Why Stack Rank Doesn’t Measure Up
In a two-part series, Ellen Chisa shares her experience at Microsoft and how its review system affected her psyche and productivity. (Read part 1 here.)
By Ellen Chisa (Product Manager, Kickstarter)
Stack ranking is a performance review system that ranks employees against each other. Critics point out that a review process that creates inevitable losers and requires managers to fight on behalf of reports is unfair and disconnected from the quality of the performance itself. Microsoft recently decided to get rid of the stack rank.
Stack ranking hurt Microsoft employees. It negatively impacted the work and damaged people’s views of themselves and their ability to improve. Here’s how:
The review process overshadowed and obscured the work
When I’m being evaluated on how good my work is, the solution is easy: do awesome work! I can tell if I’m happy with my work. I can talk to my manager to see if they’re happy with my work. If everyone is happy, I can generally expect that things are going fine.
The stack doesn’t let that resolution happen.
I was aware of the review system and stack rank as soon as I arrived at Microsoft. The first thing you do is write your “commitments”, which are what you’re evaluated against during review season. Since I started in the summer, I also saw my new coworkers get their reviews in September and how people react emotionally. It became very clear that the review process was important for many people who care about their career.
Then, you’re surrounded by reminders all the time. Your manager could respond to a new idea with “well, do you think this activity aligns with your commitments?” Or when meeting with another team, they might say something like “well, our VP committed to doing X this year.” It was hard to feel like you were doing actual work without someone bringing up commitments or review.
While I couldn’t avoid thinking about reviews, they also didn’t help me understand how I was doing. The stack isn’t only about you but how good everyone else’s work is and seems to be.
It’s hard to do well by doing good work if you’re being stack ranked against someone who focused on getting a good review. Some people at Microsoft spend their entire working year dedicated to the review process: writing status reports to elevate visibility, creating complex documents to show the manager their impact, etc. There aren’t that many, but it’s scary when people spend more time on their “career” than on the actual work at hand.
Even if both you and your manager are happy with your performance, that doesn’t guarantee a good review result after being calibrated with others in your team or in your organization (dangerous if you’re on a talented team!). Plus, your review rests on how effectively your manager negotiates for you, which makes little sense to me. How good your manager is at negotiating has nothing to do with your work.
This means you can never really know if you’re doing “well enough” for the system, which is hard to resolve mentally. When I have an unresolved question, it keeps nagging me. I was always a little bit on edge, same as when I try to work while I’m worried about something else like a health issue.
While the review and stack system was important to my career at Microsoft, it created stress and distraction. There wasn’t a clear way to actually know how things were going and move on with my work.
The forced comparison to tens of thousands of employees is confidence-crushing
I haven’t heard people talk much about how many employees Microsoft has with reference to the stack. The number I’d always heard thrown around was 90,000 full-time employees.
My memory is a little fuzzy on the exact breakdown of reviews — but let’s pretend 20% of people get a “1” (the best review). Say you get a “2” on your review. It means “exceeded expectations.” That’s great! It’s like getting a “B” in a really hard class.
When you throw the stack rank into it, it feels very different. When you get that same “2” – a good review! – the numbers say 18,000 people are doing better at their job than you are. That’s a small city worth of people, and you’re being told that all of them are better than you. It doesn’t really feel good anymore. Actually, it feels pretty lousy.
Then pretend you have a not-so-great year. You get a 4. That suddenly implies 80,000 people — that’s most people — are better at their jobs than you are.
When this happened to me, it fed into a classic case of imposter syndrome. I spent six months thinking I was a terrible project manager, worrying that everyone was going to “find out” that I wasn’t very smart, didn’t have good ideas, didn’t have a good work ethic, and that I’d been secretly fooling people for twenty-three years.
The review is about justifying the numbers rather than having a constructive conversation
Strangely, you don’t necessarily know how your review will go because of the factors out of your control.
At both of my Microsoft reviews, I walked into my manager’s office, where I was handed my review and told to sit and read before discussing. In my reviews, the “important” page was the numerical one listing your score, your merit bonus and raise, and any other associated financial changes (stock grants, R&D raises). You’re required to sign and accept the review.
Since you aren’t really involved in the discussion regarding your own review process, most of your review is spent justifying your score. In my mind, there’s a big difference between these two conversations:
- Here’s a result! Here’s some supporting data I decided to use to justify it.
- Let’s look at some patterns and trajectory throughout the year to see what the big picture is and where we can improve.
In my case, the first really in-depth conversation was when I got a final number. You can’t really change anything after you get that review because the stack model has already been locked. You don’t get a chance to respond to how people think about your work until everything is finalized.
You do also get a written review of “how you worked this year” with more qualitative discussion. In my mind, these pages are far more important. I’ve spent more time reviewing them since then and used them as markers for ways I can improve. I re-read them several times in an attempt to be fair while writing about this, and I think they’re a great part of the Microsoft process (definitely way more useful than a “stack”).
Unfortunately, managers often write their formal reviews after the numerical calibration has finished, which creates a risk of biasing the content based on the score. If you already know someone has a “2”, you might look favorably on mistakes because they’re “good”, construed as areas for improvement. The flip side is with a “bad” score, the review needs to justify it. I’d much rather have had my managers write these before the number was finalized.
Since things have already been decided, it can feel pointless and be viewed negatively to try to explain yourself. The stack makes things a lot less flexible, and that obstructs making positive progress as a result of review feedback.
So, I think the stack review process limits employee involvement. Since employees can’t participate in conversations about the stack,there’s a lot less discussion, control, and opportunity to improve. Feedback leans towards being a justification for a score, rather than being useful for future performance. This makes the review feel more like something that “happens” to you rather than a process in which you’re participating.
Without a stack, managers and employees can work more effectively together toward creative positive reviews, focused on improvement and growth.
This post previously appeared on iDoneThis Blog