Decisions – even strategic decisions made by senior leaders – aren’t always made in a systematic, let alone scientific, manner.
Anyone who has spent some time in organizations can attest to that. And organization design decisions have traditionally been among the least systematic – and the most intuitive and politically charged.
Even internal consultants and advisors who help executives sometimes accept this as the irrevocable status quo.
Last year via Linkedin, I communicated with a senior advisor in a large and global company. When I mentioned the methodology that I have been using, his response was:
In our company, we have a pragmatic approach toward organization design…I am not sure whether top management would be positive toward a formal process…decisions are often based on “gut feel”.
Well, that may be. But this situation is not inevitable.
I have myself been part of many organization design projects where we have succeeded in running a systematic process. And I know of many excellent consultants in the field who are capable of doing that.
Thanks to forward-looking leaders and excellent internal advisors, some firms are also able to gradually increase their maturity level and institutionalize a systematic approach to organization design.
Recently, I have been thinking that it should be possible to go even further – and not only be systematic – but to aim for a scientific approach. At least in those companies that are already at a moderately high maturity level.
What does “scientific” mean in this regard?
First, it means that we use a data-driven approach to inform the decisions that we are making (as I talk about in the email tutorial that I am running).
But once we start collecting data, we may also be able to measure the effects of a re-design.
Measurement is not the goal per se. The goal is to evaluate whether we achieve the intent we had in mind when making the decision. To learn what we need to focus on to further improve.
Basically: To experiment and treat your organization as a lab. Which in turn requires a particular mind-set, and perhaps a particular culture as well. As author (and former rocket scientist) Ozan Varol puts it:
To experiment is to be humble — to acknowledge that you’re uncertain how your idea will pan out.
Impossible in a real organization, you would say?
I came across an inspiring example in an article* written by a team led by Jan Vlietland, a well-respected professor in IT management in the Netherlands.
In the article, they describe how they helped six scrum teams in a large, multinational bank.
They begin by pointing out that Scrum teams are supposed to be relatively small; but this means that they must divide responsibility for different parts of a system between them, which in turn creates interdependencies that must be coordinated.
In the bank, the Scrum teams didn’t collaborate effectively and had problems synchronizing their deliverables.
Vlietland and his colleagues worked with the team to improve their procedures, for example, with regards to the feature backlog. But they also adjusted the organization, by introducing a Product team to coordinate the work between the Scrum teams.
But they didn’t leave it at that. They collected data over a time period of 9 months to check whether the intervention worked. And it did: On average, delivery time went down from 29 to 10 days!
This is how organization design should be done, and how social science should be done!
(My only modest proposal would be to say that could also have looked at the configuration of the teams; there is no information in the paper about whether the team configuration was optimal.)
By the way, if you are involved in an organization re-design effort in your organization and would like to do something similar, we should talk! Send me an email or use the Contact form on this blog to get in touch.
*Vlietland, J., van Solingen, R., & van Vliet, H. (2016). Aligning codependent Scrum teams to enable fast business value delivery: A governance framework and set of intervention actions. Journal of Systems and Software, 113, 418-429.
Image source: https://www.freepik.com/vectors/scientist
Related blog posts:
Does it take longer to use a systematic approach?