January 11, 2012
Why Meetings are Inefficient
Meetings are one of those things that people just "know" to be inefficient. In many meetings within corporate environments, you'll see participants working their laptops or smart-phones when they are not actively participating in the meetings. If asked why, they tend to say that they are attempting to be more efficient by using downtime in a meeting to get other work accomplished.
We know we're not alone in this, but we've also noticed that feeling (rather than being) more efficient is the actual goal, as entertaining themselves surfing the web seems to also be considered more efficient than fully participating in the meeting.
From the perspective of MMD™, meetings are primarily inefficient due to conversion efficiency. In fact there are two types of conversion efficiency embedded within meetings, one structural and not amenable to change, the second is process related and definitely subject to modification. A tangential efficiency effect is also seen due to information durability, but that is often something desirable for reasons to be explained below.
Meetings operate very much like a token ring computer network. For those who can't remember that far back, token ring networks had a communication bus which was connected to all of the computers in a ring topology network. All computers in the ring could "listen" at the same time, but only one could "talk". In order to talk, that computer would have the "token" (permission to talk) for a specified time before passing it on to the next computer on the ring. Since all computers were using the same bus for both talking and listening, having a defined way to regulate who talks enables the highest possible throughput for a half-duplex network. (Full duplex networks allow talking and listening at the same time.)
Meetings are similar. In a well run meeting, people take turns talking, and everyone else listens. The shared sound environment means that multiple simultaneous talkers would create confusion, as listeners would be unable to process what is being heard and therefore results in even less efficiency.
The shared sound environment and the need for all of the participants, save one, to listen creates the structural conversion efficiency issue. Human brains are capable of listening at much higher word rates than they are capable of talking. That difference means that meeting efficiency metrics will always asymptote to something approaching the unused capacity of a listener as the number of listeners increases.
Figure 1 illustrates the commonly understood averages for typical information transfer activities. Meetings involve participants who are speaking aloud as well as those who are listening. Reading is something that meeting participants should not be doing within the meeting, however we will later want to discuss when reading can be a more efficient method of transferring information than having a meeting.
The averages within Figure 1 are based on aggregate data. Obviously, the capabilities of meeting participants will vary significantly, some people will be able to speak and understand faster than others but the overall relationship between the two activities will not be changed significantly by variability in the rates.
One note, we also assume in this analysis that speakers and listeners are optimally language capable, i.e. their language skills are sufficient to fully derive the desired semantic content. Unfortunately, in most meeting environments there is typically a wide range of language skills so that assumption is wildly optimistic. Meetings lose efficiency due to this factor, but the range of participant capabilities is not necessary to address the core conclusions being demonstrated, and is substantially more difficult to make generalizations about which would be applicable to all meetings.
Meeting participants can listen at much higher speeds than speakers can talk. The unused capacity represents inefficiency. In the example shown, listeners are capable of integrating an additional 350 words per minute, rendering their listening efficiency to 30%.
There is another issue to be addressed prior to moving on, and that is a 100% capacity offset adjustment. In almost every instance where the concept of a theoretical capacity maximum is discussed there is the understanding that capacity maximums are not sustained, but represent burst capacity. In an industrial setting, a factory loaded to approximately 80% capacity is considered fully loaded. The logic is that the 80% value represents the load that can be processed on a steady state basis. Bursts above this level are possible but cannot be sustained indefinitely without breakdown of plant and personnel. The same thing is true for other capacity related endeavors, although the percentage applied is strictly arbitrary. We are content here to simply acknowledge the effect, which we will label "Capacity Load Factor", and leave the actual amount measured to some future study.
Incorporating that factor into Formula 1 yields a revised version shown in Formula 2 below.
Given a capacity load factor of 80%, we would thus see a baseline efficiency of 80% for the speaker and 24% (30% * 80%) for each specific listener.
A meeting efficiency formula, which assumes participants are only speaking or listening, is captured within Formula 3.
Figure 2 below depicts the resulting relationship in terms of overall efficiency by the number of meeting participants. As noted earlier, the more participants within a meeting, meeting efficiency approaches the single listener efficiency.
Without questioning the need to actually have a business meeting, several conclusions are immediately apparent. First, that the most efficient meetings are those with two or three participants. Second, that larger meetings (those with more than 8 participants) are little different from those with 30 from a gross efficiency standpoint. Meeting efficiency in this context is just another form of conversion efficiency. The business purpose of a meeting is accomplished by taking aggregate information from speaker(s) and converting it into some specific objective which requires the group to achieve. The economic inputs are time-related costs incurred by being in the meeting instead of other work activities. Due to the structural process of the meetings themselves, aggregate meeting efficiency provides a coefficient which can be used to calculate economic wasted during the meeting process.
For example, if a 10 person group meets for an hour, the meeting efficiency is 30%. If the blended hourly rate is $40 per hour, the total cost of the meeting is $400 ($40*10 persons), and of that total $120 is expended actually producing the objective results intended for the meeting. $280 is spent in wasted time. Some managers view the $400 figure as the metric that matters, since the decision to hold the meeting assumes that the meeting objective is worth the economic input. There is some merit to that simplification, however, it doesn't account for what the meeting participants might be doing instead which could potentially create greater economic value for the business. As mentioned earlier, the meeting process itself has a structural component which can't be changed (token-ring like speaking/listening), it is just part of the cost incurred when having a meeting. Potential mitigation can be deciding who absolutely needs to be in meetings in order to accomplish the goal, and keeping meetings short.
What about the process related conversion efficiency mentioned earlier which could be changed? An example would include participants who are not concentrating on meeting business while attending the meeting. These would be participants who read emails, text message, read unrelated materials, do other work and so forth. Again, the participants engaged in such activities think they are being efficient however the meeting itself becomes less efficient due to this activity. If the participant's input is needed within the meeting, they must stop what they are doing, catch up with the group need, address the issue, drop the group and go back to whatever they were doing. As we have stated in prior publications, humans do not multitask, they task-switch.
The inefficiencies induced by such participants include having speakers repeat themselves in order to bring the task switcher up to speed (wasting everyone else's time in its entirety) as well as the task switching overhead penalty otherwise known as task setup. If meetings are well run, participants should not be allowed to engage in any other task while in the meeting. That results in the most efficient use of meeting resources possible.
Why have meetings at all?
We maintain that there are only two reasons to hold meetings in a well-run enterprise. If a goal does not fall into one of these categories, alternate means should be found which are more business effective.
First, meetings make sense when they are the most economical means to accomplish a goal. Just because meetings are inherently inefficient doesn't mean that other methods can't be even less efficient. In cases such as that, holding a meeting is a rational approach.
Second, meetings can be an effective means to reinforce business hierarchical structures. Team-building exercises fall into this category, as well as staff meetings and so forth. Again, the key is to run an effective meeting using as little resource time as possible so that the balance can be used for other productive activities.
One final note on this topic, meetings can also be effective when there is little need for information durability. Verbal transmission of information has a very short retention half life, usually around 24 hours, and sometimes that can be very useful. Verbal transmission to a large group (broadcast model) also has a low cost in terms of the speaker's time, and company-wide question and answer events can often address large issues at one time.
On the other hand, when participants are expected to remember details concerning the issues it often makes more sense to produce written documentation which has a much higher information durability.
Copyright © 2012 Arbitor Associates, LLC. All Rights Reserved. Last modified: 1/10/2012