International Process and Performance Institute

When it comes to getting work done, the old saw is to put more hands to the plow.  This makes perfect sense in the classic sense of work.  If you’ve ever moved from one home to another, you understand that, generally speaking, more people make it happen much faster and better.  However “knowledge work”, as it has become known, seems to operate under a different set of wisdom.  In knowledge work, more hands can actually mean less gets done.

The late Peter Drucker introduced us to the concept of the “knowledge worker” in the 60’s.  He made some amazingly accurate observations about that demographic group and how they get work done.  The concept of knowledge work is significantly different than other work; yet we have a tendency to approach all work with the same tactics, and as Drucker explains, this is often flawed.

In addition to the “more hands” approach, many organizations want maximum buy-in for projects and tend to include as many stakeholders as possible in hopes of better results.  Yet our experience is often quite the opposite; research has concluded smaller team sizes are more effective.  Much of the research has said project effectiveness takes an exponential drop at 20 or more people, and the most effective teams were teams of five or less.  Indeed, more hands do not make the project go faster or give us better results.  In fact, the projects tend to take proportionately longer and the results are often convoluted compromises which tend to expand the ineffectiveness.

So is there a magic number for knowledge work project teams? I think there is.  In my consulting and coaching of BPM related content, I have often been asked if there is a magic number and started to think about it.  Yes, I have come to a definite conclusion about this question.  It’s not in any way a clinical or scientific conclusion.  I’ve arrived at it purely by years of observation.  I believe there is a magic number and it is three.  Any more or less and both the team effectiveness and project results diminish.

Why not four or five?  My observation of project teams with four people is that, generally speaking, one person will tend to sit back and let the others do the bulk of the work.  In a sense, they actually seem to check in and out of the project and thereby create “drag” on the project.  They often need clarification, ask questions which have already been asked and answered, or challenge something which has already been challenged by the core group.  A fifth person follows suit with the fourth and drag becomes exponential from there.  If you reflect back on projects where large numbers of people were involved, there were really three people who contributed the majority or core of the work on the project.  Sometimes the comment is made that the project is dominated by those three people.  I now believe there is something much deeper and less Machiavellian at work here.

Why not two or one?  Well, obviously one person does not make a team.  Yet the efficiency of a team of two is incredible.  There is often little disagreement with two because of the normal social pressure to get along.  Two people do not seriously challenge each others ideas and the project suffers.  While I see teams of two crank out more work in less time than any other team size, the work product is consistently lower quality and more highly prone to mistakes or failure.

What about projects so large three people can’t possibly get it done in a reasonable or required time frame?  I firmly believe three people can get a huge amount of work done.  However, I concede there are projects so large more hands are needed.  I recommend breaking up these projects into logical chunks and assigning groups of three to attack the chunks. This seems to work nearly as well.

Indeed, for best results, keep your knowledge work project team sizes to three people and absolutely no more than five.