Tag Archives: Business model

Prediction to Evolution

15 Mar
Darwin's finches or Galapagos finches. Darwin,...

Image via Wikipedia

In one of my earlier blogs, i had mentioned about folly of predictions. After some careful thought, combining my interests in natural processes and business, this idea popped up.

If the cost of failure for any decision is low or none, should we predict or evolve?

perhaps the question is not worded right so here is an example which might drive the point home. The case i take is that of an Book publisher.

Publisher are skeptical by nature of the business, so they read through the manuscript of the authors before they decide if they will spend considerable resources to print, bind and distribute the books and deliver the text to the reader and still make a profit. they have their heuristics to decide, but this is the old way of doing business, read 2 years ago. The cost of failure in this model is high, a failure to predict failure will red line the final figures.

Enter the eBook publisher, they have a different business model, they enable the author to do most of the work for a fee and they take a cut of the total revenues, its an over simplification but the core of the business is more or less correct.  Their model uses modern technology to deliver the same text to the reader at fraction of the cost of the model earlier. Here the cost of failure is less, people will either buy or not buy, not buying your eBook is not going to adversely affect your bottom line badly, since much resources are not sunk in that activity.

So, if the cost of failure is close to zero, should the eBook publisher publish all and any manuscript that comes along its way, do its thing and put it on eBook shelves and see what happens. By doing this they will be the trend setters, some will click and become best seller, many will languish in the middle, and some will be just sinkers. Just as natural selection chose the best in a given condition, here market selection will decide what will be bought and what not, and just as evolution creates several ecosystems, you will also enjoy several reader ecosystems. So here there is no need of prediction, you evolve, or rather your evolution is the market itself. Perhaps alternate source of income like intelligence on readers. Perhaps long tail markets will come in play here.

There is no way one can predict the market to any degree of accuracy, instead of predicting  what it will do next, you are at every step when the market changes. Most forms prediction using big data revolves around serendipity, the case of Target knowing if a daughter was pregnant or not, i dont think they went around asking that question, but some one noticed it. This is a perpetual problem with big data mining, you don’t know the question, there are answers and mountains of it.

Well, i havent run the numbers to see if this will work out, i am going to do that shortly(perhaps run a simulation) and report back in another blog. But the idea is compelling.

%d bloggers like this: