KnewSearch measures whether a brand is cited inside AI answers on ChatGPT, Perplexity, Google AI Overviews, and Gemini. It runs real daily probes through DataForSEO against the buyer prompts that decide your category, records which brands and sources each engine cites, and tracks the changes over time. It is direct measurement of live engine output, not a modeled estimate: when a number moves, an engine changed.

The free scan is an AI-modeled estimate, and it is labeled that way on the report. It predicts what the engines likely say about a brand, and it is a fast, honest signal of whether you have a problem. Paid plans replace the model with real measurement: your top prompts run against the live engines every day, with history, competitor tracking, and gap recommendations.

Being linked is not being the answer. Pew Research found that only about 1 percent of visits with an AI Overview result in a click on a source link inside the summary. The value has moved from the link to the citation itself: whether the engine names your brand as the recommendation. And a link on one engine tells you nothing about the other three, because each engine cites a different web.

The foundations are SEO, and Google says so itself. Two things are genuinely new. The measurement: four engines answer differently and their cited sources rotate constantly, so the job needs daily probes, not a quarterly audit. The work: 84 percent of AI citations are earned media (Muck Rack, 2026), so the levers are third-party mentions, listicles, reviews, and community presence more than on-page changes. Your SEO instincts transfer. Your rank tracker does not.

No. Google has said it does not support llms.txt, and nearly all published llms.txt files receive zero requests from AI bots. What correlates with being cited is earned presence: brand mentions across the web outweigh backlinks roughly 3 to 1 in Ahrefs' 75,000-brand study. KnewSearch's recommendations focus there.

Measurement is immediate. Movement is not. Expect the first citation changes in roughly 4 to 8 weeks of executing recommendations. The churn works in your favor: cited sources rotate constantly, so openings appear every month for brands doing the earned-media work. We re-measure daily so you see the movement the day it happens.

We will not pretend the engines are ours to command, and anyone who guarantees their behavior is overselling. Here is the honest version, with the track record behind it: every client we have taken on has earned citations. What we guarantee is the work, and the refund: if your dashboard does not surface a single citation gap in your first 30 days on a paid plan, email us for a full refund.

Yes. The Agency plan starts at 499 dollars per month for up to 10 client domains at a flat rate, each measured across all four engines, with white-label PDF reporting. No per-seat or per-domain stacking. A partner program pays 30 percent recurring commission to consultants who bring their client lists.

You start a plan with a card on file. Your first real probes run the same day, so you see where the modeled scan was wrong within hours. Nothing is charged until day 15, and you can cancel from the dashboard in two clicks. The trial exists because two weeks of daily measurement is enough to watch your citation picture actually move.