The Regenerate Web

You are a manager or leader. You want to make a difference. You recognize that the way that the organization does things is getting in the way. There are processes in place that need to change if things are going to get better. Most people want things to get better. Virtually everyone in every organization has some process that they would like to improve. Trouble is that it is difficult to get two people (let alone the whole team, division, department) to agree on what or how to change.

How do you ensure that the changes that you implement are having the effect that you hoped that they would? How do you ensure that the changes stick, that the people experiencing the change are helping drive the change, rather than resisting it. How do you collect sponsorship from those managers in positions to help you drive the change? I have one suggestion: Measure the change!

Here are two strong arguments in favor of measurment:

Measurement signals importance!

Anything worth doing, is worth measuring. I have come to believe this: if you are changing something to achieve a particular value outcome, then you darned well ought to be measuring that outcome. Why? Because organizational change is hard. People get frustrated, have strong opinions, and generally resist change, unless they can see tangible evidence of the outcome. Measuring results shows people (your staff, your boss, key project stakeholders) that you are serious. As the initiator of change, measurement is also a way of making yourself accountable: When you measure the outcome, you are saying, "If the change doesn't work, it will be evident, and we can "undo" it". Measurement sends a signal to sponsors of change, and to those undergoing the change. It says this change is important, and it says that you are willing to be accountable for the results. Both messages are important to both constituencies.

Measurement fosters accountability!

People aren't stupid! They recognize that measuring things provides visibility into what they are doing, good, bad, or indifferent. I am not proposing implementing productivity metrics with a view to performance management. I am proposing implementing metrics to manage process change with a view to help your team improve some value outcome (i.e. productivity). In time, performance expectations will align with the team values, but when you initiate the change, you will probably need to reassure staff that you will not beat them about the head and ears with the metrics. This visibility that measurement provides gives you insight into team behaviors. When implementing the change, you should resist the temptation to penalize team member when their bad behavior patterns are illuminated, this should be viewed as an opportunity for you as a leader to coach the team member into better behavior patterns. Hint: you should be prepared by knowing what behavior patterns lead to the desired outcomes. In the best situations, you can get the rest of the team to help with this by encouraging valued practices, and holding their teammates accountable.

Here are two tips for how to make measurement work:

Metrics must be relevant and simple!

The team must understand what you are measuring, and be invested in the value outcome that you are trying to achieve. The way that the metrics are derived should flow obviously from the process as designed. Stated another way, metrics should not be forced or reverse engineered. You want your team to hold each other accountable, at minimum this requires a suspended disbelief on their part. They don't have to be excited about the change, but they should agree that the outcome that you desire is an improvement. They should also agree that the metric correlates well with the desired outcome. Mostly, your metric should be simple and intuitive, you should not need to spend much time "selling" the metric.

Measurement must start at high frequency!

Process changes measurement must be frequent to provide adequate feedback on behaviors. High frequency measurement means that effort expended in collecting the data must be small. Moreover, feedback lag must be much shorter than the measurement period. That is, if the measurement period is a week, the feedback should be given in a day. If the measurement period is a day, than the feedback should be immediate. When coaching athletes, training to perform a new technique, feedback is delivered on each and every attempt. As the new technique is developed, they can self-diagnose, because the athlete knows what it feels like to do it right. Once most of the team know how to do it right, the team can hold each other accountable, and the measurement - feedback loop can relax.