I’ve spent the last decade mostly working at big companies. They were some of the very best, and well run big companies: Facebook, Google, EMC. There was a lot going on within them that was, and is, worthy of admiration. But there was also some to be frustrated about.
During that time, what I became most frustrated about was the way that the system of performance management and incentives worked.
When you work at a state of the art big US corp, the typical system is something like this: Every quarter, or sometimes 6 month period, you are assessed on your performance vs. some middle point. So for example, if you are a Product Manager at level x, there is a very well defined standard for what you are supposed to achieve at that level for the period. If you achieve more, your manager will rate you as exceeding expectations, usually by one or two degrees, kind of like getting a birdie or (more rarely), an eagle in Golf. If you are judged to have achieved less, you will get a bogie, or very rarely a double bogie. Your bonus, in the form of cash or stock or both, will be bigger if you outperform, and smaller if you slip up.
To make sure things are nice and fair, people within a team or a group are calibrated to fall on a bell curve, or close to it. So if your manager wants to give you a birdie, they will have to justify why you demonstrated better results than a peer getting a par for the quarter.
Outperforming par over a few cycles consistently means promotion. I’d rather not talk much about what happens on the other side — and to be fair it’s quite rare. You will get a lot of coaching and a lot of help to set you on the right path, no matter where you are on the spectrum. Usually you will have good managers and good support by HR. These companies hire managers well and train them even better.
The system is quite sensible, on the face of it. It allows for fairness, up to a point, and as much transparency as possible, without violating employee privacy. But after going through through more than a few cycles of it, I’ve come to lament its shortcomings.
The essence of the downside is that it creates extrinsic motivation that causes you to do less than you best work, and optimize for the process rather than the objectives.
To illustrate how incentives like this can achieve less than stellar results, my favorite example is the career of the great Ukrainian pole vaulter Sergey Bubka. Bubka was a wonderful athlete, and completely dominant in his field for the better part of two decades. During his illustrious career he pushed the world record from 5.83 meters before him, to 6.14m by the time he was done, including breaking the iconic 6 meter barrier.
Bubka was so good that he ran into a bit of a bind. No sponsor was willing to pay him a large win bonus, because it was just assumed he would win. Finally Nike came up with a special incentive plan: They would pay him a $100,000 bonus for breaking the world record.
This incentive, combined with the peculiar way pole vaulting works — you set the bar at a specific height, making it easy to determine your result, unlike, for instance, the 100m dash, where it’s not easy to determine you will win it with a time of 9.67 seconds — lead to the result that Bubka broke the record again and again by 1 cm at a time, a total of 35 times.
All of this may seem to be a victimless crime (excepting perhaps the shareholders of Nike, and I don’t suppose they minded much). But the point is that it is widely believed that Bubka could have done better overall, if it weren’t for this unique system. He was good enough in his prime to push the record higher than 6.14m. He might still be holding it. Focusing on what matters could also have helped him win more than the one olympic gold he collected in Seoul.
And that, I believe is the deep flaw within how most corporate performance management systems work. Economist Dan Ariely’s popular and academic work has taught us a lot about why intrinsic motivation is better than extrinsic motivation. I’ve witnessed this time and again. When people are given a set extrinsic system, they work towards it, or sometimes game it outright. For example, if you have just been promoted and you know in the next cycle it’s going to be someone else’s turn at the right end of the bell curve, why not take a step back? Or if you are on a project where you know the right thing is to shut it down, how easy is it going to be to sell that as an above par performance, when your peer has set a measurable goal of closing 10 deals and nailed 15? There were many times when I witnessed uncomfortable right decisions and course corrections get delayed, sometime by weeks or months, so that they would fall after, and not interfere with, the current performance cycle.
I think it’s time we found ways to measure performance more smartly, and pay for it differently. I don’t claim to have all the answers. But a good place to start would be to take a long hard look at how flexible we are about incentives, and how much room we leave for judgement calls and good old common sense. It would have Served Bubka better in the long run.