Performance reviews are defended as development and experienced as judgement, and the research is not kind to them.

What the evidence suggests

Ratings correlate more with the rater than with the person rated.

Which is a substantial and well replicated finding.

Feedback and pay

Combining developmental feedback with compensation decisions.

Which causes people to defend rather than listen.

Separating the conversations improves both.

Forced distribution

Requiring a set proportion in each rating band.

Which damages collaboration and has been abandoned by many large employers.

Frequency

Annual feedback about events from ten months ago.

Which is too late to change anything.

Continuous alternatives

Regular short conversations in place of an annual event.

Which several large employers have adopted.

They require managers to actually hold the conversations, which is the hard part.

Goal setting

Objectives set annually in environments that change quarterly.

Which produces either irrelevant goals or constant revision.

Shorter cycles fit most modern work better.

Calibration

Managers comparing ratings across teams.

Which reduces individual rater variation and introduces a different politics.

Documentation

Records supporting employment decisions.

Which is a legitimate reason the process persists in its formal shape.

Being honest that this is a purpose would improve how it is designed.

What people actually want

Clarity about expectations and honest information about how they are doing.

Which the annual process delivers late and diluted.

Managers who provide this continuously find the formal review uneventful.

Recency bias

Recent events dominating an assessment of a full year.

Which is well documented and largely unavoidable without notes.

Keeping brief records through the year addresses it.

Self-assessment

Asking people to rate their own performance.

Which produces systematic differences between groups and is still useful as an input.

Peer feedback

Input from colleagues rather than only the manager.

Which adds perspective and adds administrative weight.

It works better for development than for pay decisions.

The minimum that works

Clear expectations, regular honest conversation, and a documented summary once a year.

Ratings and motivation

Receiving a numerical rating tends to reduce receptiveness to accompanying feedback.

Which is a reasonably consistent research finding.

People process the number and discount the rest.

Manager time

The process consumes substantial management time each cycle.

Which should be weighed against what it produces.

Organisations rarely measure whether the process changes anything.

Development plans

Actions agreed and revisited.

Which is where the developmental value would come from if it were followed up.

Most plans are written once and never referenced again.

Poor performance

The formal process is not the place to first raise a serious concern.

Which is a common and damaging error.

Concerns should be raised when they arise.

Redesigning it

Separate pay from development, shorten the cycle, and reduce the paperwork.

A general note

Employment law affects how performance processes must be run, and requirements differ by jurisdiction.

What good feedback looks like

Specific, timely and focused on behaviour rather than character.

Which is straightforward to describe and difficult to do consistently.

Managers who do it well rarely need the formal process for anything except documentation.

Receiving feedback

The defensive response is normal and is worth noticing.

Which is the reason to separate feedback from pay decisions.

Upward feedback

Employees assessing managers.

Which requires genuine anonymity to produce anything honest.

Cultural variation

Directness in feedback is interpreted very differently across cultures.

Which matters in distributed international teams.

The practical position

Have the conversations regularly, keep notes, separate pay, and treat the annual form as documentation rather than as the management.

Small employers

Formal processes designed for large organisations transplant badly.

Which does not remove the need for clarity and honest conversation.

A short documented annual summary is generally proportionate.

The one-line summary

Feedback continuously, document annually, and keep pay decisions in a separate conversation.

A closing caution

None of this is prescriptive. Businesses differ by sector, by scale and by stage, and practices that work well in one context fail in another for reasons that are not always visible from outside.

What is consistent is that the businesses handling these questions well tend to have written something down, measured it in a defined way, and reviewed it on a schedule rather than when a problem forces the issue.

Where a decision carries legal, tax or employment consequences, professional advice specific to your jurisdiction is worth the cost, and this article is general description rather than advice.

One more thing worth saying

Most of what is written about running a business is written by people selling something, which shapes what gets emphasised and what gets left out.

The unglamorous parts, keeping records, reading the contract, updating the forecast, rarely feature because nobody can sell them. They are also the parts that most reliably separate businesses that survive from businesses that do not.

Whatever you take from this, take the habit of writing the number down and looking at it again next month.