“Change is easy, improvement is far more difficult” – Dr. Ferdinand Porsche
When we make a change, we want to make things better. However, it’s not always easy to ensure that the change is actually positive overall. How can you increase the chances of actually making an improvement?
Good news, there’s a set of simple (not easy!) steps you can follow to vastly improve the odds on hitting that improvement you are seeking.
First, be very clear what the problem is. Write it down. State it in the simplest possible terms, which means you might need to refine it several times. Get specific, watch out specifically for weak or ambiguous terms. “We’re slow” is a very weak problem “We consistently take twice as long as our initial estimate to launch a product” is much stronger.
When you have a strongly stated problem, you can then work on what that improvement would look like. Do you want to improve your estimates, reduce the actual shipping time even if the estimates are still bad, or do something else entirely?
Next up, get explicit about what you are willing to spend to seek improvement. Are you going to invest more resources? Drop something that’s not important or high value? Maybe even make something else harder or not as great as it used to be?
Now you get to start trying things. You’ve got a framework to know if you are going in the right direction, and the guardrails to correct if it’s not going well. This is where the change gets to be implemented. Be as brave or incremental as needed for your problem and constraints, but be ready to measure and correct as you go.
Before making each change, record your hypothesis. “By doing this, I believe that we will move X to Y”. Take the actions, measure the impact and review against the hypothesis. If it’s going well, then keep it up! If not, don’t be afraid to cut the initiative and return to the status quo to try again.
Put in the effort to bring clarity to your proposed change, add the effort to measure as you go, and you are much more likely to find that improvement you seek.