Are we managing projects or are we conducting experiments? Managing you say? Are you really sure?

My teenage daughter complained that breakfast was not getting her through the school morning. She said that when Mom made her smoothies (frozen bananas, peanut butter, honey, milk, all blended together) she had energy and alertness until lunch. It was obvious to me, let’s fix her a smoothie for breakfast. Problem solved.

Not so my wife. She started to think, which is always scary (she graduated magna cum laude, inducted into Phi Beta Kappa, in computer science, hired by IBM). She started thinking up something that would have proteins and carbohydrates …. I intervened and strongly suggested, “experiment at another time, let’s make her a smoothie.”

My wife didn’t see it as an experiment. Instead, she was logically working out the perfect solution. Concoct a breakfast, based upon her vast knowledge from the Food Channel and Good Morning America on how to create the perfect breakfast (did I mention she was one A shy of a perfect GPA in college and recruited by IBM?). Instead of using something that worked, based upon past performance and what our daughter asked for, she was going to do something bigger and better.

Ugh. How many times have I seen this in a business and project management environment? A problem comes up, and there are lots of potential best practice solutions, but someone — too often senior management — decides we need a new super duper sexy edgy sounding solution. Off we go to implement that solution and if it doesn’t help, well then people just didn’t do it right.

Don’t get me wrong, there is nothing wrong with trying out something we have not done before. In fact I encourage it. Yet we need to be aware of the distinct difference between what is an experiment and what is an established best practice.

We had had years of poor project performance. Sometimes we would produce great products, that sold well in the market place, but even those were delivered late, were over cost and had less quality than we planned. Senior management decided to take action to fix this situation. Clearly product and project management was poor (theory alert!) and so they decided to implement a project management advisory board (experiment alert!).

This advisory board, consisting of senior managers who were known to be technically inclined, conducted regular reviews of all ongoing projects. Over time, we project managers learned that the best way to handle these reviews was to give the board whatever it wanted (as to slides, documentation, etc.) and to implement whatever they asked (regardless of cost or impact, as our projects were struggling anyway so they couldn’t make it much worse). Our on-time performance and quality did not improve but the advisory team became a permanent fixture.

Instead, in this case, what eventually worked to get us on time and with good quality was finding a way to finally estimate a realistic schedule and then letting people do the job they knew how to do. Often the key role of the project manager was to act as a protective buffer for the project teams from such things as advisory boards, QA initiatives, Black Belts, CMMI teams, etc. Any of these other initiatives could have been great, if implemented in a smart and integrated manner.

In the case of my teenage daughter, the success criteria was clear. If she got through the morning still feeling alert and energetic, it was a winner. For the advisory board, there was no such criteria. There was only activity and compliance. There was nothing wrong with this experiment, per se, only that there was no clear success criteria. Instead, it created a significant group of people incentivized primarily to keep the board going (money, control, prestige).

We should know the difference between when we are managing using tried and true techniques and when we are conducting a management experiment. Both are essential in managing projects, but confusing the two can result in some pretty bad results and even project failures.

Do you have any management experiments, not seen as experiments, going on in your project?