OP here. Thought a lot about the auction structure and decided the Dutch style (reverse) was the best given my goals.

I wanted to combine getting a good value with the ability to showcase the bid/price - and the process - publicly.

If I did a normal open auction, there is a chance everyone would wait until the last minute to bid. I'm managing bids manually, so it could get hectic. Plus, it wouldn't be as fun to watch for people reading.

Also, if there were NO bids early on it would really hurt the potential sales price with a lack of interest.

A full-on closed auction wouldn't be nearly as exciting for people to watch, and I wanted to generate buzz and interest, plus show the whole process.

With the reverse auction, the bid management is simplified - I only have to manage one winning bidder. It also allows the process to be showed publicly. Finally, it creates a sense of scarcity. If someone really wants it, they'll be more likely to bid early to prevent others from getting it.

So is it perfect? Definitely not! :-) Probably leaving some money on the table but I thought that this was the best structure to maximize value AND interest.

You forget the value of impulse when someone is bidding :)

This would maximize intrest, but not maximize value (money).

A typical (sealed-bid) auction has the exact same problem. I value the business at X, but know that I can bid less than X, probably still win, and keep more surplus.

To give bidders the incentive to bid their true value, you need a Vickrey auction where the winner only has to pay the next highest bid. But, this is suboptimal if there isn't a lot of liquidity. Also, you won't have the credibility to do this if you are auctioning off your own business (you have an incentive to lie about the next highest bid).

Ah, credibility issue. Never considered that one before. Couldn't you get around it by giving the bidders certain time slots to bid (and then reveal them in real time)? That way you could submit a "fake" bid, but you'd also risk pricing out the real bidders.

Since the first bid wins, the bidder in your example also risks losing his chance to purchase the business at his desired price by waiting.

Not true in the presence of other bidders, where the threat is that someone else's true valuation may be below X. This is game theory domain, and I haven't modelled it, but I'm quite confident the outcome for each player is to bid at their true valuation.

This was my opinion also. Thinking out loud, I think a regular auction might work better.

Downside of that might be that the reverse auction probably generates more buzz though... here we are, on the front page of HN.