We’ve taken a deep dive to analyse more than 1,300 digital teams and discovered some surprising results. While stability is easier to achieve in small teams, stability in larger teams could impact negatively on diversity and innovation, and sometimes lead to a toxic work environment.
Harvard University Team effectiveness guru Richard J. Hackman wrote that; “Real work teams in organizations have four features: a team task, clear boundaries, clearly specified authority to manage their own work processes, and membership stability over some reasonable period of time”. At the time of writing this, digital teams were in their infancy; but few would argue these same four features are just as relevant for off-line teams as they are for digital online teams … or would they?
What we have found from our recent ‘deep dive’ into more than 1,300 self-described digital teams is that digital teams are much larger than Hackman would normally consider a ‘team’. We would argue that for digital teams that meet Hackman’s rule of “less than double figures”, Hackman’s four factors should equally apply to both digital and non-digital teams.
But what about those self-declared teams that are considerably larger than 10 members? Does ‘Digital’ invalidate the rule of an ideal team size being less than 10 through the increased functionality teaming platforms can afford?
We took the last factor mentioned – “stability over time” – to test this proposition.
Measuring Team Stability
What decades of research tells us is that team productivity falls off once team sizes get into double figures. Team productivity research reported by Ivan Steiner in 1972 was more precise, indicating peak team productivity occurred in teams sized between 4 and 6 members, peaking at 5 members:
We therefore decided to measure relative stability based on the top 5 most active team members. The degree to which the top 5 varied on a week-to-week basis over a 12-week period was used as our measure of team stability. If the top 5 remained the same each week over the 12-week period the team was assigned a 100% stability score. If the top 5 varied by just one member from one week to the next we assigned an 80% stability for that particular week. The overall %stability was the average of the week-to-week stability scores over the full 12-week period.
We applied our stability formula to 316 teams for 4 different organisations to find the following distribution of stability scores:
What we found was a full scope of %stability scores, from 0 to 100%, with 79 teams or 25% of our sample achieving 100% stability.
We would therefore anticipate that these teams have at least one of the four factors that Hackman believes are required to form a high performing team.
How does %Stability Relate to Group Size and Group Type
While we did not have access to information that might validate these findings precisely, what we were able to do is to identify how group %stability relates to group size and the group types we identified in our prior Teams research:
- Single Leader
We can see that, as anticipated, there is a relationship between ‘Team’ size and %Stability, with the smaller the team, the greater the potential for stability. While the relationship is statistically significant, the above plot does indicate there are some sizeable groups that do exhibit reasonably high %Stability.
To further explore this relationship, we calculated the average %Stability for each of our inferred group types:
As we might anticipate, the disconnected teams/groups have the lowest average stability and the community and single leader have the highest levels of stability.
All group/team types had instances of 100% stability. A good community of practice will have a strong and stable core of members. Single leader teams may have a small number of trusted lieutenants, supporting the leader. A high performing self-directed team will have all members consistently active. A forum may be like a single leader team, having a ‘leadership’ sub-group sustaining the forum. What the averages hide however, is that there were a significant number of disconnected groups that also have the maximum stability. So what is happening here?
Taking a look at some of these groups we show the following example of a stable but disconnected group:
The most connected members are shown as the 5 largest nodes, that over time are consistently the same top 5 on a week-to-week basis. However, we can see that each of these nodes appear to be engaging with separate disconnected members.
In fact, what we are seeing here is the formation of separate and stable cliques within a single group/team.
Highly stable disconnected teams/groups should send up a warning flag. Like a political party infected by competing cliques and factions, group/team effectiveness will be hugely impacted unless the cliques/factions can be brought together.
The impact of Team stability on performance is often taken as a given. In this article our analysis reinforces our expectation that stability is harder to achieve, the larger the group/team becomes.
Beyond this anticipated finding, however, we have discovered some nuances. While in general stability is good for team performances, we need to be aware that for larger groups, a stable core can also lock out diversity and innovation.
On a more sinister note, disconnected teams that demonstrate a stable core could in fact be demonstrating a stable set of competing cliques or factions, something that can result in a toxic working environment and low productivity.
Watch for our Team Stability Widget when we launch SWOOP for MS Teams in July 2019.