Penlid
Start free trial
Free tool · no account

Does AI recommend you — or your competitor?

We ask one engine — Claude, with live web search — four of the questions your buyers actually type, then show you who gets recommended. Results come with real confidence intervals, because four samples is four samples and we won't pretend otherwise.

One engine, not four. Adding ChatGPT, Perplexity, Gemini or Google AI Overviews needs a configured key for each, and every report names the engines it actually sampled — never the ones we intend to add.

Takes about a minute. We fetch your homepage to work out what you sell.

Keep this tab open while it runs. The questions are asked one at a time, driven from this page, so closing it stops the run where it got to rather than finishing it in the background.

What this actually measures

Honest by construction

Unbranded questions

We never ask "is your brand good?" — that guarantees a mention and measures nothing. We ask the comparison and buyer's-guide questions a real customer types.

Recommended ≠ cited

Being listed as a source and being recommended as an option are different outcomes. We report them separately, because only one of them wins you the customer.

Intervals, not vanity numbers

LLM answers are stochastic. Every figure carries a 95% Wilson interval, and we show you the exact prompts so you can judge the sampling yourself.