Running Multiple Improvement Experiments in Parallel


Running multiple experiments in parallel

In Continuous Improvement Through Experimentation we described how a team could improve their way of working (WoW) via the strategy depicted in Figure 1.  In Better Decisions Lead to Better Process Outcomes we showed how you can increase the rate of improvement by identifying potential improvements that are likely to work in your situation.  In this blog we describe how to accelerate your team’s improvement by running multiple improvement experiments in parallel.

Figure 1. Running an experiment to evolve your WoW (click to enlarge).

Experiment to evolve your WoW

The strategy is exactly as it sounds.  Instead of running a single process improvement experiment at a time you run several at once.  You may decide to do this for one of several reasons:

  • You want to compare several strategies that potentially address the same process issue you hope to resolve.
  • You have several process issues that you hope to resolve at the same time.
  • Your team is currently running an experiment and feels it has capacity to take on another improvement experiment.
  • Techniques are dependent on one another, such as automated regression testing and continuous integration (CI).

Although this strategy will help your team to increase its rate of improvement, there are two fundamental challenges with it.  First, it is harder to assess the effectiveness of an individual technique when multiple experiments are being run in parallel because they can interact with each other.  For example, you decide to experiment with holding stakeholder demos and with active stakeholder participation at the same time (you weren’t doing either one until now).  When you see improvements in the quality of stakeholder satisfaction with the product you’re developing, is that the result of one of the practices but not the other or both of them in combination? If can’t definitively answer that question, it will make it hard to assess the effectiveness of each technique.

Second, you only have so much capacity for experimenting with new WoW. It can be hard enough, sometimes, to convince your organization’s leadership to allow you the time to experiment in the first place.  Getting time to run multiple experiments is even harder.

Our point is that running an experimentation-based approach to evolving your way of working (WoW) makes sense.  In some cases running multiple experiments in parallel can be even more effective.

MORE INFORMATION

For more information about choosing and evolving your WoW, we humbly suggest that you consider reading our book Choose Your WoW! A Disciplined Agile Delivery Handbook for Optimizing Your Way of Working. If you want to succeed at enterprise agile you need choices, not prescriptions.

Have any Question or Comment?

One comment on “Running Multiple Improvement Experiments in Parallel

Mark Lines

Scott and I find a lot of complimentary views with Dave Snowden, creator of Cynefin. He talks about running “Safe-to-fail” experiments in parallel. And those experiments should have a coherent view as to the merit of investing in each of them.

Reply

Leave a Reply to Mark Lines Cancel reply

Your email address will not be published. Required fields are marked *

Categories

Archives