Case Study I
Startup A was formed in 2000 by two computer science academics to provide solutions for data distribution on the internet. A small team was set up and they started to develop an innovative product for transparent web caching on the net. In 2001 the company recruited a new VP R&D, Howard, a few weeks short of the crash of the dot-com bubble. The changing economy pressed hard on the company to change and adapt, and they had to realign. They converted their core technology and switched to distributed storage in corporate WANs. However, their new VP R&D came with a fresh and clear vision of how a team works and adapts to changes.
Howard insisted that although the product runs in a complex environment of distributed nodes, every developer, tester and support person should have access to a production-like environment of his own. The company invested heavily in hardware to set up a laboratory where field-like simulations could be run. In addition, Howard put one of the technically strongest engineers to be in charge of an automatic build environment that would compile and test the product each night and store the produced artifact on an intranet server. The nightly build gave the development team timely feedback if they would break something on the one hand, and confidence in the quality and stability of their progress on the other.
A small additional team of programmers was recruited to develop automatic tests of the product. Within weeks, more and more of the test specification was being covered by automatic scripts and within half a year over 85% of the cases were fully automatic. This allowed each release cycle to be completed in a very short time while ensuring that the version underwent sufficient quality assurance to be given to the sales team for their demos in the field. In addition, the automated tests easily lent themselves to a combination of configurations to test under many different scenarios.
These were the post-bubble days and the company had to constantly test different waters in order to reach rewarding markets. The company finally found a promising niche in the data centralization market. The VCs behind the company reluctantly provided another round of funding. Customers showed interest but were wary of trusting their data to a small startup. The company soon entered the radar screens of one of the big players in the market who found that the company’s product could fit well into the solution stack for their distributed enterprise. The big company was very impressed both by the product and by the team and the technical prowess of its members. In a matter of months a deal was cut and the startup was acquired by the big company in one of the most significant exits of the early post-bubble years.
The company faced difficult business situations all through its history. But Howard’s vision allowed the technology developed to be an asset and not a liability, freeing the board to take bold decisions and confront changes with adaptability. The quality-driven approach enabled the engineering team to maneuver and the company to survive – and succeed.