Case Study: How to Improve the Performance of An Already Brilliant Team

by Pete Carapetyan


An outsider such as myself can come into a mediocre or poorly performing team and offer many enhancements - new techniques, technolgies, practices, approaches to get allow a team to move at a higher rate of delivery.

But how can I help a team that's already world class ?

In 2008 I got the chance to answer that question first hand. I spent almost the entire year as Product Manager for a team which had been performing top notch work for years, and genuinely did not need (me or anyone else) to show them how to do things better.

The answer arrived on my first month on the job - they were too smart, and too effective. So much so, that they did many tasks through diligence alone, that could be farmed out to automated routines to do for them. For competitive reasons, the techniques used might be better left off of a web page, but the net effect of months of work was a team that was freed up from a huge percentage of their daily workload - and a product that was even more consistent and well tested than when we started. That is saying something, because this is a well regarded product.

Automated Testing

The boldest move made was a total migration to automated UI testing for a product which spanned over 700 OSGi code modules, and almost a gig in size. Using a popular UI testing toolkit, this greatly reduced the build-release cycle by making the tests run 100% of the time all the way through for every release - something that was clearly not possible when all testing was manual.

Build Automation

A very large percentage of the team's capability was being charged to the budget category of manual builds. The process of automated builds was explored and moved forward, to the point of mitigating perhaps half of the manual tasks. As I finished my contract, the remaining work of automating the build was still progressing to completion.


My year with this software firm proved to be the most challenging year of my career. I may never know just how effective I was or was not, but it proved to me that even the best teams can use a boost from outside perspectives, even if it's just temporary relief to the pressure of a rocketing success path.