Should You Measure Agile Adoption Effectiveness?
A question I frequently get when helping organizations transition to agile is "What metrics should we use to measure our agile adoption?" I think this question is premature. People really should be asking, "Should we measure our agile adoption?"
Measurements themselves are not inherently bad and can be useful when used appropriately. So if you can create appropriate measures related to an agile transformation, yes, you should measure it. The trick is in figuring out what an appropriate measurement is.
Patrick Kua has an interesting perspective on what entails an appropriate use of metrics, and I find that I generally agree. I think metrics are helpful for a situation when they are outcome based and facilitate learning. Outcome-based metrics are those that are tied directly to the goals of the organization and therefore show how the change you are examining contributes to meeting those goals. Metrics support learning when they can be used in a way to revise your actions, intentions, and choices to better align with achieving your desired outcomes.
So if we take it as a given that you should measure agile transformation, the next question is whether you can measure your agile transformation. The most common approach to answering this question is to try to quantifiably determine how agile your team is, with approaches including creating a spider chart, a fairly simple dashboard that looks at the characteristics of the team, and this minimum viable measurement.
ScrumMaster Sean McHugh lists a couple of problems that are inherent with these sorts of measurement schemes and suggests measures that could be used in their place. I think the characteristics of the measures he suggests are helpful but disagree with the measures themselves because adding value to every story drives teams to deliver more functionality than may be needed.
Another point of view is that lean agile adoption can never be the goal and should not be thought of as a project. Therefore, you can't measure the success of your agile transformation directly, but proxy measures that look at outcomes can give you a sense of how the transformation may be impacting your organization overall.
If you think proxy metrics may be helpful, consider tracking a trend instead of concrete numbers in metrics such as the ratio of fixing work to feature work, cycle time, and number of defects escaping to production. If you are interested in finding out what actual teams have done and found to work, this Adobe experience report details how the company approached measuring the success of an agile adoption.
After considering these different perspectives on measuring agile adoption, my advice is to consider why you are adopting agile in the first place, find metrics to help you determine if you are reaching your goals, and use the measurements as a source of information to guide adjustments to your activities.
