Thanks Trevor, hopefully it helps get this idea a little further ahead.

You bring up a really great point about the transaction details. From my experience, credit card companies simply aren't given the purchase details from merchants. Given how valuable purchase data is, I can understand why business would not share that data with a credit company. Mint.com for example, will categorize your spendings based on the merchant name. This would limit the logic behind the concept to encourage/discourage purchases at the vendor level.

If we were to challenge this limitation, I believe the options would be:
1. Manually have the person scan their receipt in via phone camera then process with OCR and such
2. Manually have the person indicate their purchases via an app
3. Work with stores to collect purchase details directly from them
4. Have the stores "do the math" and change the merchant name accordingly (e.g. "Safeway (a34)" means healthy purchases were made!)
5. Build a new payment processing system that collects line items on transactions

I'm really stretching it with some of those ideas (just to be thorough). I feel that really only #1 might be feasible if it was prompted only for certain merchants such as groceries stores while an inference is made for restaurants and others.

The merchant name may be all we will have to work with. From your point of view, how critical it is to solve that particular problem?

Hey Richard - thanks for adding me to your team here. This is the first team I've been part of, so when I saw the comment from Trevor, I thought it was an opportunity to help out.

In the past, I've written up some proposals such as the one that was suggested. I hope you don't mind that I've put together a first draft of how the prototype might look.

In summary, the pilot would take 4 months total and cost about $700. This would provide credit card transaction data for up to 100 testers and an additional layer of survey responses. This data would provide insight into whether the concept works and to what degree.

Here's the doc, feel free to edit and make your own: https://docs.google.com/document/d/1U1JTS2-HXHFer_0QbIgUZC-DCHT4HSRBNHPZl75QP-o/edit?usp=sharing

Chris - Those are some really interesting calculations! It's taken me from one idea to another.

Knowing the scale, we could motivate people by linking their workout to a good cause perhaps. Say there was a device intended to be sent to countries in need that needed to be charged. A workout could be framed according to the number of such devices being charged.

The one that came to mind was the "Crank Laptop" which took me here after a quick search: http://techreport.com/news/10420/100-laptop-ditches-hand-crank

Interestingly, the article talks about how the crank laptop has switched to a pull string because it creates energy more efficiently. I feel like this point links back to my comment about building an exercise machine that's designed to produce energy first.

The hand crank and the bike pedal system are one and the same in principle. In the effort to mimic the "upgrade" to the pull string - what if we redesigned the exercise bike to have an additional free spinning wheel ("power wheel"). Pedaling would keep the power wheel going at the most efficient rate possible while avoiding any actions that would create resistance. Another possibility would be that the energy is stored and "fed" to the power wheel as needed over time.

The question is whether a change like this would result in a meaningful increase of power generated. Any thoughts Chris?