peekr is an AI search visibility tracking tool. It measures whether generative AI assistants name a brand when they answer questions in that brand's category.
peekr asks the generative engines listed on its help page a set of real buyer questions, repeats each question multiple times because generative answers are not deterministic, and stores every answer in full with its timestamp. From those stored answers it reports four things. Visibility: the tracked questions in which the brand is named. Average position: how far into the answer the name appears. Share of voice: the brand's share of mentions against the competitors named alongside it. Negative mentions: answers that describe the brand unfavourably, including claims that contradict the brand's own published facts.
Because the same question can return a different answer, and a different set of cited sources, on each run, peekr treats one response as a sample rather than a result. Questions are tracked one by one, so a team can see which specific prompts name the brand and which name a competitor instead, rather than a single blended score.
peekr also runs a page check on the site that was submitted: what an AI assistant can and cannot read there.
Anyone can run a free snapshot from a single homepage URL, with no account and no card. The free limit counts runs per person rather than per site, and resets daily.
peekr is early. We have not set a price for continuous tracking yet, and this profile will say so until we do. We publish what we measure on our blog, including the experiments that did not work.
The current list of engines we query is published at trypeekr.com/en/help#engines.
