Annual or continuous reviews: what you gain and lose with each

Echipa HR 365 · reviewed 2026-09-12 · 5 min read

The choice is not between modern and outdated, but between two kinds of cost. Annual reviews concentrate the effort into one period and produce stale information; continuous reviews spread the effort and produce fresh information, but demand a managerial discipline few companies have. What works in most cases is a combination.

The comparison, on five criteria

The recency effect, quantified

In an annual review done from memory, the last two months weigh, in practice, more than the first ten. Not out of ill will: a whole year simply cannot be reconstructed in one sitting. The consequence is that someone with a good year and two weak months gets a weak review, and vice versa.

The correction does not require changing the system, only writing things down as you go. Three lines a month, after a one-to-one, turn an annual review into one based on twelve observations rather than two. It is the cheapest possible improvement and it works whichever model you choose.

The mixed model

What works in most companies under 200 people

1. Monthly — a one-to-one, with three lines noted: what went well, what did not, what comes next
2. Quarterly — a thirty-minute conversation about direction and objectives, with no scoring
3. Half-yearly — the formal review, against criteria, with a self-assessment and the manager’s ratings
4. Annually — the perspective conversation: role, development, and — separated in time — pay

The separation in the last row is what changes the quality of the conversation. When the review and the salary are discussed in the same meeting, everything said about performance is heard through the filter of the financial consequence — and the feedback no longer lands.

What it costs, in time

Two hours a week for eight people is probably the best-invested part of a manager’s time — and the first thing cut when pressure arrives. The figure is useful precisely for that: it makes visible what is being sacrificed.

What matters less than people think

• The tool. A good form inside a bad process changes nothing.
• The number of criteria. Five well-chosen ones beat fifteen generic ones.
• The scale. Four described levels work as well as ten undescribed ones, and better than them.
• Frequency on its own. A quarterly review done superficially is less useful than an annual one done seriously.

How do I convince managers it is worth the time?

Not with arguments about the importance of feedback — they know those. With the figure: two hours a week for eight people. And with proof that last year’s reviews led to a concrete decision.

Should I drop formal reviews entirely?

Rarely. Even in companies with excellent one-to-ones, a formal moment serves a function ordinary conversations do not cover: synthesis, comparability between people, and a written trace for the decisions that follow.

How do I move from annual to mixed without losing everything?

Add monthly one-to-ones with notes for six months, keeping the annual review. By the second cycle the review is far easier, and only then do you move to half-yearly, without anyone feeling a loss.

Where to start

Whatever model you have, add three lines of notes after every one-to-one. It is the one change that improves any review system and requires nobody’s approval.

Then check whether the salary conversation shares a meeting with the review. If it does, separating them is the second most effective improvement, and it costs only a move in the calendar.

What changes with company size

The third row introduces the only extra component that appears at scale: a meeting where managers compare how they rated, against the same criteria. Without it, “meets expectations” in one department can mean what “exceeds expectations” means in another — and the decisions made on those ratings become arbitrary.