AI brand perception: how AI describes your brand

A marketing manager asks this when an AI answer names the brand and then adds a sentence the manager did not write. An agency owner asks it before a client review. This guide shows how one perception theme scores, and reads the themes of our own project in its last 28 days. Our guide on AI visibility metrics defines perception and its theme score beside the other six metrics. For how one prompt set measures every rival at once, read our guide on competitor presence.

How do you measure AI brand perception?

You measure AI brand perception with fixed questions that name the brand, and you group the quotes of the answers into themes. A visibility question asks for the best tools in a category. A perception question asks what one named brand does well, or why not to buy it. So the answer describes the brand instead of listing rivals.

FullReach AI asks each platform eight fixed questions about each brand every week: five on qualities and three on objections. Our methodology states how the questions run. The Brand Perception page shows where the result lands. A theme is one label for every quote that means the same thing, such as "Pricing" for every sentence that calls a tool expensive. Our guide on AI visibility metrics counts the answers that described another company.

How does a perception theme score?

Inside one answer, the first theme gets 100 points, the second 90, and each later rank 10 fewer. Our guide on AI visibility metrics defines the average that turns these points into a score.

Take ten answers about one brand. Three raise a theme first, two raise it second, and five never raise it. The theme scores 48: 300 plus 180, divided by 10.

Read a quality score and an objection score in opposite directions. A quality at 20 means the answers raise it rarely or late, not that they rate it poorly. An objection at 80 is the worse result, because the answers raise it early and often. Read the answer count beside every score. In the 28 days to 30 October 2026, 82 answers scored our qualities and 38 scored our objections.

What does AI credit a brand with?

The answers credit every brand in our project with the same first quality. In the 28 days to 30 October 2026, on 7 AI platforms, "AI Search & Visibility Tracking" scored at least 77 for all six brands. For our own brand it scored 86, raised in 72 of 82 quality answers. The five rivals scored 77 to 89 on the same theme, so it does not separate the six brands. The table lists our five highest qualities in those 28 days, with the count of the five rivals that scored higher on the same theme.

A row with several rivals above us shows a theme where their case reads stronger. "Monitoring Coverage & Scale" scored 45 for our brand in the 28 days to 2 October 2026. In the 28 days to 30 October it scored 20. The quotes behind it name the platform count and the daily run. A theme that falls is a reason to open its quotes and the pages the answers cite.

The answers describe the category before they describe any brand in it. So the first question is whether the answers name your brand in that category at all.

What does AI hold against a brand?

The answers raise the same objections against every brand in our project, at different strengths. The table pools the 28 days to 30 October 2026 on 7 AI platforms.

"Pricing" is the highest objection for four of the five rivals, and it hits Scrunch hardest, at 83 of 100. Against our brand it scored 57, from 24 of 38 objection answers. Profound is the one brand where "Usability & UX" scored 50, 23 points above the next brand. The quotes behind our "Pricing" objection name the monthly price and the trial length, and set free checks against them. The quotes behind our "Data Quality" objection say that the brand has few independent reviews yet, and that AI answers change from week to week.

Our "Pricing" score rose from 23 in the 28 days to 2 October 2026 to 57 in the 28 days to 30 October. "Feature Limitations" appeared against our brand for the first time in those 28 days, at 50. "Privacy & Security" fell from 53 to 18 between the same two windows.

How do you act on an objection?

Answer the objection on a page of your own, with the fact the quotes ask for. The methodology grades that move and names its source. FullReach AI turns an objection that three or more answers raise against a rival into an Opening in Actions, with the quotes as evidence. Each Opening names the rivals the objection hits and the platforms that raise it.

Three moves follow from a theme:

1. For an objection against you, publish the page that states the fact: the price, the trial terms, the review count.
2. For an objection against a rival, publish the comparison that states where you differ.
3. For a quality the rivals score and you do not, check whether your pages state it at all.

Our guide on getting mentioned in ChatGPT answers reads the pages ChatGPT cites and lists the four moves that follow. The perception questions run every week, and each run shows whether the quotes changed after you published.

What to do next

1. Ask one platform what your brand does well, and why not to buy it. Note each sentence that names a fact.
2. Check whether each answer describes your company or a namesake.
3. Pick the one objection with the highest score and a fact you can state. Publish that page first.

For how FullReach AI collects the answers, read the methodology.