Re-engineering a Better Performance Review

There’s a dirty little secret about performance reviews, which is this: most of them are useless. Managers dread them because they’re tedious and time-consuming and potentially volatile. Employees hate them because all too often a performance review is the only objective data point informing compensation and bonus. Rarely in my experience does a performance review actually accomplish its primary goal of encouraging superior performance or correcting unproductive behavior.

In fact, performance reviews generally fail for the same reason pyramid schemes succeed: there’s little incentive to be honest.

After grappling with this problem for several years, I’ve come to the conclusion that the best metric is the one favored by sales departments everywhere: profit. I try to base a large percentage of every employee’s performance review and bonus on project profitability, which accomplishes two things: it provides manager and employee alike with quantifiable data that balances out subjective feedback, and it incentivized back-office production staff to work collaboratively with front-office sales staff to their mutual benefit.

I do this, however, with a couple caveats:

  • For this sort of no-nonsense approach to work, employees have to have a reasonable amount of control over the success of their projects. If they don’t share responsibility for the estimate and/or aren’t allowed to micromanage their own time, they shouldn’t be assessed on the eventual actuals. (Note that reasonable control does not mean complete control. Part of being an effective team member is managing the risks of the team.)
  • Profitability should never be weighted so high that it inhibits innovation and risk.

Goals and Targets

To provide structure for the less quantifiable parts of a performance review, I tend to use a system of goals and targets which are negotiated at the beginning of the evaluation period and then revisited either quarterly or semi-annually. Goals are kept high-level and accommodate both peer / stakeholder feedback and supervisor feedback.

Each goal is ideally supported by specific targets which act as a roadmap for achieving the best possible score on that goal. Wherever possible, targets should be binary. “Create such and such an application” is a better target than “participate in product development.”

Although goals should be shared between all employees performing a similar task (to ensure equitability of assessment), targets need not be. Whichever path gets an employee to the desired goal is the right path.


I tend to use percentages because they’re easy to work with. Quantifiable goals can be scored with precision, whereas less quantifiable goals are best captured as thresholds (I use increments of 25%). A particularly dysfunctional company I used to work for had the dubious plan, perhaps inspired by a certain scene in Spinal Tap, of allowing scores as high as 135% in the belief that it would encourage their employees to strive that much more (35% more!), but obviously all it accomplished was the devaluation of 100%. Keep it simple: stick to the basic math.

No Surprises

The teams I’ve led will tell you that I’m candid, probably to a fault. I don’t believe in the big reveal, and I apply the same philosophy to performance reviews: if I’m doing my job as a manager, a performance review need take no longer than a handshake and a few words of encouragement. If, on the other hand, I’ve scheduled the better part of an hour to walk through a laundry list of suggestions and concerns, I probably haven’t been providing enough event-specific feedback through design reviews, status meetings and project post-mortems, and you can therefore expect more of these check-ins in the months ahead.

Comments are closed.