I feel like I've been writing about performance evaluation a lot recently.
It's not that I find it a particularly interesting or engaging topic. It keeps coming up here because it keeps coming up in my professional life, and I write about the things I experience or need to work through mentally.
So, without further ado, lets get cracking and write about it some more!
You Say Black, I Say White
Performance evaluations are probably a necessary evil.
Assuming that the core intent is to identify those who are not meeting expectations and help to bring them up to speed, they feel like they sit on the stick side of the carrot and stick model of motivation.
I'm not a fan of the stick.
I'm not here to denounce performance evaluations though. Maybe one day, but not today. Also, they are something I have to do in filling my role as an Engineering Manager, so it's best to look for the bright side.
Performance evaluations are a highly effective, regular, structured way to generate feedback for people, which in turn helps them to grow into a more awesome version of themselves. Everyone wants to be more awesome, right?
At Atlassian, a performance evaluation considers the following pillars:
- Project impact
- Craft excellence
- Team direction
- Organisational impact
Each pillar is viewed through the lens of our values. For example, if you create amazing impact, but you completely fuck the customer while you do it, you're not performing well. You're just a dick.
The entire performance evaluation distils down into a set of written feedback and one of the following possible ratings:
- Does not meet expectations
- Meets most expectations
- Meets expectations
- Exceeds expectations
- Greatly exceeds expectations
I'm not going to go into any more detail about the model here because I've written about it before and I don't like to repeat myself too much. Too much.
These sorts of formal performance evaluations are done once every twelve months, to align with the end of the financial year. They are also a key input into any compensation changes.
You Say Bark, I Say Bite
Twelve months is a long feedback loop and long feedback loops suck.
It doesn't matter what you're doing, if there is a long time between when you do the thing and when you see the ramifications of the thing, it's easy to lose the connection between the two.
Performance evaluations are one of the most obvious ways to get feedback on how you're doing professionally. To get an understanding of if you are going to achieve the level of performance you set out to achieve, whatever that may be.
So why not do them twice a year?
Break the financial year up into two halves and then semi-independently evaluate each half. Twice the feedback, half the waiting period.
It makes sense. A tighter feedback loop means more chances to improve for the individual and a better chance of higher overall performance for the organisation. Low performers can be identified and helped sooner, high performers can be recognised more rapidly and subsequently accelerated further.
You Say Shark, I Say Hey Man
In case it's not already obvious, Atlassian has decided to do two performance evaluations a year. Our annual performance process is called OneCycle, so the idea of doing it twice has unofficially been dubbed Bi-Cycle, which, honestly, is a pretty good name.
I'm sure a more official name will be released at some point.
As for the change itself, I'm not really sure how I feel about it.
My first reaction is that it feels like a significant increase in the amount of performance evaluation work that I'll need to do. You might even say it's twice as much, because that's how math works. This is exacerbated by the increased expectations around how many engineers a front-line manager is responsible for.
I already have plenty of things to do. More things than I can effectively do actually, so adding more things feels like a recipe for disaster. Especially for me, because I won't compromise on quality, I'll just work harder and longer.
I should probably work on that trait.
It means that I'll need to be pickier about where I focus my effort, which is fair, but I don't want to just focus all of my effort on evaluating people. There are a bunch of other things that I want to be doing, like facilitating the delivery of value to our internal and external customers.
The second reaction is about stress. The annual performance review period (aka June-July) is easily one of the most stressful times of the year for me, so having to do it twice isn't something I'm looking forward to.
Granted, I should only have to go through half as much information in order to generate a rating, but the gathering of information has never been the main source of stress. Finding and reading stuff is easy, if time consuming.
The stress comes from actually having to evaluate someone. To decide how they are doing when compared against a constantly shifting and fluid set of expectations and subsequently deliver that information to them in a way that doesn't make them sad.
The third reaction is more positive, in that this is an opportunity to improve. I'm a big fan of the idea that if something is hard you should do it more often. After all, frequency reduces difficulty. Doing performance evaluations twice a year is a great opportunity to get better at them, which will make them easier. Less effort, less stress.
I Choose To Ride This Bike!
At the end of the day, twice yearly performance reviews are something that is happening, so there is no point in lamenting. Best to just accept it and move on with my life.
Like all things, I'll give it my best and if it's terrible, I'll give that feedback as loudly and as clearly as I can.
Ironically I was already trying to do a pseudo performance evaluations for everyone once a quarter anyway. I wanted to do this because of the reflections I had from previous cycles.
It feels a bit different when someone tells you have to do it though.
Management load, stress and questions about the fundamental value of performance evaluations aside, I kind of like that I'll get two evaluations a year.
More visibility into how I'm doing can't hurt.