Conventional wisdom suggests that leadership is important in business, so naturally the various types of leaders in and around Agile teams are important in making the teams healthy and productive. Or are they? After all, Agile organizations are often far from conventional. At AgilityHealth, we are all about understanding Agile teams and what makes them healthy, which is what our SaaS product, AgilityHealth®, was created around. Since launching our platform, we have amassed thousands of Team Health surveys from tens of thousands of participants, which gives us a unique data set to really look deeply into what makes teams successful and healthy.
Having all the data at my fingertips, I was curious as to what unique insights we might be able to gather from it. To start with, I wanted to find out how the different roles and levels of leadership (formal or informal) affect the teams’ health.
In order to answer these questions, I used correlation analysis which tries to determine if variables are related. To pick an example from our day-to-day life, we could ask: How is driving time related to the distance we travel? There are many factors affecting speed such as traffic conditions, road hazards, route taken, etc., but we know from experience that there is a more or less linear relationship between time and distance: The more time we spend driving, the longer the distance we cover. So these two variables are strongly correlated. Instead of getting deep into statistical analysis, let’s look at this visually via a scatter plot, which could look something like this for our example:
It would be relatively easy to draw a trend line through these dots to show the linear relationship.
As a counterexample, we can probably all agree that the volume of our car radio has no direct correlation to driving time (except maybe for those of us who might be tempted to drown out passengers on long road trips). So drive time and radio volume are not correlated. A scatter plot with these variables might look like this:
As we can see, there’s no clear pattern of the dots on this chart and we would be hard-pressed to draw any kind of trend line.
So let’s apply this technique to some of the key roles in Agile teams and see what we can learn!
The Scrum Master (SM) is a team-level facilitator and coach and one might expect that this person is quite influential within the team. In other words, if we looked at the various competencies that pertain to the SM in our Team Health assessment, how would they relate to the overall team health? Does a capable SM tend to make the overall team healthier? Here’s what the data shows:
Apart from a number of outliers, we can see that the dots distribute closely and evenly around the regression line, i.e. as SM health improves (x axis), so does the overall team health (y axis). So it appears that the data supports our thinking about the strong influence of the SM.
Not surprisingly, the Product Owner (PO) is also a critical member on an Agile team as he is responsible for the backlog, prioritization, has a deep knowledge of both the product and the market, and needs to be able to collaborate with the Agile team effectively.
So what does the data show us?
The graph looks somewhat similar to the SMs. So it appears that the PO also has a strong influence on the team’s overall health. So far, so good!
The Tech Lead (TL) is not an official or formal role in Scrum, however many organizations designate a senior technical resource or architect on the team to lead and mentor the other developers. SAFe® has formalized this role, not at the team level itself, but at higher levels in its structure as System Architect/Engineering (release train level), Solution Architect/Engineering (solution train level), etc.
Does the data confirm a strong correlation?
Somewhat. While we see very competent TLs on high performing teams, we start to see more of a spread at lower levels: Some teams may be quite healthy despite a not-so-healthy TL. What does this mean? The data itself can’t tell us this, but here are some potential explanations:
- The overall team health is less dependent on having a very healthy and effective TL.
- The team may consist of highly skilled developers (and testers), so it makes less of a difference if the TL on the team is very good or not. The team knows what they’re doing.
This was interesting, wasn’t it? Look for the next part of this post in which we’ll look at the effect of the functional manager as well as whether or not various people’s confidence in the team is actually a reliable predictor of its health.
By Rene Rosendahl, AgilityHealth Product Strategist