I’ve never experienced a company with a completely healthy performance review culture. To be fair, it’s quite likely that my sample size is just skewed. Still, giving feedback well is tricky, and performance reviews always have the added complexities of introducing a power dynamic and being fundamentally transactional. Not the coziest starting point.
And yet there is room for a middle ground between pretending it’s a coffee chat between friends and a professional colonoscopy. On paper, performance reviews should work great for both parties: the manager gets to help propel the report into a position as a more productive member of the team, and the report learns what is expected of them to reach the next stage of growth and compensation—or sometimes just avoiding getting terminated.
Mandatory disclaimer: my background is not in HR. I’m reporting what has worked for me, as well as some wishful thinking that I hope is still actionable. As always, my bias is towards small to medium-sized teams.
We need to start defining stuff before we get into the weeds, so here goes: a performance review is a periodic written summary of impact and behaviors over a defined time window, plus an agreed growth plan. It is often accompanied by an in-person check-in, but doesn’t have to be.
It is not a compensation negotiation (which is related but separate), a stack ranking ritual, or a place to air our grievances from the last six months. We don’t just collect feedback so we can shove it into a two-page document about all our latest failings, ideally we have a better process for giving quick, actionable, kind feedback.
So, what do we want to achieve, then?
After the performance review is done, we want to feel aligned and ready to tackle the next period of work as an even better team. To do this, we need to strive for clarity in our communication, fairness in our evaluation, and stay actionable. Let’s go through these in turn.
Clarity is achieved by avoiding hedging while staying kind. Hopefully there are no surprises anyway, since your company has an ongoing feedback culture and avoids in-office politics that reward backstabbing and careerism, right? On the off chance that there is a political component, we do our best to stay evidence-based and bring receipts.
This brings us to fairness immediately. We need to factor in things like the individual’s experience in the job vs. in the wider industry, team and company situation within that period1, and any potential biases we might have (this includes things like charisma and manager proximity).
Finally, our review needs to be actionable. No matter whether there are things to improve or we mostly want to stay on our current trajectory, we still need to set explicit goals to make performance measurable. I also advocate for at least two levels of goals: goals for satisfactory performance, and goals for exceptional performance. This makes room for stretch goals while reducing anxiety about what is necessary vs. good to have.
We should probably also be explicit about what we don’t want.
Conversely to our goals above, we do not want to be vague, personal, or give surprise feedback. These non-goals are just derived from the above.
But there’s more failure modes I’ve seen, often related to problems in middle management and their KPIs.
Firstly, managers shouldn’t be obligated to hit a quota or give out a certain number of positive or negative reviews. We’re not doing curve-fitting here.
Secondly, we should keep the number of goals manageable. 20 tiny goals will lead to less overall change than 3-5 big ticket items. And of course all of them should be explicit; hidden criteria that only pop up once review rolls around are embarrassingly common and directly lead to frustration and, consequentially, a loss of valuable team members.
Finally, we should look at “non-primary” work as well, and reward it. Mentoring, incident response, reviews, quality documentation should all be rewarded, especially the higher the report’s level. Of course this must be tailored to the career track the person is on, especially if there is a “manager” vs. “technical” track. Again, be as explicit about everything as possible.
We should briefly talk about how long our review cycle should be, and how it should work.
I’ve seen everything from yearly to bi-annual to quarterly. Generally, the younger the company (and thus faster moving), the shorter the cycles. I’d personally try to keep the cycles as short as possible organizationally while still providing enough input into how performance actually trended. Six months is what I see most often. I personally find a year too slow, but I’ve also never worked at a corporate behemoth.
The review is informed by two reviews (one by the report and one by the manager), artifacts the report produced (code, documentation, PRs, incident reports in my domain), expectations set by our career ladder2, as well as any on-topic 1:1s and ongoing feedback logs we still have. This gives us more than enough data to figure out a rough trajectory.
You don’t need a formal document for the reviews themselves, but I’ll sketch up an example that more or less contains all the inputs and output I’d expect:
# Performance review <name>: <period>
**Role**: <role with link to description>
**Next follow-up**: <date>
## Summary
<2-3 lines of summary, longer if desired>
### Role alignment
<how did we do in comparison to our ladder, including flagging over-performance>
## Outcomes last period
| Description | Achieved | Notes |
|-------------|----------|-------|
| | <marker> | |
## Growth plan next period
- <bulleted list is good enough>Reviews are always scary, and it’s probably impossible to make them completely anxiety-free. But an organization that commits to making growth the goal goes a long way towards making reviews productive and actually motivate team members rather than scare them.
In the end, I just want to have people leave the room feeling seen and ready to tackle the next 3-6 months. If we achieve that, we did most things right.
1. It should go without saying that if you had a lot of attrition or a market crash throw a wrench in our last quarter this should factor into our performance review and whether we reached our goals. I can tell you, however, that it doesn’t go without saying.
2. I’ve seen a lot of places that have roles without descriptions what they actually mean (i.e., staff, senior, and junior handled at the discretion of the manager). Don’t do that. You don’t have to spell out every single metric, but having expectations for your roles makes everyone’s lives easier: managers know what to expect, and reports know what to strive for to climb the ladder.