You Are Not a Number

One of the most important components of being a great manager is effective performance management. To do it well, it can’t just be an annual event; you have to continually give feedback to your employees — which can be difficult, especially when feedback is tough. At year end, you may have to make hard choices on compensation: You have to pay your stars incredibly well — but those funds sometimes come from employees who haven’t performed as well. And, you may need to terminate underperformers. I’ve had my share of terminations, and I can’t tell you how much I hate having those conversations. They haven’t gotten an inch easier through the years.

In our early days, we had a structured performance management system that was designed to drive managers to make the right (and sometimes tough) choices. As part of our system, all managers rated their employees:

• 2.5 meant you were underperforming and that you had to improve significantly or you would be terminated.
• 3.0 meant that you were performing your job adequately. You were solid, but not great.
• 3.5 meant you were doing a good job. You went above and beyond when needed.
• 4.0 meant that you were a star. You went above and beyond regularly and you had potential to be on the way to a promotion.
• 4.5 meant you were a true superstar.

A key aspect of our previous system was the requirement that managers align their teams along this numeric curve. While we would accept some deviation from the standards, we wouldn’t accept a manager who declared that everyone on his team was a star. This created real discipline amongst our leadership — they had to raise the bar on a high performing team. It was often the case that our managers needed to move out the lowest 5% of our employees (the ones with 2.5s) every year — this is similar to the famous “GE” rule. The performance ratings also determined the compensation allocations. Higher ratings meant higher raises, bigger bonuses, and more equity.

We were a performance management machine. At year end, my leaders would come into my office and present their recommendations. We’d look at their distribution curves by employee level. We’d discuss their senior management teams and career planning for their stars and superstars. And, of course, we would look at compensation. The process ran like clockwork.

So, we decided to blow it all up. Why?

Because the ratings took over. Teams were spending hours and hours going over their ratings distributions to make sure that they fit the designated curves. They then spent hours trying to fit their compensation pools within their ratings ranges while being as fair as they could to their employees. And employees became obsessed with their ratings scores. From an employee perspective, their year-end feedback sessions went something like this: “blah blah blah, blah blah blah, blah blah, blah blah. Your rating is a 3.5.” When we talked about employees, we’d label them based on their rating: “He was a 4.0 last year, but he got a 3.5 in his last round.” In the end, the nuance got lost and the rating took over.

Last year, our HR team launched what we call Passport to Performance (we call it “P2P”), which was a complete redesign of our processes. We defined how we measured effective execution (e.g., results - what you got done), but also what it meant to be an effective team member (e.g., how you got it done). And we asked our managers to have real, structured and ongoing conversations with their teams, but without the crutch of a rating at the end.

What hasn’t changed is our expectation that managers give tough feedback when and where appropriate, and that they continue to pay for performance. Our managers initially had a tough time determining appropriate pay levels. In the past, ratings made it easy. A 3.5 expected a certain formulaic range for bonus, raises, and equity, so the compensation conversation was relatively straightforward. Now, our managers must give qualitative but specific feedback regarding performance and competencies and connect that feedback to dollars and cents. This aspect requires a ton of training which our terrific HR team delivered effectively in a relatively short time.

So how did it go?

Well, it has only been a year, but feedback regarding the new program has been overwhelmingly positive. We are an intensely data driven culture, so it has been difficult not having a wide swath of ratings and distribution data. However, our feedback conversations are much richer — they are about the person and their development, not why they got a certain rating. Better yet, the conversations happen regularly, giving employees the opportunity to hear both what they are doing well and where improvement is needed.

Interestingly, our managers have remained disciplined on pay for performance: we took a look at actual pay distribution and we found that our teams continued to differentiate pay based on strong (or weak) performance. I worry that, with time, we will lose our discipline and revert to “peanut butter” pay, spreading it out nice and evenly. It can be easy to do and is something we will need to watch.

The most surprising element of our new program is that it has become a great element in recruiting and differentiating us from our competition. Sadly, our HR team has rejected my recruiting motto ideas: “Come to Expedia and be a human, not a number,” or “Come to Expedia if you like numbers, but don’t want to be one.” I’ve been told to stick to my day job.

Year two of our P2P experiment is coming up. I’m hoping that it is as good as, or better than, year one.

Photo: lunchtimemama/Flickr