I’ve been thinking more about time-based & ‘sponsorship’ ad models since reading about Project Wonderful and Coudal’s The Deck. Project Wonderful and The Deck serve very different audiences, and they operate differently, but the core principle’s the same – the advertiser’s not buying impressions or clicks. Instead, they’re buying the right to show their ad on a publisher’s site for a period of time. If I agree to pay $100 a day to show my ad, I’ll pay that whether it’s shown a million times or a thousand. In other words, it’s a forward contract on a site’s inventory.
Here, both parties are taking risk. The publisher risks underestimating what his future site will be worth, and selling the advertising rights for too little. On the other side, the advertiser risks overestimating the value of the inventory. Just like a regular CPM campaign, where payments aren’t based on performance, the advertiser risks overestimating the quality of the audience and therefore paying too much. However, the advertiser also risks overestimating the size of the audience, which could decline between the time the advertising was booked and the time it runs. The overall cost is fixed – a decline in the number of impressions just raises the cost of each one.
My biggest initial concern: the risks are all associated with the performance of the publisher’s site, and publishers know more about their sites than the advertisers. (The publisher knows that they’re going to take a week’s vacation next month and aren’t going to update. The publisher knows that the traffic stats from the last month are a complete fluke based on an exceptionally popular outlier. The publisher knows that traffic always takes a header when school lets out for the summer. And so on…) Traffic stats help even the sides, but they can only do so much. That creates a moral hazard – this pricing model makes the most sense for publishers when they know their sites are going to underperform what advertisers would expect from publicly-available information. On the other hand, publishers who anticipate that they’ll exceed expectations are better off selling CPM or any other type of performance based advertising.
I haven’t seen this moral hazard in the wild – perhaps because the existing systems either only accept publishers with good reputations (The Deck) or are rooted in a supportive community (Project Wonderful, which started with webcomics.) Looking forward to being accepted into the Project Wonderful program, and posting some more detailed impressions.
{ 0 comments… add one now }