Why the Spotify Model Is Good AND Bad

 

The Spotify model is being touted by some as a way to scale Agile. While it has many useful insights, this is not what it was designed for. At Net Objectives, in 2010-11 we had created a way of coordinating teams that exactly matched the model but with different terms. When, the Spotify Model came out in 2012 I remarked to myself “cool, not a surprise since the chief consultants know Scrum, Lean and Kanban.” Good structures tend to follow the laws of Lean and Flow. I had seen different consultants who knew Lean creating similar results before.

While the model provides great insights on how to organize groups of teams as they grow, I didn’t really think about it much until recently. Unfortunately, it is now being used in ways for which it was not designed and which can cause damage. The first is when it is used to scale. Arbitrarily binding teams together in the model just adds a structure that makes coordinating the teams more difficult. Large orgs need to be decomposed, not organized together.

I have also seen it used to organize teams within a silo. This creates a structure to abide by that doesn’t add value. The overhead will make finding a useful solution more difficult.

See There Is No Spotify Model for Scaling Agile by Henrik Knieberg, one of the early coaches at Spotify who provided a window into how Spotify uses agile.  Here are some excerpts: in the words of its creators – “the ‘Spotify Model’ is not an Agile Method”, “Nor is Spotify a Scaling Model”, and “Mimicking the Spotify Agile Model is lazy.”

Personally, I'd run from any consultants promoting it as one instead of as a source for insights.