Annual performance reviews should be illegal

I’m being slightly hyperbolic with the headline. Slightly. But I’d genuinely support a workplace law that banned the annual performance review in its current form, because I’ve watched it do more quiet damage than almost any other standard management practice, and almost nobody defends it on its merits. They defend it because it’s what everyone does.

Let me make the actual case, and then I’ll tell you what to do instead, because “abolish the thing everyone uses” is only useful if there’s a better thing.

Key takeaways

• Annual performance reviews systematically over-weight the last month: human memory can’t fairly compress a year of behavior into one sitting.
• Bundling feedback with compensation turns a development conversation into a negotiation; neither goal gets met.
• Single-manager reviews measure the manager’s perception as much as the employee’s performance.
• The fix has three parts: separate feedback from pay, make it frequent and small, widen the input to more than one perspective.
• You can do all three with conversations and a shared document. Software is optional.

The annual review fails at the one job it claims to do

The stated purpose of an annual review is to help people improve. Hold that next to how it actually works.

You take a full year of someone’s work, hundreds of decisions, dozens of projects, countless interactions, and you compress your assessment of it into a single conversation in, say, March. By then, most of what happened is forgotten or distorted by recency. The thing they did brilliantly in May is gone. The thing they fumbled three weeks ago feels enormous. Research on this is unambiguous and decades old: human memory can’t fairly weigh a year of behavior in one sitting, so the annual review systematically over-weights the last month and whatever was emotionally vivid.

So the feedback arrives late, distorted, and all at once. If your actual goal were to help someone improve, you would never design it this way. You’d give feedback close to the event, in small doses, while it’s still actionable. The annual review does the precise opposite and then we act surprised that people find it useless and stressful.

It makes feedback high-stakes, which makes it dishonest

The deeper problem is what happens when you bundle a year of feedback into one event and, this is the killer, attach it to compensation.

The moment a conversation determines someone’s raise, it stops being a feedback conversation and becomes a negotiation. The employee’s job is no longer to hear honestly; it’s to defend, justify, and protect their number. The manager’s honest observations get softened or weaponized depending on the budget. Nobody in that room is doing the thing the meeting supposedly exists for. They’re both playing a different game with money on the table.

You cannot get honest developmental feedback and make a compensation decision in the same conversation. They are opposing activities. Trying to do both at once guarantees you do neither well, and everyone leaves the room having learned to perform rather than to listen.

It concentrates all the power in one person’s memory

Traditional reviews run on a single input: the direct manager’s recollection. One person, one perspective, one set of biases, assessing a year they only partially witnessed. The manager wasn’t in most of the meetings, didn’t see most of the collaboration, and has their own blind spots about what good work looks like.

This is why the same employee can be a star under one manager and a problem under the next, with their actual behavior unchanged. The review measures the manager’s perception as much as the employee’s performance. For something that determines careers and compensation, that’s an absurdly thin and biased instrument.

What to do instead

So if I’d abolish the annual review, what replaces it? Three things, none of them revolutionary, all of them better.

First, separate feedback from compensation. Make them different conversations at different times for different reasons. Have compensation discussions as their own thing, driven by role, market, and contribution. Have feedback conversations purely to help people grow, with nothing financial riding on them. The instant you decouple them, both get more honest.

Second, make feedback frequent and small. The fix for “a year compressed into one distorted conversation” is feedback close to the event, in small doses, all year. A quick word after a project. A standing prompt in one-on-ones. The goal is for feedback to be so normal and low-stakes that no single conversation carries existential weight. Building a genuine feedback culture requires more than good intentions, though; that post covers what actually has to be true before the rhythm sticks.

Third, widen the input. Stop relying on one manager’s partial memory. A few times a year, gather lightweight, structured feedback from the people who actually work with someone: peers, reports, the manager. Not as a verdict, but as a fuller picture than any single person can offer. This is the part of multi-source feedback worth keeping: more perspectives, lower stakes, aimed at growth rather than judgment.

Notice that none of this requires the dreaded ritual. It requires a rhythm.

The honest caveat

I run a company that builds 360 feedback software, so you should discount my view accordingly, I clearly have an interest in you doing more structured feedback rather than less. Fair.

But notice what I’m not arguing. I’m not arguing you need to buy anything, including from me. You can do every one of these three things with conversations and a shared document. The argument isn’t “buy software.” The argument is that the annual performance review, as practiced almost everywhere, is a genuinely bad design for the job it claims to do, late, high-stakes, distorted, single-source, and that nearly anything built on frequent, low-stakes, multi-source feedback beats it.

Abolish the ritual. Keep the rhythm. The tooling, if you ever want it, is the easy part. Lynxify is built for exactly that stage, no annual contract, no credit card required.

FAQ: annual performance reviews

Are annual performance reviews effective?

No, not in their current form. The core problem is design: annual reviews compress a year of work into a single conversation, are typically tied to compensation, and rely on one person’s partial memory. Each of those choices works against the stated goal of helping people improve. The research on recency bias in performance assessment is decades old and unambiguous.

What should replace an annual performance review?

Three things: separate feedback from compensation (different conversations, different timing), make feedback frequent and small (quick notes, one-on-one prompts, lightweight rounds throughout the year), and widen the input (structured feedback from peers and reports, not just the manager). None of this requires software; a shared document and a commitment to rhythm gets you most of the way there.