The source of the information is surprisingly less important than the cues we use as heuristics to determine it's validity.
We need to feel that the contributor is writing in good faith. Over time we begin to recognize usernames and consider their information trustworthy. This affects member retention rates, but rarely the frequency of our contributions.
Contributions must genuinely reflect the thoughts, feelings, or experiences of the contributor. We don’t look at who they are to judge that. We don’t have the time (or inclination) to match every contribution against the perceived trust cues contained within the author's profile.
Instead we look for cues within the information itself.
If a review is too positive, too closely reflects corporate material, there are too many similar reviews, too many reviews from people without profile pictures, or something strikes us as unnatural – we consider the information untrustworthy.
We don't participate when this happens too frequently.
This is where allowing members to report suspicious contributions with a single-click, cross-referencing a contributor’s IP address with the headquarters of possible relevant organizations (surprisingly effective), and explaining to newcomers the typical structure of discussions/responses can have an impact
