The Economics of Analytics (I)

I’ve been working for quite some time this month on the problems around the Economics of Analytics.

In what I think will become a series of posts – I talked with some fairly brilliant people and I’ll be getting their permission to quote them over the next few weeks.

I spent a lot of time figuring out how to ask the baseline questions that were troubling me.

At the highest level, the argument runs like this:

It starts off with private discussions. We lamented that “the hourly model is broken” and were generally “frustrated with the way that the market rewards performance” in the client-agent relationship. I was also particularly frustrated with an inability to understand the scalability of our methods in Analytics and the social technologies that can change the economics.

This led a careful exploration of incentives and risk communication.

When I argued all the way through the assumptions, I realized that this was increasingly a not a simple “cost of agent = benefits of optimization – margin” kind of a solution.

—————

There was a period of time during the ‘role of analytics’ fights where the discussion shifted, quite suddenly, from the legitimacy and effectiveness of Business Analytics, to how a specialized skillset could be applied to business problems.

We quickly settled on the accepted thesis that companies should focus on core competencies, that Analytics was not in the DNA of most organizations, and so a client-agent relationship made the most sense.

Once we got over the “access to data” consraints – (don’t assume constraints that arn’t there) – we get into how to actually price Analytics.

The more I worked on the issue of pricing, the more I understood that this problem was just a sub-set of all client-agent pricing problems. There wasn’t anything in this that required wholesale re-invention.

The mental hangup, at the time, was that I could not express the value of an analytics programme in a single sentence, in much the same way, five years ago, I could not express the value of public policy analyics in a single sentence. There wasn’t a magic Lowest Common Denominator arguement I could make.

Until one night at Bier Mrkt.

In the next post – I want to hit the aspects of risk management and trust in client-agent relationships.