The four-part structure
1. Evidence review: what happened since the last conversation, with dates and specifics — not a general impression of 'how the period went'.
2. Discussion: the employee's perspective on the same events, including context the manager might not have — workload, blockers, changes in priority.
3. Forward objectives: two to four specific, measurable goals for the next period, each with how it will be checked.
4. Written summary: a short document both sides can refer back to, ideally acknowledged (not necessarily 'agreed', since disagreement can still be recorded) by the employee.
Worked example: weak vs strong
Weak: 'James needs to improve his communication and be more of a team player. Overall a fair year.' This gives James nothing to act on and gives the organisation no evidence if performance becomes a live issue later.
Strong: 'Since March, three deliverables were handed to the team without a heads-up in the shared channel (12 April, 3 May, 20 May), which caused rework each time. James says workload spiked in this period and he didn't think the channel update was necessary for internal handoffs. Agreed: all handoffs go in the shared channel going forward, reviewed at the four-week check-in on [date].' This version is specific, includes the employee's side, and sets a concrete forward objective.
Why ratings often make reviews worse
A single numeric or letter rating tends to become the entire conversation — the employee focuses on contesting the number, and the underlying detail (what actually happened, what would help) gets lost. Many UK employers have moved away from forced ratings and stack ranking for this reason, favouring narrative, evidence-based summaries with clear forward objectives instead.
Objectives the employee can accept or decline
Rather than issuing objectives as instructions, propose them and record whether the employee accepts, wants to discuss, or declines with a reason. This surfaces disagreements while they're small and cheap to resolve, instead of six months later when a missed target becomes a bigger dispute about whether it was ever realistic in the first place.
When a review surfaces a bigger problem
If the evidence review shows a genuine, ongoing performance gap — not a one-off — that's usually the point to move from a routine review into a Performance Improvement Plan, with dated objectives and checkpoints, rather than carrying the same unresolved issue into the next annual cycle. See our guide on writing a PIP.
Frequently asked questions
How often should performance reviews happen?
Many UK employers now run short, frequent check-ins (monthly or quarterly) alongside a longer annual summary, because notes logged through the year are far more accurate than a single reconstruction at year-end.
Should reviews be linked to pay decisions?
They can be, but mixing a developmental conversation with a pay decision in the same meeting often makes employees defensive and less willing to discuss genuine problems. Some organisations deliberately separate the two conversations.
What should happen to a manager's private notes?
They should stay private and never be shared verbatim with the employee or carried into a formal process — only confirmed, evidenced points that have been discussed with the employee belong in the shared record.
This is general guidance on UK workplace practice, not legal advice. If a case is contested, high-risk, or involves discrimination, ill-health or long service, take advice from a solicitor or ACAS before acting.
Next step
GoMediation.ai runs reviews without ratings — voice notes through the year, objectives the employee can accept or decline, and private notes that stay private.
See the Employee Reviews module