Seven Keys to Effective Performance Measurement

by Harry Chapman

October 2009

One of my clients, a non-profit that is a leader in clinical research, asked me to expand the scope of my “On Demand CIO” services to take over the Bioinformatics Research development team and system. The predecessor left to pursue “other career opportunities.” At the time, I knew very little about genetics research and the underlying genetics research database. The system was a cluster of 12 servers running on IBM’s DB2 database. The system had a reputation for being unstable. The support staff spent virtually all of their time just keeping the system working and had little time left over to develop new capabilities the researchers were requesting.
Believing in the power of measurement, I had the technical team research some of the key performance parameters. We learned that the peak demand on any of the 12 CPU’s NEVER exceeded four percent of capacity. We learned that the clustered architecture was appropriate for high volume input/output environments such as a point of sale (POS) system used by large retailers with hundreds of locations and thousands of terminals constantly feeding transactions back to the data center. We learned that this architecture was less than ideal for a data-intensive, calculation-based environment. Before making any changes, we developed a set of measures of the current environment and used the measures as the basis for assessing the architecture and re-architecting the underlying system. The net result was to covert the system from its original clustered architecture to a de-clustered environment, eliminating 10 of the 12 servers, and moving from a system that needed to retrieve data from 12 servers to one that only needed to access only two servers. This change dramatically improved performance, dramatically improved reliability and enabled the development staff to move from focusing on keeping the system running to actually building some new functionality for the users. The key was using performance measures to diagnose the “root cause” problem. Knowing the root cause allowed us to build a business case for the time and effort needed to re-architect the system. Without the insights of the extremely low CPU utilization coupled with extremely low reliability, we would not have had the insights to greatly simplify and there by improve the design. Performance measurement was the key.
While the system is complex, the underlying lesson in the example applies across many areas beyond simply the technology. In order for performance measurement to work most effectively, use of performance measurement should be based on some key principles. Here are the principles that have worked for my firm, the Bay Area Consulting Group LLC (www.baconsulting.net) over the years:

1. Start with a disciplined management process. If you commit to a performance measurement process, then you need to “walk the talk” by imposing appropriate discipline to live the commitment. GE is an excellent example of a company that has implemented a Six Sigma culture to obtain a competitive advantage. They are either number one or two in their industries or they get out of the industry. Clear, disciplined and effective.
2. Create a supportive climate: If the measures reveal something that is counter to your intuition, accept what you are being told and deal with the “brutal facts.” Often the reaction to a counter-intuitive result is the belief that the measure is wrong and therefore we can ignore the results. Yes, indeed, the measure could be wrong, but be careful in discarding the results until you fully understand what the results are telling you. To do this, you need a culture that is supportive and able to view the results in a broader perspective – “What is this really telling us?” Enable the organization to “think outside the box.” Encourage people to take chances. Understand that the road to success may be littered with failures.
3. Transform intuition into a measurement-based process. Take your “beliefs” about cause and effect and transform those “beliefs” into a process. What is the process? What are the key measures for the process? After all, defining the process forces you to clearly articulate your intuition into a form that can be measured. By establishing the right measures, you will learn from the results and be able to improve your performance.
4. Challenge conventional wisdom; rethink the essence of the business in order to understand what really matters. Every industry has its beliefs and assumptions about what really matters. Often these preconceived notions, no matter how long they have been adopted or how deeply they are ingrained into the industry, are flat wrong. The truth often defies conventional wisdom. Uncovering these gems of insight can lead to important strategic break-throughs. An example is the utility that believed during power outages that their customers valued quick service. They set up a measurement process and learned that what the customer really valued was not quick service, but predictable service. The customers would rather have the utility tell them that power would be restored by noon tomorrow (and have it restored NO LATER THAN NOON) than provide no information and leave the customers guessing – ever if the power company were able to restore power sooner. They learned that satisfied customers care more about having the utility set the right expectations and then follow up the commitment with predictable delivery. The utility learned this by establishing measures, carefully observing the results and learning from their experience even when the results defied their long-held and deeply ingrained beliefs. 
5. Identify the “right” things to measure and identify the “right” measurements for those things: In a measurement culture, we often are lured into measuring the processes and activities that are the easiest to measure with the measures that are the easiest to collect. Wrong on both counts. Identifying the most meaningful aspects of your business is fundamental to a successful performance measurement culture. Along with identifying the most meaningful aspects of your business is the need to identify meaningful measures for those activities. Take, for example, the case of a Fortune 500 company that identified as one of its strategic goals to “increase the breadth of talent.” One of the measures was the “percentage of people with college degrees, MBA’s and registrations/licenses.” On the surface, these may seem to be good measures. They are discrete and easy to compile. But are they meaningful? These measures are only meaningful within the context of what the organization is trying to accomplish with the “breadth of talent.” Out of context and with no linkage to the overall objectives of the organization, these measures are meaningless as well as potentially misleading.
6. Analyze the results: Carefully analyze results in order to take appropriate action. Work is, after all, comprised of people, process and technology. If the results of the process are unsatisfactory you may have the wrong people performing the process or the people may not have the skills, training or motivation necessary to be successful. Are the people the reason the results are underwhelming? Alternatively, perhaps the process itself needs improvement. Consider either improving the process or replacing it altogether with a process capable of reaching the overall objective. If you have the right people and a good process and you are still unable to achieve satisfactory results, rethink the strategy and the process goals. Start over.
7. Trust the facts, don’t kill the messenger. How often have you seen the person who reports the facts which contradict conventional wisdom being blamed for the outcome? Our belief systems and conventional wisdom often run so deep that we find it difficult to believe what the actual facts are telling us. Accept the information and learn from it – build a culture that uses the results of the performance measurement culture to challenge preconceived notions and gain new insights into your business. Don’t kill the messenger!

There are a handful of organizations that have distinguished themselves as industry leaders by defying conventional wisdom, using performance measurement as the basis for achieving new levels of success. Take Dell computer, for example. One of their key measures is their “Cash Conversion Cycle” (CCC). The CCC is a measure of the time it takes to turn expended cash into cash in the door. It is determined by taking the days of inventory, adding the days of receivables and subtracting the days of payables. The industry norm typically ranges from 60 – 100 days. At Dell, the CCC is a negative 38 days. That’s right, MINUS 38 days, not PLUS 60 to PLUS 100 days for conventional businesses. Extraordinary! How do they accomplish this? It is a byproduct of their incredible business model resulting in Dell actually building working capital as their sales increase. They have defied conventional business belief by developing an awesome business model built on a performance measurement culture.
How about your organization? Is your organization using performance measurement effectively?