How we collect data

Everything in your dashboard comes from one repeatable loop: run your prompts against the engines, read the answers, detect mentions and sources, and track the change run over run. Here is exactly how that works, including what the data can and can't tell you.

#Running prompts

A project's prompt set is the list of questions we send to the engines on your behalf. We submit each prompt to every connected engine the way a real user would, capture the full answer, and store it. Prompts run on a recurring schedule, so each run is a fresh, dated snapshot rather than a one-off.

#Detecting a mention

For each captured answer we determine whether your brand was mentioned. This is harder than a text search: brands have variations, abbreviations, and names that collide with common words. We match against the variations confirmed in your brand profile, which is why keeping that list accurate directly improves detection quality. The same logic runs for every competitor you track, so mentions are comparable across the whole field.

Once a mention is found, we read its context to capture prominence, sentiment, and any citations and sources attached to the answer.

#Headline metrics vs. trends

The headline numbers on your dashboard — Share of Voice, mention and citation rate, sentiment, top competitor — report your most recent run on each engine. They answer “where do I stand right now?” and refresh when your prompts re-run, so a strong new snapshot moves the number instead of being diluted into a rolling average. Each metric reads from the same latest run everywhere it appears, so a prompt card and the answer behind it always agree.

How those numbers got there lives in the trend charts. Every dated run is plotted so you can watch the direction of travel, and we keep the underlying answers, so any point on a trend line can be opened to read the exact responses behind it.

#Signal versus noise

Because answers vary, the headline snapshot is your latest reading — but whether a change is real is a question for the trend line beneath it. When a number moves between runs, ask: did it persist across several runs, did it span multiple engines, and is there a matching change in your sources or a competitor's activity? Sustained, multi-engine shifts are signal; a one-run blip usually isn't. That's why the dashboard pairs current metrics with trend charts rather than showing snapshot numbers alone.

#What this can and can't tell you

• It cantell you how you appear in AI answers for the prompts you track, how that compares to competitors, and how it's trending.
• It can'ttell you about questions you don't track, coverage is only as broad as your prompt set, so invest in prompts that reflect how customers really ask.
• It can'tguarantee an engine's answer to a given user at a given moment, since answers vary and are personalized. It reflects representative behavior, measured consistently.

For the list of engines and why they differ, see AI engines we track.