Managing Change by Measuring Behavior

Getting organizational change to make an impact is very difficult under the perfect situation, but getting an impact quickly and to achieve the desired results in most situations requires the perfect storm. Rapid and effective change is more important now than it has every been given the current economic contraction and chaos over the last 18 months.

Change Model showing the Starting Point (now) and the Desired State

The groundwork, or Step 0, must be laid first.  This is what we call the environment of change – where we have created the “luxury of a crisis” (now) the desired state (vision) and a path to get there (change management path).

Step One – Identify the Lever

By performing an analysis of the gaps that prevent your company from achieving its goals, you will begin to focus the change process.  There is a Chinese expression that “A problem well stated is half solved” and this provides a guiding principle for successful change initiatives.  A project history approach is one of the best analytic tools for determining how to unlock an organization’s potential to achieve its goals.  By looking at the last set of projects/programs, one can examine the barriers that prevented the achievement of the objectives.  By doing a set of project histories or by ranking the impact and frequency by a cross functional management team, the organization can identify the most important problem to solve.

Project History combined with Root Cause Analysis

Once the top set of problems are created, the cross functional management team should investigate the best practices (tools, workflows, or organizational) to address the barriers.  With the top (i.e. single) barrier identified, and the corresponding solution identified, the team should begin to frame a “Predictive Metric” around this best practice.  It is important to recognize that one can only take on so much, so it is better to start with a focus on one initiative and make significant change in one area before tackling more programs.

Deriving Predictive Metrics from Goals via levers

Step Two – Identify a Predictive Metric

Once the biggest barrier is identified, and we have the solution in hand we need to derive a predictive metric that would place a lens on behavior that would be modeling the new way of doing things.  The metric should describe a behavior/deliverable/process step that happens with relatively high frequency so that you could see a change in one week’s time and significant change in 4 weeks time.  The metric, does not have to be a measure of the end of the process, in fact it should not be.  What it should be however, that it be causal to the end behavior that you are seeking.  It can be a proxy for the final benefit and often times will appear simplistic at first glance.  The metric should also have some other qualities:  It should be easy to acquire (by an administrator or equivalent), not require an IT intervention (keep to spreadsheets), and be simple (no difficult calculations or judgment).

List of Metric Components

Step Three - Set Targets using ½ Life Principles

How fast does one expect to change?  It depends on many factors including urgency, simplicity, number of dependencies (people, process, or technology) and the organizational scope.  In the extreme, to avoid an oncoming truck coming at you on the wrong side of the road at 80 MPH, it takes seconds.  However, changing the supply base of a large multinational organization can take years.  However, if there is not a well developed plan, how can you estimate the rate of change in a large organization?

Art Schneiderman, VP of Quality at Analog Devices wondered exactly that.  When in trying change initiatives at his company, he performed a survey of successful change initiatives and developed a theory based on the how long the change took based on external factors.  (Reference:  Schneiderman, Arthur, "Setting Quality Targets," Quality Progress, April, 1988).

Table showing time to achieve 50% improvement in behavior

The greater the technical complexity and the organizational complexity determines the rate of change he found.  Furthermore he found by estimating these two factors, one can predict the “Half Life” period of a change initiative – the time it takes to get behavioral change ½ the way to the goal.  We have applied this in our work and found it to be amazing accurate.

Graph showing theoretical improvement over time with a half life of approximately 2 months

Step Four - Monitor by Governing Body

By creating a dashboard showing the actual and the targeted behavioral change, with a read out in real time, management can see the deviation very early in the change initiative.  This is critical.  Also critical is that there is a governing body that can take action to close the gap.

If the “Changers” and the “Changees” are all in one organizational unit, this monitoring function is simple since the head of the unit can provide the motive force to close the gap.  More commonly, the change spans several units, organizations, functions, or geographies.  In that case, you need a governing body that can have binding authority across the unit.

Step Five - Course Correct

Armed with a clear metric, an a clear remediation, one can achieve rapid change.  The system is closed loop by acting on the difference between actual and desired behavior.


By using simple tools, with sufficient resources rapid change can occur on a predictable course.  Just like real life, doing fewer things, better, will yield greater output with higher speed.