Sean O'Driscoll has posted a really great POV on Online Community ROI on his Community Group Therapy blog.

Sean runs the MVP program at Microsoft, and has created on of the most mature influencer programs I've seen.

Sean's post breaks down ROI into three categories:
- Sales & Marketing
- Support
- Product /Program development

In particular, I found his thoughts on Affinity, Loyalty and Satisfaction interesting:

# Affinity / Loyalty / Satisfaction: Most companies of some size run some sort of annual broad customer satisfaction survey/research methodology. Most surveys are trying to get at driver analysis and correlation data. This might tell you obvious things like product quality is the #1 driver of Satisfaction, but how much does support contribute? What about content? other indicators. Most of us do this stuff. The relevant issue here is whether you embedded community participation questions into your satisfaction measurement process. Can you correlate any of the following to your users who participate in community verses those that do not? (If you don’t have the “verses” part, you have nothing!!). So do your users who participate in community have…

Higher overall satisfaction with your company? or particular product/service.
Higher likelihood to recommend?
Higher likelihood to repurchase? or purchase companion/related products/services.
Higher satisfaction with support?
Higher perception of Image or Brand?
hint: I didn’t just invent this list as random examples

I can attest to the fact that "If you don’t have the “verses” part, you have nothing". At Autodesk, when we presented data to senior staff, we were generally asked "compared to what"?

The ROI conversation is a much larger, and frankly, ongoing discussion that really needs to be set in the context of a particular business to have true value to that organization. With that said, I think Sean's observations are valuable, and I for one appreciate his perspective.