EA Measurement - How not to fail

Measurement is critical to any business initiative in today’s business climate but the mountains of data collected by various business applications can also be detrimental. Agonizing over metrics can reduce speed to market, by creating a culture that needs to “fully” analyze data before decision making. Gartner’s “Why EA Measurement Programs Fail provides a simple high level viewpoint of what to avoid when constructing measurement programs. The validity of the article became apparent when looking back at a past business interaction from several years ago.

Management wanted to institute a new time management program and track time everyone in our division was spending on various activities. Sounds all fine and good, right? WRONG. Management decided to have everyone track their time in 6 minute increments (one tenth of an hour) because that was what other companies were doing and justified it by stating it was  industry standard. It was a way to reward our productivity and gauge the IT tools provided to us. That’s not how it was perceived by the majority of the employees and this effected the results of the measurements.

One of the core tenants of the company had always been having a client focus and to the employee base this new time tracking standard was seen as micromanagement. It took away time from clients and put an undue burden on the employee to manage this active metric (most metrics gathered by the company were passive). After a few weeks of adoption and correction to standardize the compliance department who was running the time tracking measurement program touted their progress and how the program was giving them insight into how time was spent for projects vs.  support vs. administration.

However, there was an elephant in the room they could not see. There were no controls around how employees were entering their time. After being questioned on “why an entry was a certain way” and/or “what one was doing” employees began constructing their time cards in a way that corresponded with what the compliance department wanted to see. There were few controls around this active metric so compliance thought they were seeing improvement when in fact all they were doing was scolding employees into cooking their timecards. All of the employees were on salary so it had no effect on payroll dollars, but invalidated all of the measurement that was being done by compliance.

The measurement program was not well explained, aligned with the company culture so it was not trusted by the participants. Needless to say it failed after a couple of trying months. Measurement programs must be well articulated to those being measured in order to build trust in them and around them. If the participants can see the program is designed to furnish them with better tools and the right resources for success – the measurement will have value and garner support. 

Works Cited

Weiss, D. (2006). Why Enterprise Architecture Measurement Programs. Stamford: Gartner .