A cooperative can invest in a service that a part of the members uses little. Low usage does not automatically prove useless: knowledge, skills, time or suitable conditions may be lacking. Before changing the service, the cause must be identified.
I propose to distinguish those who do not know the service, those who have tried it and those who use it permanently. The questions to the three groups should be different. A problem in starting requires a response other than a benefit that is too limited compared to the cost incurred by the member.
In a hypothetical example, a commercial platform offers useful functions but requires complex initial input. Support can improve access; if the problem concerns a function of little relevance, training alone does not solve it. The result should be verified after the intervention.
The collective cost must also be made visible. A service used by a few can have a justified specific function, or absorb resources without sufficient benefits. The evaluation should link usage, recipients, results and alternatives. This allows you to choose whether to maintain, adapt or reduce it on the basis of observable elements, without interpreting each difficulty as resistance of the members to change.