NEWFree AI visibility report. Tracking from $99/month
Browse the docs

Prompt Battles.

Every buying question is a contest between you and named rivals. Prompt Battles scores each tracked prompt on the latest run as won, contested, or lost, and shows you who took the ones you lost.

What it is

Battles takes the same collected answers your scores are built from and asks a blunter question of each one: on this prompt, on this run, did you win the answer or did somebody else?

It is usually the fastest read in the product, because a list of lost questions is a to-do list in a way that a score is not.

Won, contested, lost

  • Won. You appear in the top three brands the answer names. In practice this is the band where a buyer actually notices you.
  • Contested. You are mentioned, but below the leading group. You are in the conversation without leading it.
  • Lost. You are absent from the answer entirely, and a rival is not.

Who won, and on what

For every prompt you did not win, Battles names the brand that did and shows the sources the platform cited when it answered. That pairing is the point: the winning brand tells you who to study, and the cited sources tell you where the answer came from, which is usually where the fix lives.

How to use it

Work the contested prompts before the lost ones. A contested prompt means the platform already knows you belong in that answer and is placing you low, which is a much shorter distance to close than a prompt where it has never heard of you in that context.

Then follow the cited sources into Sources to see which domains keep deciding these answers, and whether you are listed on them.

A felted indigo plug with its cream cable, ready to connect

Connected? See it work.

Start free. Your AI visibility report runs itself, and every integration on this page works from day one.

Free AI visibility report · No credit card · 50 prompts, 8 engines