Yesterday, I discussed what makes a good pilot project in inner source. The main thrust of the suggestion was not to start with a big bang but rather to choose a relevant but not too large project. This begs the question of complexity of projects, specifically viewed from an inner source perspective. How should you escalate and grow your ambition for inner source projects? I see a 1 + 3 structure of levels.
The first level is the natural occurrence of inner source projects, traditionally called submarines or skunkworks or the like. Such projects are projects that have been performed by engineers, typically without management approval, or at least outside the usual budgeting processes. They may or may not have been inner source but the nature of the project lends itself to bottom-up collaboration across (the sometimes useless) organizational structures.
The second level is the inner source pilot project level that I talked about earlier. These are projects selected for their limited complexity. Often they don’t generate revenue but at least provide engineering utility, like the aforementioned internal tools projects. Thus, engineers can collaborate, but no product is touched directly, and organizational complexity is kept at a minimum.
The third level is when you up the ante and go for revenue relevance (talking about vendors now; if IT is just a cost center, this level might include projects that power in-house tools with which money is earned). Most large companies which are successful in their respective markets will have structured their products into product lines. There is a natural collaboration opportunity between the product units on top of a product line platform; inner source can play well here, see our paper on inner source in product line engineering. A key assumption and complexity reduction here is that the product line is owned by one business unit, not many.
The fourth level then is when you collaborate across legal boundaries, whether within a company holding structure or across country boundaries. In this case, do you not only have all the above complexity, but you will now also have to solve the transfer pricing problem or your financial and legal compliance departments will stop you dead in your tracks. Transfer pricing is a key challenge that we are actively investigating in its multiple solution options, so if you are interested in this, please contact me.
These are the four escalating stages of complexity in inner source projects that I see, simplified and broken down into a neat list. Feedback, as always, is appreciated.
Go back to: Getting started with inner source.