Presenters often extol the value of more communication. At the same time we hear complaints about “information overload” and “too many meetings.” I am among those who have advised managers to err on the side of overcommunicating, and in The Truth about Teambuilding I cover evidence that talking more improves team decision-making. However, I added the qualifier up to a point, without being sure where that point is. So I was excited to come across a couple of studies that provide evidence teams that talk either too little or too much are not strong performers, and some hints about where to draw the line.
Researchers surveyed “60 cross-functional teams from 25 corporate and government organizations” representing a variety of industries and projects.[1] Team members rated how often they communicated by various means like e-mail or face-to-face. Meanwhile, department heads rated team performance not only on schedule and cost, but on “goal achievement.” This meant “the degree to which the project was expected to be able to overcome all technical hurdles, meet its technical objectives, meet its business goals and provide the expected commercial value to the company,” the study said. These are closer to the kinds of goals customers really care about, so I will focus on that metric.
In line with the Agile Manifesto, face-to-face (F2F) communication correlated most strongly with performance. But the highest-performing teams did not speak to each other “often.” (That was the label for a 5 on the related questions; a 1 was “never.”) Specifically the average was a 3.81. This is significantly higher than the midpoint of 3, so my earlier comments about erring on the side of more talking held true. The main point, though, is that the line graph of the relationship between communication and performance was a hump: teams averaging around a 1, 2, or 5 on F2F communication were weaker performers. E-mail communication had lower correlations, and goal achievement peaked lower for that, around the mid-point. Phone use was not related to goal achievement.
“The downturn in team performance when more frequent face-to-face communications took place may indicate confusion or conflict within the team about project goals,” the study says. “This confusion or conflict could provoke more face-to-face discussions and could hinder project performance.” On the other hand, “low-communication frequencies may not supply enough information to team members and may not facilitate the innovative combination of information and expertise required for high performance.” Remember that correlations don’t prove causation, so we don’t know whether the poor communication patterns cause or merely reflect poor performance. In practical terms I’m not sure that matters. Either way, you should try to find the right balance.
Another team used the same dataset for a series of computer simulations that modified the complexity of projects along two dimensions[2]:
- Multiplicity, meaning “multiple approaches to complete the task are feasible,” (and) “multiple end states exist that the task must satisfy.”
- Ambiguity, meaning “conflicts among approaches and uses that require making tradeoffs must be addressed,” and “decisions regarding the approaches to be employed and the end states to be satisfied must be handled.”
The communication needs appeared to change based on the combination of these factors. The curvilinear nature still held for goal achievement, but the ideal point for each form of communication shifted as multiplicity and ambiguity did.
Most fascinating to me was that two of the resulting curves were U-shaped, the opposite of a hump: for projects that had low ambiguity, middling levels of phone conversations were related to lower performance than either low or high levels of calls! The scientists speculated that where ambiguity is low, more successful teams don’t need much information, but if they do, moderate amounts merely introduce more confusion without enough clarification. Hence in that case the need for higher communication. Though computer models have often proven accurate, I need to point out that these apparently haven’t been tested with real teams yet.
Too bad the questionnaire didn’t ask for quantified instead of subjective measurements. You, dear reader, might consider two hours of face time each week a lot, while I would argue that is too low for most teams. Actual numbers would have provided us better guidance. Then again, we humans are bad at estimating figures accurately.
“Too many meetings” is one of the complaints lodged against Agile. But there is good evidence people don’t mind meetings when they are run efficiently and accomplish results. A posting in an online-class discussion gave a perfect example. A student described a manager in his company so determined to be loved that she would not push people to change their behaviors. Thus her team meetings repeated the same issues week after week. (I recommended he suggest the use of action items.)
Someone coming from that low-performing environment might gripe about the additional face-time in Agile, but in my experience they come around when they see things actually changing. Indeed, often when this complaint is voiced in a group, someone else will point out the number of ad hoc and informal meetings the pre-Agile organization has to have to address problems. The lower number of formal meetings in a poorly run org can create the impression that new Scrum ceremonies are introducing more F2F time when they really aren’t.
On a side note, being physically collocated did not relate to performance in the original study, but being in a different city or country did, to the negative. This was consistent with earlier research I’ve seen suggesting time-zone similarity is more critical to teamwork than sitting in the same room. It’s also more evidence against “global teams,” which I put in quotation marks because I don’t believe global groups can ever be true teams. The 2003 dataset for these studies obviously can’t tell us to what degree video-conference and chat tools can capture the F2F experience. But I’ve blogged in the past about the considerable evidence they can’t make up for the time gap. Remember, the ideal F2F amount for goal achievement was nearly a 4 on a 5-point scale. Even with today’s video apps, I’ve never seen any virtual team come close to that level of interaction, much less a global one.
Regardless of the medium, the easiest answer to striking the balance between too much and too little F2F time is to ensure your team discussions have a goal and then achieve that goal as efficiently as possible. When I was doing teamwork consulting, I offered training on meeting facilitation, active listening, and persuasion skills. Together these helped ensure every perspective was heard and valued, which is critical for innovation and the best decision-making, while eliminating time-wasting behaviors. I believe a system like Scrum aids in this by prescribing specific goals for each meeting plus limits on what is discussed:
- Grooming—Focused exclusively on clarifying and prioritizing user stories.
- Planning—Deciding what stories to do in the next sprint and how to do them.
- Standup—Answering three questions (and only those questions: if you do a “16th minute” discussion, you are missing the point to the ceremony).
- Demonstration or Sprint Review—Showing what work was accomplished and getting customer feedback.
- Retrospective—Answer what went right, what went wrong, and what you’ll do differently.
In terms of saving time, the most important duty of the Scrum Master is to keep meetings on-task through good facilitation. If you’re an SM and haven’t been trained on facilitation, pick up a book or attend a seminar. The underlying principles have been well understood roughly 150 years, since the first version of Robert’s Rules of Order was published, so there’s no valid excuse for an inefficiently run Scrum ceremony.
As in many aspects of work and life, a balance between extremes proves the most effective route to effectiveness. Listen to the complaints of your teammates and the team’s stakeholders. Whether you hear lines like, “no one tells anybody anything,” or like, “we yak about this stuff too much,” pay attention. Implement a system for work planning and apply formal meeting facilitation to make the system as efficient as possible. Either the complaints should go away, or more likely, you should hear a balance of both of these complaints, at which point the majority of folks probably feel you’ve hit the right mark.
Please share this post at the bottom of the page.
[1] R.R. Patrashkova-Volzdoska et al., “Examining a Curvilinear Relationship between Communication Frequency and Team Performance in Cross-Functional Project Teams,” IEEE Transactions on Engineering Management 50, no. 3 (August 2003): 262–69, https://doi.org/10.1109/TEM.2003.817298.
[2] Deanna M. Kennedy, Sara A. McComb, and Ralitza R. Vozdolska, “An Investigation of Project Complexity’s Influence on Team Communication Using Monte Carlo Simulation,” Journal of Engineering and Technology Management 28, no. 3 (July 1, 2011): 109–27, https://doi.org/10.1016/j.jengtecman.2011.03.001.