In my last couple of presentations both at university, companies and to friend I explain the “move towards software ecosystems” by showing a 80′s software stack that consists only of IBM tools and a “modern” software stack, with a lot of options for each layer in the stack. I often sidestep in my presentation or story by telling that this move also explains why I don’t expect that new big big 1,000 plus employees companies will spring into existence anymore. Since more companies focus on a small part or even a small component within the stack, there seems to be no need to grow a company up to these extends.

Companies will stay smaller (exceptions are there: Google) and have to work within the ecosystem to reach their goals. Those companies that best manage to operate in the ecosystem on all fronts (including, ecosystem strategic thinking, cooperation, knowledge management and so on) will be the winners of the future. However, in order to harden this claim research need to be done. I hereby pose some hypothesis, and maybe one or more members want to join me in writing an article about it, we’ll need some research money to gain access to the info. Slinger… are you / is the university interested?

Hypotheses: rough and need polishing, this is to get things started:

If: There is/was a change from monoliths to ecosystems
Then the following things must be true 

1. The average number of employees per IT company (separate legal or administrative entity) is declining.
2. The turnover (NOT profit!) per employee of the  companies closest to the end customer (most probably integrators) is increasing
3a. The relative turnover per employee declines if the “distance” between the end-customer and the software company increases.(Basic value chain hypothesis)
3b. There’s a decline in the average decline mentioned in 3a.

I came up with at least a dozen more hypothesis, but this I think is a good start.

How to research this?

a. I think we need to grab the Software Supply Network methods here, since we don’t have a way (yet) to measure distances to end customers  in full ecosystems.
b. A lot of data for the Netherlands is available in the databases of the chamber of commerce and the centraal bureau voor statistiek
c. I think if we carefully select a data set from IDN or Gartner we might have all the data we need for europe or the world.
d. We collect data from the 1980′s till now

Pitfalls

1. All hypothesis touch on “Efficacy” and “Efficiency”, which of course changed since the 80′s. So all analysis should correct for the standard increases over time.
2. Be careful not to measure basic value chain numbers, but longitudinal changes in the value chain numbers instead!

Since I already have a thesis subject, and want to finish that, I hope somebody else will take the lead in this. I am very willing to help, but I think this is in better hands when done by a PHD (student) or a student that makes a thesis out of this research. I think it would be wise to do this in cooperation with the Economics Institute of Utrecht University. They have far more experience in this kind of research, they might easily know where to get the data and the most common pitfalls. Since some of the teachers in MBI come from the department of Economics this shouldn’t be too hard to arrange.

Looking forward to your opinions and ideas.