What a tracked prompt is
A prompt is one question, worded the way a customer would type it. "Best home care agency in Whittier for a parent with dementia" is a prompt. Tracking it means asking every engine that exact question at a set cadence and recording the answer each time. Over a few runs you get something a single scan cannot give you: direction.
Where the prompts come from
Renomly suggests prompts from real search data for your market, grouped by intent, and lets you add your own. Discovery prompts never contain your brand name, because the whole point is to see whether the engine volunteers it. Comparison prompts pit you against named rivals. Legitimacy prompts check what the engine says when a customer is deciding whether to trust you.
Each plan includes a number of tracked prompts alongside its scan runs, from a couple on Starter to a hundred on Agency. Rescans draw from the same monthly pool of runs, so there is no separate meter to watch.
What you see per prompt
- Named or not, per engine, per run, with the sentence that names you.
- Every other brand the answer named, so a lost prompt tells you who won it.
- Web search versus memory on discovery prompts, to separate being findable from being known.
- Change since last run, highlighted, so a scan reads as a diff rather than a fresh pile of data.
Cadence
Starter rescans monthly, Growth weekly, Agency daily, and you can run a prompt on demand at any time. AI answers move more than search rankings do, so the cadence you choose is a trade between spotting change early and spending runs. Most local businesses do well on weekly; agencies tracking many clients tend to want daily.
Prompt tracking is one half of the picture. The AI search tracking page covers the scoring that turns tracked answers into a number, and the visibility score page explains the number itself.