« Cognitive Maps of Future History | Main | Perils of Prediction: The Elusiveness of Certainty and the Value of 'Simulated Hindsight' »

01 June 2007


Media Predict

We do encourage debate about Media Predict. In response to some remarks above, we would direct you to comments already made on Midas Oracle.

There, we said that the “infinite reflection” criticism is not a slam-dunk. The important point is that selection for publication is in many ways a distributed end-result question, albeit not on the scale of a national election: “The key thing for traders to bear in mind is that thousands of books are published every year. We’re just trying to predict whether this will happen to the proposals on our site. All books on Media Predict can secure publication via getting a book deal on their own, or via winning the contest.” Thus the projection at stake is whether these book proposals are of a quality to please the publishing community in general, Touchstone being one potential publisher.

In the context of your discussion above, choosing books according to sales projections is entirely possible, and indeed will be done at a later phase in Project Publish. However, lead-times are long in book publishing. A decision market based on sales projections would not run its course until two or three years from now. It would seemingly be difficult to recruit participation for such a long-term market at this stage.

In general Media Predict believes that markets can be used – in a wide diversity of ways – to improve the status quo in media content selection. In media, 10 percent of products generate 90 percent of revenue. The more we introduce markets into the process of content selection, the more this situation will improve for everyone.

Erik Sherman

I find myself of another mind here. I think there are limitations of what prediction markets can do. Just think about this: what’s the most unreliable type of polling/market research you can do? Asking people what they would do under a hypothetical circumstance. They pick the answer that makes them appear more the way they’d like to appear, and it’s all an unconscious psychological play. In this case, they're trying to be smart and pick the winner, because then they win as well. So it becomes a contest of outguessing everyone else, but, as you indicate, has nothing to do with what readers would actually buy, and that's the real question. Publishing houses have shown themselves inordinately capable of making bad decisions. That effectively reduces the prediction to how they'll screw up this time.

Readers Digest used to be bottom line smart about this whole issue. When trying to decide what types of articles they’d write, they would create a bunch of booklets on many topics. Consumers could get them for a nominal price, but they actually had to put money down. The results told the magazine where to aim its editorial focus. That way they spoke to their readers and got solid reactions. You could do the same with book readers. Offer articles or short stories on similar themes and see to which ones they gravitate.

Only, that still leaves the real problem - how can you ultimately tell what will burst out in success in ways that no one ever expected? Those are often the books that carry publishers into serious black ink.

This also makes me wonder whether publishers are any better at picking winning titles than they were when they were run by people who genuinely loved books.

The comments to this entry are closed.

My Photo

June 2009

Sun Mon Tue Wed Thu Fri Sat
  1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 30        

Sites, People, Blogs