesk or Gorgias in this list, "we are recommended by AI" is a true sentence. It hides that you are the first name in six answers and five. The column that probably decides the click is the second one, and I have not seen anyone measure it.

The limit, out loud. 14 answers per assistant and 7 from Grok. Enough to show the split, not enough to rank the middle. The full ranking with intervals is on the index page for the category.

Same offer as last time, still open. Name your category in the comments and I run it the same way, both columns, and post it here.