Why Changes Should Be Called Experiments

Change agents have biases. Sometimes we make the changes we think makes the most sense. Sometimes those aren't the right changes though. Thinking about changes as being time-bound Experiments is a more effective way to manage change programs
June 13, 2014
Jason is the author of Lean Change Management and founder of the Lean Change Management Association and Spark the Change Toronto

I’m working with a large organization that is 7 months into an expected 3-year ‘transformation’ Today was the day we host our bi-weekly, in-person Lean Coffee where people can bring their questions about Agile, Scrum, Kanban or the transformation program itself to the table.

Not many showed up today, and we weren’t sure why.

Were people too busy? Did the people that attended previous sessions not find them valuable? Was there not enough focus for the sessions?

Sounds like a bunch of assumptions to me and the best way to challenge assumptions is to run an Experiment!

If you’re familiar with the first edition of the book, you probably recognize this diagram:

lean-change-cycle
After I had released the first edition of the book in the summer of 2012, I stopped using the term “MVC”, a term created by Jeff Anderson, and started using Experiments. There are a few reasons why:

  • MVC stands for Minimum Viable Change and I think ‘minimal’ is a matter of perspective.
  • “Minimal” doesn’t always mean small.
  • “Viable” also suffers from perspective bias.

Hence the new diagram!

lean change management cycle

The Lean Change Management Cycle

Using the term Experiment implies a few things:

  • We cannot predict, with complete certainty, the outcome of a change
  • An Experiment typically has a set of expectations and assumptions
  • Experiment deliberately implies, as a change agent, I’m starting with my own perspective, or bias, and I want to validate it by doing a small action
  • Using the word Experiment explicitly invokes a different thought process in the brain

Since we weren’t sure why the attendance at Lean Coffee was lower today, we talked about running a few Experiments:

  • Put business cards on the tables in every conference room: “Lean Coffee, June 21, 2pm, Room 002-129: Let’s talk about the Agile Transformation! Bring this card with you!” This would help us validate if this communication channel was more effective than the posters we put up.
  • Recruit people on Agile teams to create and present a monthly theme for Lean Coffee
  • Hand out flyers in the lobby…kinda like those annoying people in the mall!
  • Start tracking acquisition and retention numbers (how are people finding out about lean coffee, who’s coming back for more than 1 session)

The problem we’re trying to solve is whether or not the low attendance is an awareness problem, or interest problem.

Our hypothesis is that by running a low-cost, and potentially high value, Experiment, we’ll gain some Insights into what to do next.

This is a simple example.

When it comes to more complex change programs, like transformations or re-orgs, it’s perhaps more important to think of changes as Experiments. Reason being, the word “transformation” implies a definable end state. The “transformation” metaphor is less relevant in today’s global economy because the pace of change is constantly increasing. Thinking of changes as Experiments will create a new metaphor, and approach to transformation:  continuous change.

Continuous change is about create a series of small Experiments where the outcomes of these Experiments create Insights into new Experiments. Think of this approach as adaptive-planning or what I call, a ‘feedback-driven’ approach  to change.

Constant change can lead to change fatigue so the idea is not to constantly run Experiments. Sometimes it’s necessary to take a ‘wait and see‘ approach during organizational change. Decide how long to run certain Experiments before moving on to the next.

Next, it’s important to create the Experiment by provoking conversations with these questions:

  • Who’s our “customer”?
  • Why would this change (Experiment) help them?
  • How would we know if the Experiment was successful?
  • How long should we run this Experiment?

Outcome Based

Experiments force a discussion around outcomes. The challenge with these outcomes is that people often assume that a binary success or failure outcome can be tied to a change. The scientific method works great in science, not so much when the variables of an Experiment are the behaviours of people. Sometimes change is more art than science.

That said, changing the conversation to focus on the outcomes of Experiments puts the people affected by the Experiment front and centre. As a change agent, it isn’t about your or me. It’s about the people who ultimately have to live with the consequence of the change that matters.

Interested in learning more about Experiments? Sign-up and get a free sample chapter of Lean Change Management!

Get a FREE Sample Chapter

1 Comment

  1. Ella Mapple

    More and more companies are trying to get nimble to enable them to respond to change with agility. Over the years, there has been a clear shift in momentum about the ways how companies manage projects. So, the project manager should be a PMP certified, who can better handle the planning, execution, and closing of any project. To get yourself prepared for PMP Certification, http://www.pmstudy.com is the best source.

    Reply

Trackbacks/Pingbacks

  1. The best in change management and innovation - […] Why Changes Should Be Called Experiments | Lean Change … […]
  2. The Change Equation is Wrong | Lean Change Management - […] Given the pace of change, the rise of the creative economy and the nature of disruptive technology, the only…
  3. What will make your change program succeed? | Karthik Suresh - […] Instead of Vision and First steps you might have options and experiments. Jason Little has an interesting blog post…

Submit a Comment

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.