This is exactly the point of disagreement. If ESPN can consistently charge customers a set of prices, then their perception is *correct* and reflects reasonably the value to advertisers. And if they can make this happen on their own terms, then they will survive.

I realize -- and directly benefit from the fact -- that the entire online ad world is a gradual shift of risk from advertisers to publishers (and secondarily exchanges/networks), but strong, focused publishers can hold off this kind of risk accumulation indefinitely with a well run internal salesforce and disciplined remnant floor pricing.

The publisher capitulation that you are suggesting is excessive and unrealistic.

Regarding your argument about ESPN loss of pricing, editorial and account control - I disagree. Pricing should be according to a true value of the media which technology reveals. An "on average good price" or "local maximum" seems based on what ESPN thinks the price should be. I argue the price should be on the value of the media to advertisers.

The plug-and-play nature of the exchange allows for this insight. Sure, scale will play a part of it, too (more auction players, more bidding, higher price) but so will a better understanding of what the media represents bringing together the right buyers. Loss of account control is offset by the improved yield opportunity the exchange provides. This isn't to say that ESPN will let anyone (such as Sportsline) buy. Controls will be in place to continue to allow for brand control for the branded publishers such as ESPN.

As for "Selling across a scaled exchange's volume will certainly raise numbers somewhat, but the economic benefit of scale to publishers is a decreasing marginal return." Time will tell - respectfully, I disagree with you again. You appear to base your argument entirely on scale - I base it on scale and insight.

Your final words regardiing an illegal monopoly a la Google is an important point. The winner(s) of the exchange "race," i.e. the companies whose exchanges we all end up using, will deriving revenue per transaction rather than the black box or through selling "seats" on exchanges. APIs will allow value providers (behavioral, contextual, etc.) to offer their services to buyers and seller on the exchange for a small fee (minute per transaction). No doubt there will need to be strict regulation in place that prevents monopoly, trade front running, and more.

Perhaps you are being deferential to the highly speculative business plans you named this blog after.

There is an additional issue with exchanges that isn't being looked at. On its own, ESPN can sell media at an on-average good price (call it the "local maximum" for argument's sake). The exchanges have to use their scale not only to achieve higher revenue for ESPN, but also make up for ESPN's loss of account control and editorial control, and to provide profits for the exchange. How high must that exchange premium be in order to meet that economic burden, 20%, 30%?

Selling across a scaled exchange's volume will certainly raise numbers somewhat, but the economic benefit of scale to publishers is a decreasing marginal return. It's difficult to construe numbers that show success in any situation where the exchange is not an illegal monopoly a la Google.