Will one metric make communication easier?

Will a single metric make communication easier?

Measuring software development (inclusive of development, enhancement and support activities) generally requires a pallet of specific measures. Measures typically include productivity, time-to-market, quality, customer satisfaction and budget (the list can go on and on). Making sense of the measures that might be predictive (forecast the future) or reflective (tell us about the past) and may sent seemly conflicting or contradictory messages is difficult. Proponents of a single metrics suggest simplifying the process by developing or adopting a single metric that they believe embodies the performance towards the organizations goals and predicts whether that performance will continue. Can adopting a single metric as a core principle in a metrics program enhance communication and therefore the value of a measurement program?

The primary goal of any metrics program in IT, whether stated or not, is to generate and communicate information.  A metrics program acts as a platform to connect metrics users and data providers. This process of connection is done by collecting, analyzing and communicating information to all of the stakeholders. The IT environment in general and the software development environment specifically is complex. That complexity is often translated in to a wide variety of measures and metrics that are difficult to understand and consume unless you spend your career analyzing the data. Unless you are working for a think tank that level of analysis is generally out of reach which is why managers and measurement professionals have and continue to seek a single number to use to communicate progress and predict the future of their departments.

Development of a single metric that can be easily explained holds great promise as a means for simplifying communication.  A single metric will simplify communication needs if (and it is a big if), a metric can be developed that is easily explainable and is it as useful in predicting performance as most metrics are in reflecting performance.  While there are many elements of good communication such as a simple message, ensuring the communication has few moving parts and is relevant to the receiver are critical.  A simple metric by definition has few moving parts.  The starting point for developing a single metric are the design requirements of simplicity and relevance which can be controlled and tuned (hopefully) by the measurement group as business needs change.

Developing a single metric is a tall order for a metric program, which is why most approaches to this problem use indexes (such as Larry Putnam’s PI). Indices are generally more difficult (albeit there are exceptions, such as the Dow Jones Industrial Average) to understand for wider audiences or fall into the overly academic trap requiring a trained a cadre to generate and interpret them. Regardless of what has been pursued, a single metric done correctly would foster communication and communication is instrumental for generating value and success from a measurement program.

Advertisements