Ten AI models, weekly and instant.
AI answers vary widely between platforms and change from week to week. Measure only one platform and you see a fragment – ask by hand now and then and you get a lucky hit.
Weekly and at the push of a button.
Every answer is evaluated: is the brand mentioned? Is the website cited? Who is recommended instead?
The cadence depends on the time since the last measurement, not on a fixed weekday. Without a previous measurement, the first one is due immediately.
“Re-evaluate” immediately queries all active prompts on all platforms again and evaluates the answers.
This run is billed in credits.You choose the prompts.
BrandZaps never creates prompts without asking – every prompt incurs costs from its first measurement. BrandZaps suggests prompts based on confirmed topics; a suggestion is only added once you click.
There is no fixed limit. Only duplicates are checked. That keeps the measurement volume a deliberate project decision.
Prompts via action cards in Project Chat →The ten platforms.
The individual platform pages explain how each measurement works.
Retroactive evaluation without new queries.
If competitors or brand spellings change, BrandZaps re-evaluates the saved answers. The historical series stays usable with the current classification.
In the full answer, your own brand, competitors and other brands are highlighted in colour. The measured values are kept permanently.
A failed query is shown but does not count as a measurement. A failed platform query is not interpreted as zero.
A complete time series across all relevant AI platforms – not just what a single answer says.
A time series instead of a lucky hit.
How mentions, citations, sources and competition develop over repeated measurements.