Oversight for your Oversight

Last night at Lean Software Austin, I posed the question "Do managers have to be good programmers?" and the usual soft affirmations, serving more to make a position obtainable than to prop one up*, were absent. Instead I was met with a resounding "Yes" and I got the impression that I was the only individual in the room would have made the "mistake" of wondering as much aloud. Scott Bellware asserted most managers are failed programmers and are disconnected. He spoke of an aristocracy in which the upper tier of an organization pads its pockets in lieu of building a slush fund of cash padding for research and rainy days. This leads to playing the short game (short term) instead of the long game. How does one keep this from happening? What will make a manager a better manager when you can't expect his manager to do so in the aristocracy model? Some insights, serving as a light in the dark in the name of preventing bad management:

• Program or be programmed. Everyone has to be a programmer. So what we are really talking about is how to hire in general. The programmers become/are the managers.
• Tristan Slominski suggested there should be no hiring bonus and no artificial incentives in hiring. People should join a team for the RIGHT reasons. This drew the most applause of the evening. It is alright to start a team with the good people you know, but when you ask good people if they know good people in hopes of baiting other good people, you are sliding into bad.
• Marissa Mayer of Yahoo apparently has a practice of screening everyone herself. I heard about this for the first time last night. This could be a good thing if the person on top is good. Another example of when this doesn't work was given too, an example of a Vice President inheriting a team full of bad hires and then setting ridiculous hiring practices that only PhDs navigate well as an overreaction. The hiring hero has to be wise.
• Greg Symons of DrillingInfo suggested that one does not need to play the short term game of trying to find the right people quickly. He simply suggested that if you cannot find the right people in a short enough amount of time that you will lose anyway in spite of your efforts to the contrary, so you might as well take the pressure off of yourself that leads to panic and bad hiring decisions.
• Matthew McCullen was hired at GitHub after they were well-established, successfully, and hard to break into. He did it by being a renowned blogging expert at Git. Try to hire people who are smarter than you. This is tricky, but here Matthew MuCullen is an example of finding someone smart.
• Don't play to the stockholder. That is courting the short game. It is vital to have someone protecting the company from the stockholders.
• Think about who you want to go into combat with.

 
 

*Debate picks apart would-be solutions at Lean Software Austin. Once and again, I've been to this group and come away without any new solutions. Instead, challenges are discussed in such a manner that one is left with the belief that software development is just really, really hard and that there are no easy answers. Do not read this as criticism, as indeed we are talking about the hard problems people haven't solved and not the easy stuff. This session was different from the others I've been to in that there were some insights that everyone agreed upon as listed above. That said, we also struggled with:

• Kenneth Olsen's Deck was brought up as a potential company to envy, although it is long gone. They had a very flat structure and there was some debate over whether the way Deck did things was wise. Supposedly no one ever got fired there and open debate and bickering was rampant. Jay Paulson spoke of his own experience in a no-one-gets-fired-ever environment and didn't make it sound like an environment to envy.
• GitHub and Valve were brought up as companies that were doing everything right too, but Tristan Slominski counter argued that the people working at these companies have themselves as clients (as geeks need source control and geeks play games) and thus both enterprises had an advantage that other enterprises would likely not, skewing how miraculous their accomplishments really were.

 
 

Thanks to Sukant Hajra for moderating the event. This was his event. Thanks to Compass Learning for hosting and Olivier Champet of Compass Learning for keeping the doors open.