Perception problem

A perception problem is when AI answers paint your brand wrongly because the pages they read tell the wrong story, while the product itself is fine.

A perception problem is when the AI platforms say something unflattering or untrue about you, and the cause is the pages they read. The product itself is fine. The story about it, as the web tells it, is wrong or out of date.

It's worth naming because it changes what you do next. If a criticism is true, it's product feedback, and the fix is in the product. If it's a perception problem, the fix is in your content and your press: the pages the platforms read.

How to tell the two apart

Ask whether the criticism is true today. Then look at where it comes from.

A perception problem usually traces back to a small number of pages. An old review written before you fixed something. A forum thread from three years ago that keeps getting read. A comparison article that lists last year's prices. A competitor's page that describes you unfairly. When one or two pages sit behind most of the answers that repeat a claim, you're looking at a perception problem.

When the same criticism comes up across many independent sources, recent reviews among them, it's more likely to be true. Take it to whoever runs the product before you try to fix the story.

An example

Answers keep saying Kestrel Roasters' coffee "often arrives late." Kestrel checks its delivery data: almost every order this year arrived on time. It looks at the pages behind the claim and finds one review roundup from two years ago, back when Kestrel used a different courier, that most of the answers read.

That's a perception problem. Kestrel writes to the roundup's editor with its current delivery record and asks for an update, publishes its delivery times plainly on its own site, and asks recent customers for reviews.