(n.b. this is a continuation of this discussion)
Both situations have an individual being compensated (or having the potential to be compensated) for talking about something. Both situations have a behind-the-scenes intermediary (Ketchum in the former, BzzAgent in the latter) that is itself compensated to have individuals start a conversation. These conversations take place in situations where the other parties in the conversation would typically feel that the commentator is speaking from the heart, and not as part of a part of a program (or under contract). In both cases, the others in the conversation feel duped afterwards, upon learning that an interaction that seemed genuine was actually staged and part of a program of payola.
Despite all the metrics and process, I still feel the BzzAgent model is broken. How to fix?
1) Explicity lose the incentives (per here). If only a small portion of the BzzAgents are redeeming them anyway, what's the harm? Even if half the current participants drop out, there still are (if the claims are true) many tens of thousands of people who are participating.
2) Require disclosure. When BzzAgents are buzzing, anything less than stating (either verbally or in writing, if blogging, etc.) "By the way, I'm a volunteer part of an organization that's getting compensated to promote this product, and I will be writing a report on it at some time in the future," is disingenuous. Just say it. The Marqui people do. (I'm not thrilled with the Marqui model, but I do respect their upfrontedness about it.)
3) Aggressively change the meaning of the word "agent." For this to work, "agent" needs to mean "agent, as in catalyst," not "agent, as in shady operative."
Communication is good. An increase in interpersonal interaction is good. Making money is good. But doing the first two as a means to the third without disclosing it is a good way to rile up a lot of hornets. And that's not-so-good.