Pitching a new technology

Making the case for a big change needs to be pitched as the right decision for the organization. I shared one example from my personal experience in an interview with the My Open Source Experience podcast. The interview actually happened a while ago, but the podcast recently highlighted part of this interview on YouTube, which I've shared below.

I shared several examples of using Linux in the back-office to support business processes, including NIS and NFS. Later, I joined the Web Team at a large university.

The university was working on a major systems migration, which required re-integrating a number of add-on systems such as registration. While most processes worked smoothly, the team struggled to support web registration on the web—despite using a "full IBM stack" including hardware (IBM SP2) and software (IBM AIX, IBM Java, and IBM WebSphere). Every registration cycled failed dramatically on the IBM systems, until we had the idea to migrate web registration system from "Big Unix" to Linux systems.

This was made possible because IBM decided to support Linux as a first-party platform for both IBM Java and IBM WebSphere. After running exhaustive tests, we found that one Linux system could support what several IBM SP2 systems could not. We successfully pitched Linux to our leadership, and solved the problem.

A big part of our success in pitching the new technology was that we didn't approach it as "Linux is cool" but instead "Linux performed better in these tests, it solves the problem. And it's cheaper, and the full stack (even the hardware it ran on) is supported by the vendor—no finger-pointing if something were to happen."

Another reason the idea was well supported by leadership was we documented our idea in a white paper that laid out things in a clear way, including costs, risks, and risk mitigation. And we wrote it in plain language, oriented for the senior leadership. That attention to audience helped make a positive impact.