Trust problem
A trust problem is when AI platforms know you do something well but don't recommend you for it, because too few outside sources back you up.
A trust problem is when the AI platforms know you can do something and still don't recommend you for it. Ask them about you, and they'll say you do it. Ask them who to choose for it, and they name other brands.
It sits alongside two other problems. A comprehension problem is when the platforms don't know something true about you. A perception problem is when they say something wrong. A trust problem is when they know and say the right things, and still don't put you forward.
How to spot one
Ask two kinds of question. First, about you: "Does Fernhill Coffee do office subscriptions?" Then, about the need: "What's the best office coffee subscription?"
For example, the answers to the first question say yes, Fernhill does office coffee, and describe it correctly. The answers to the second name Kestrel Roasters and two others, and never Fernhill. The platforms know. They just don't rate Fernhill for it.
Why it happens
When a platform picks brands to recommend, it looks for outside confirmation: reviews, articles, list pages and forum threads that name a brand for that need. If those name your competitors and not you, your own pages saying you do it well aren't enough.
How to fix it
The fix is outside your own site:
- Get onto the list articles and comparisons the answers use for that need
- Ask customers who use you for it to say so in reviews
- Give journalists and reviewers a reason to cover that part of your business
In AnswerOrbit
In Perception, each quality has an Association, how often the answers describing you bring it up, and a Standing, your rank when people ask about the quality itself. High association with a low standing is a trust problem: the platforms tie you to the quality but don't name you for it. The Gaps tab in Sources lists the sources that name your competitors and not you.