I tested whether ChatGPT/Claude/Perplexity/Gemini recommend my own app.

Got zero mentions across all four.

Spent the last few days building what's basically the standard GEO playbook now, llms.txt, markdown mirrors of every page, honest comparison content against competitors, the works.

Then actually ran the test instead of assuming it worked: asked all four models "what's the best calorie tracking app" a handful of different ways. My app came up exactly zero times. Not even cited as a

source. Same five competitor names every single time, regardless of engine.

Not entirely surprising since most of the infrastructure went live in the last few days and these models aren't re-indexing in real time. But there's a difference between assuming that and actually seeing the zero. Planning to re-run the same test in a few weeks as a real before/after. Curious if anyone here has actually tracked this properly for their own product, what's a realistic timeline before you started showing up in AI answers, and did any specific tactic move the needle more than others?