Hire the best — and let them go

When I had my software company, in a typical year half of our employees were software testers, for which we hired local college students or occasionally high school students. We mostly got students who grew up nearby, and were attending the local junior college (three miles away) or the CSU or UC a half hour down the road.

The expectation was that this was a temporary job while they were in school. We hired a few college seniors, and a few didn’t work out, but most stayed 3-4 years. We taught them everything they needed to know about both running routine tests and isolating bugs so they could be reproduced by the programmer when it came time to fix it. (I’d like to say I created these processes, but they were created by our former director of product development and QA manager — who both sadly died well before their time).

Most of our business was with HP. When we went to visit their local office for a testing meeting, I was surprised to see how they used a completely different approach. There, QA was a career that you could do after a four year degree and stay with for many years. Because the best college graduates wanted to be programmers, the testers were (for the most part) those interested in computers but not talented enough to be programmers. (Our QA manager had years of experience, but his lack of even an A.A. meant he’d never be hired by HP).

For pretty much all of our student employees, this was the best computer-related job they could get in our local area without a degree. We got brighter employees without prior experience who didn’t stay very long, but we got the benefit of their aptitude and good work habits while we had them. Later on, I realized how fortunate we’d been to have them, as they went on to work at HP, IBM, Microsoft and other tech companies. One has started two companies and has mentored other entrepreneurs. The one student programmer we hired is now a full professor of computer science at MIT.

I was reminded of this when reading a Wall Street Journal column today by Sydney Finkelstein, a Dartmouth leadership professor. I know him from his great book on leadership that (in part) chronicled the leadership failures that brought Motorola crashing to earth in his book Why Smart Executives Fail.

The column has a provocative title and opening:

Why the Best Leaders Want Their Superstar Employees to Leave
By Sydney Finkelstein

Should bosses try to hold on to their star performers?

For most of corporate America, the question might seem like nonsense. Star performers are seen as so valuable that managers should pull out all the stops to keep them—or else see their companies take a big hit in productivity.

Yet some of the best managers not only allow their top performers to leave, but actively encourage it.

He talks about his study of leading execs across 18 industries for his new book, Superbosses: How Exceptional Leaders Master the Flow of Talent. While I might quibble with a point here or there (notably the idea of Larry Ellison as one of “the world’s greatest bosses”), the data are compelling. For example, a sidebar to the article says that at one point, 20 of the 32 NFL coaches either trained with Bill Walsh or someone trained by him.

This is the part of the article that really resonated with my experience:

The stories of these bosses reveal a crucial shared belief: You’re better off having the best people for a short time than average people forever.

The bosses were uncompromising when it came to recruiting. They didn’t want average; they wanted mind-blowing. They searched high and low for unusually talented individuals, often experimenting with nontraditional hires (and tolerating higher levels of churn when some of these hires didn’t work out).

But once unusually talented people were inside the organizations, the bosses accepted that some would leave. Top employees, the bosses realized, were almost always on a rapid growth trajectory. They were ultra-ambitious, perpetually angling for the next big opportunity, and it stood to reason that at least some of them would eventually need to leave the company to keep their careers moving ahead.

Even at our little software company, we were fortunate to have some truly exceptional teenagers and 20-somethings just starting their careers. This lesson is a reminder that — even if I’m no superbosss — for my next startup we should keep a policy of hiring the most talented people we can find, even if we have to train them from scratch or can’t hope to keep them for more than a few years.