Good fit + recommended
Positive
The recommendation is appropriate to the buyer’s need.
Learn · AI recommendation behaviour
Being mentioned by AI is not the same as being recommended
A useful recommendation matches the buyer’s need, explains the fit and respects the business’s boundaries. maesee makes those differences observable.
Reading a response
Different states mean different things.
An educational guide to recommendation behaviour, not the framework or a scoring funnel.
AbsentMentionedShortlistedRecommendedRecommended with reason
Absent → Mentioned → Shortlisted → Recommended → Recommended with reason
A response may name a business without endorsing it. A bounded consideration set, an explicit recommendation and a supported rationale each tell you something different.
Recommending the right customers
More recommendations are not always better.
Good fit + not recommended
Gap
A suitable business was not put forward in the tested scenario.
Poor fit + not recommended
Positive
Correct non-recommendation respects a genuine suitability boundary.
Poor fit + recommended
Unsuitable recommendation
The assistant puts the business forward for a need it does not suit.
Rationale quality
The reason matters as much as the name.
Does the explanation hold up?
A recommendation needs a supported reason that connects the offer to the customer’s situation. Generic praise, incorrect eligibility claims or invented differentiators can undermine an apparently positive result.
Reliability and Recommendation offer related perspectives on that same response.
Buyer scenarios that do not name your business.
Scenarios are neutral by default, without naming the business being assessed. Reports show selected representative output and observed indicators, while the full testing instrument remains protected.
Counts and suitability findings describe tested behaviour, not rankings or guaranteed commercial outcomes.