Know-How Exchange : E-Marketing

It seems your constraints preclude using a Vickrey auction as the methodology. The bids in a Vickrey auction are submitted simultaneously, and without any prior knowledge of how others value the deliverable. As soon as you have sequential bidding you no longer have the conditions for a Vickrey auction.

So you need to decide (a) if using a Vickrey auction as your methodology is absolutely necessary, or (b) whether the importance of using Vickrey auction is greater than the value of sequential bidding. I don't see how you can have both.

Maybe you need to go back to the real objective of the research and the selection of a methodology that will give you what you need to best achieve that objective. Deciding on the tool before you are clear on the objective seems like backward thinking. You wouldn't buy a screwdriver for a woodworking project that hasn't yet been defined, would you?

Why can't you simulate the test using avatars to submit simultaneous bids based on different strategies, and then you can compare real-world results?

As I understand the Vickery auction, it's a sealed bid auction. That means bidding is done once and done blind (without knowledge of other bids). So long as the bidder understands the mechanics of the auction - and you'd need to be very careful to explain and show via examples how they work, and as long as there's no collusion/communication between respondents - each bidder will remain blind to other bids an so can be done sequentially keeping older bids hidden.

If you are planning to reveal a 'winner' you can delay the reveal, or in the worst case use artificial data - but the key is that it's not an open auction.

In analysis, you'll know the bid of each person and can use this to create a demand curve.