A simple guide to incremental improvement toward any objective

Neil Killick
3 min readMar 25, 2020
I started from not being able to run down the road, to running a marathon, in less than a year. I did this by addressing the underlying reasons why I wasn’t doing enough exercise (including eating too much junk food!), then taking small steps toward achievable goals (e.g. being able to complete 4k without stopping, then 5k, then 6k, then 8k, etc.), all the while with a true north of getting fit.

Here is a simple guide (fewer than 10 steps) to making incremental improvements using a scientific approach, whether it be in your personal life or in your workplace.

Through the steps (below), we’ll follow two different examples — one related to an agile transformation, and one related to personal health.

1. Define an objective

  • Increase speed to market of product features
  • Get fit

2. Define possible root causes* for why you or your team are not where you want to be right now with regard to the objective

  • Team dependencies, different release schedules, work decomposed as batches ==> Slow speed to market
  • Eating junk food, not exercising enough, not sleeping enough ==> Being unfit

* Note that in complex adaptive systems, such as human bodies and product/service teams and organisations, there is rarely a single root cause for any given outcome (good or bad), hence the need to consider multiple possibilities which should form the basis of incremental and iterative improvement experiments — fishbone and causal loop diagrams are both useful for this purpose

3. Define a problem hypothesis based on one of the possible root causes (this is now the “problem to solve”)

  • I believe that [unsynchronised release schedules and work dependencies between teams working on the Acme initiative]
    are causing [longer than desirable lead times (speed to market) for releasing even small changes]
  • I believe [eating kettle chips every day]
    is causing [me to be overweight]

4. Gather the baseline data for the problem

  • 2 teams required to make a label change and release it to production — Team 1 and Team 2 release dependent components every 4 weeks, but 2 weeks apart ==> Minimum lead time = 2.5 weeks (including integration/testing)
  • Eat ~100g of kettle chips every day — Weight = 90kg — Height = 188cm — BMI = 25.46 (should be max 24.9)

5. Define an improvement hypothesis — an [activity] and [expected result] which you believe to be a small step toward the objective in a measurable way

  • I believe by [Team 1 and Team 2 releasing on the same day every 4 weeks]
    we can [reduce the minimum lead time for a label change release from 2.5 weeks to 3 days by April 20th]
  • I believe by [eating no more than 50g per day of kettle chips, and not adding any other junk food to my diet]
    I can [reduce my weight to 88kg in by April 10th]

6. Do the improvement activity

7. Compare the actual results to the expected

  • New minimum lead time = ?
  • Ate ? grams of kettle chips — Any other junk food? — New weight (and BMI) = ?

8. Consider what you learned by doing the activity (it should guide your next steps)

9. GOTO step 5, unless

  • There is a more potentially impactful problem hypothesis to focus on — GOTO step 3, or
  • There is a a more impactful objective — GOTO step 1

If you have any questions on this approach, or need a hand with your lean/agile product development or agile transformation endeavours, please reach out directly to me or my company Hypothesis.



Neil Killick

Software/product coach and leader. Expert in agile product development and product management.