Designing the Performance Review Process for a Startup

Problem

I joined a small company with only about ten people as a head of engineering. Like many startups, we initially lacked any formal process for evaluating employee performance. As our employee count crossed about 50, we ran an engagement survey. We saw that people longed for more feedback from their managers: they wanted to know how well they are doing, what they can improve on, and how they can achieve their career goals in our company. However, many line managers were in a first-time manager role. Having difficult conversations, that sometimes accompany feedback giving, did not yet come naturally to some of them. I was chartered to come up with a process where employees could get the feedback they need while managers will level up in their ability to lead those crucial conversations.

Actions taken

I started off with the basics: talking to industry colleagues, reading blogs, figuring out processes that existed in similar companies. I also compiled a list of resources for managers: online courses, books, articles on feedback giving and coaching. Combining it all together we initiated the fairly standard process: a form for every employee where the manager would evaluate them on a half a dozen of attributes a couple of times a year.

After doing it twice, we received pretty lukewarm feedback from our team. From an employee perspective, the quality of feedback was pretty uneven and feedback was not timely enough. From a manager’s perspective, the time investment needed to do it was too high. Six months was a very long period of time for a company like ours. Some managers ended up just checking off the boxes: they were too busy to do it right. As such, our first iteration was not quite a success we needed.

At this point, there was a lot of feedback about the process pouring in. Some wanted to add numeric ratings, add more questions and attributes that matter, and even stack rank everyone. Others wanted to make it less frequent to reduce the time investment, or even questioned whether the company of our size needs reviews at all. I felt like we were stuck between two bad choices: either drown in more and more process or go back to where we started.

In a somewhat controversial move, I decided to go another way and do something way more extreme and a lot more simple. We pared down the review form to just two questions (“What did you do well?”, “What can you improve on?”) and we made reviews very frequent -- monthly. My idea was inspired by a concept of an MVP (Minimal Viable Product): let's put in place something simple, get feedback fast, and iterate.

And it worked! After a few months of practice, the time investment for managers went down dramatically. Employees were getting performance feedback in near-real-time. And if something went wrong (missed review, poor quality, disagreement), then the next opportunity to fix it was just a month away. There was a natural incentive to improve on both sides, and company leadership had a chance to review what managers are doing and coach them on their management skills. Our engagement survey was clearly showing improvement too.

Over the next several months we introduced a few minor tweaks, such as incorporating 360 feedback. But the MVP we put in place has scaled amazingly well even as we continued growing to 100 employees and beyond.

Lessons learned

My goal was to get all managers comfortable with giving feedback, and there is no better way of doing it than getting them to practice it often. Doing reviews monthly made all the difference: it would take years to develop the muscle of feedback giving with the annual cycle. Avoiding a pitfall of too much process was also important: once managers are comfortable with feedback conversations they do not need checklists or long forms to zero in on most important points.

If you want to become better at something, you just need to practice it often!