Why you should separate OKR and compensation

OKR is a management tool, not an employee evaluation tool. As such, a second tenet of the OKR framework is to separate OKRs from compensation and promotions.

As Intel’s Andy Grove, wrote:

It is not a legal document upon which to base a performance review, but should be just one input used to determine how well an individual is doing.

Rick Klau wrote:

OKRs are not synonymous with employee evaluations. OKRs are about the company’s goals and how each employee contributes to those goals. Performance evaluations – which are entirely about evaluating how an employee performed in a given period  –  should be independent from their OKRs.

This is very different than the traditional model, which is showing signs of aging. A study by Willis Towers Watson showed that traditional pay for performance tools are not effective at driving improved individual performance, nor at rewarding it:

  • Only 20% of employers in North America say merit pay is effective at driving higher levels of individual performance at their organization;
  • Employers give short-term incentives low marks. Only half say short-term incentives are effective at driving higher levels of individual performance, and even fewer (47%) say that these incentives are effective at differentiating pay based on individual performance.

The tale of two bonuses

There once was an organization that had two employees in the same team: Paul and Mary.

  • Paul was smart, focused and delivered results. But he was driven by monetary rewards and was always trying to figure out how to make more money.
  • Mary was also smart and focused but she was driven by her achievements. She believed that if she delivered, money would follow.

The organization used a simplified bonus formula, connecting goals to bonuses:

Bonus paid = ƒ(% of goals achieved * salary grade)

This means that the size of the bonus was a function of the employee salary grade and the percentage of goals that the employee achieved.

And then, the following happened:

  • Paul achieved 110% of an easy goal, that he successfully reduced after several rounds of negotiation with his managers;
  • Mary achieved 80% of an extremely hard goal, going way beyond anyone in the company thought was possible. A true stretch goal.

Who deserved the higher bonus? Mary, of course.

But who got the bigger bonus in the end? Paul.

This tale is a classic example of a perverse incentive. Our incentive system is, for all practical purposes, rewarding the inappropriate behavior.

We are all Paul and Mary

Everyone has a bit of Paul and Mary inside. Your incentive system should work for real people in real life. And even if you have team full of Marys, why would you have a system that incentivizes something you do not want to happen?

If you want to create a culture in which setting stretch goals is the norm, you should really think about dropping the formula-based (or tightly coupled) model for both bonuses and promotions.

What’s the alternative?

The alternative is to adopt a system in which the achievement of goals is an input to the performance evaluation process, in which bonuses and promotions are defined.

In this model compensation and goals are loosely coupled.

Basically, the performance evaluation considers not only the percentage of the goals achieved but also the goals themselves: the difficulty and the impact on the business.

Think about it as the Difficulty Score in gymnastics: you get more points for performing routines that are more difficult.

“But this is too subjective”

One of the common complains about this model is that it is “subjective” while the formula-based model is “objective”.

The problem is that simply using a formula at the end of a process does not make it objective. People simply think it is objective because all they can see is a bit of math:

  • Several companies around the world use, at least sometimes, spot bonuses or discretionary bonuses to compensate or complement the bonus policy. Both are 100% arbitrary, following subjective rules;
  • Calculating the bonus based on who has the best negotiating skills to reduce the goals is “subjective”;
  • Project/resource allocation is arbitrary. Sometimes the organization needs somebody to turnaround a trouble project, which may hurt his/her bonus in the short run – which is usually compensated by spot bonuses.

As with stretch goals/moonshots, I strongly recommend that you don’t adopt this model in the beginning. Do not change your compensation model before having a stable and mature OKR capability in your organization.

And how about sales quotas?

Sales teams are different, since the result is easier to measure. You can attach compensation to a sales quota, but you should avoid any model that rewards employees that negotiate a reduction in the quota.