AI NativeA scan is one measurement of how AI answers represent your brand right now. It asks your prompts the way a buyer would, reads the answers, and scores whether AI mentions you, recommends you, and how much of the conversation you own. This guide covers running your first one.
What a scan actually does
When you start a scan on a product, AI Native takes the prompts attached to that product, those persona questions discovered during onboarding, and runs each one several times across AI answer surfaces. It then reads every answer and works out, for each one:
- whether your brand was mentioned at all,
- whether the answer actually recommended you, not just named you in passing,
- the sentiment of how you were described,
- and which other brands appeared, so it can compute your share of voice.
For branded prompts, where the buyer names you directly, the scan also fact-checks the answer against the facts you verified in the Brand Truth Studio, so you can see where AI states something wrong about you.
A scan runs each prompt more than once on purpose. AI answers vary between runs, so a single answer is a sample, not a verdict. Repeating and averaging gives a number you can trust enough to act on and to compare against next time.
Demo versus live
A scan, like onboarding, runs in one of two honest modes.
- Demo produces a deterministic, believable result instantly and costs nothing. It is the right way to learn the screens and to give a walkthrough, but the numbers are illustrative, not real.
- Live queries real AI answer engines and reads what they genuinely say about you today. These are the numbers you act on. A live scan makes external calls, so it takes a little time and consumes credits.
Credits and cost
Live scans draw down your workspace credit balance. The pricing is transparent by design: you are billed the real, measured cost of the calls a scan made, multiplied by a fixed rate. Before a scan, AI Native shows an estimate that calibrates from your own recent scan history rather than a flat guess. After the scan, it reconciles against what the calls actually cost, so you are billed for reality, and every movement is written to a ledger you can read, with the underlying cost and the multiplier shown. Demo scans never touch your balance.
To prevent a runaway bill, a workspace can only have a few scans in flight at once; start another and AI Native asks you to wait for the running ones to finish.
Watch it run and read the result
Starting a scan takes you to its run page, which updates as the scan progresses from queued to running to done. When it finishes you see the headline numbers, the gaps ranked as actions, and, from your second scan onward, a before-and-after comparison with the previous scan so change is on the record.
You can also set a recurring cadence, daily, weekly, or monthly, so scans keep running on their own. That is what turns single snapshots into a trend line you can watch move.