What is continuous performance management? It is a way of running performance in which goals, feedback and development are handled through short, regular conversations across the year, kept on a written record, with a lighter formal review that summarises that record and feeds pay and promotion decisions. It does not abolish reviews or ratings; it moves the work out of one annual meeting and into the weeks where it happens. The CIPD’s current introductory factsheet, Performance management: an introduction (January 2026), describes performance management as a continuous cycle rather than a one-off event, with objectives that change as business priorities change and feedback that is regular, timely and aimed at improvement. For a company of 100 to 1,500 people, the practical question is which conversations run on which rhythm, and what has to be true of managers before the model works.
Key takeaways
- It is a system, not a meeting. Four parts work together: goals that can be updated, regular one-to-ones, a running record of feedback and evidence, and a periodic review where decisions are made.
- Reviews and ratings usually survive. McKinsey (2016) reported that nearly nine in ten companies worldwide still produced performance scores and used them to inform pay. Continuous management changes when evidence is gathered, not whether decisions are made.
- More feedback is not automatically better. Kluger and DeNisi’s 1996 meta-analysis found feedback improved performance on average, yet more than a third of feedback interventions made performance worse, especially when attention shifted from the task to the person.
- Manager skill is the bottleneck. In a 2017 CIPD field trial in three UK public-sector bodies, manager training made one-to-ones more useful, but more frequent performance conversations needed training plus aligned HR policy and guidance.
- The shift is a decade old, so the evidence is real. By 2016, Cappelli and Tavis wrote in HBR that more than a third of US companies had dropped traditional appraisals.
Follow us on Google
What continuous performance management covers
The term gets used loosely, often as a label for “more check-ins”. A more useful definition starts from the problem it solves. Elaine Pulakos and colleagues argued in Industrial and Organizational Psychology (2015) that formal systems had turned performance management into periodic procedures cut off from everyday work, while the behaviours that actually drive performance happen daily: setting expectations as work shifts, giving informal feedback close to the event, and developing people through the work itself. Continuous performance management is the attempt to build the system around those daily behaviours instead of around the form.
In practice that means four components:
- Living goals. Objectives are set for a quarter or a project and revisited when priorities move, rather than fixed for twelve months. The goal-setting guide covers how to write goals that can be scored at that pace.
- Regular one-to-ones. Short manager conversations, typically weekly or fortnightly, covering progress, obstacles and how the work is being done.
- A running evidence record. Brief dated notes of outcomes, decisions and feedback, so nothing depends on memory in December.
- A periodic decision point. A formal review, usually quarterly or six-monthly, followed by calibration, where the record is turned into ratings, pay and promotion outcomes.
Take away any one of these and the model degrades. Check-ins without a record leave the review to recall. A record without a decision point leaves employees unsure where they stand.
What it is not
It is not the end of ratings. Some well-known employers removed ratings altogether, but McKinsey’s 2016 article Ahead of the curve: the future of performance management noted that the large majority of companies still scored employees and tied those scores to pay. The common pattern it describes is narrower: differentiating clearly at the high and low ends, and spending less effort on fine distinctions in the middle.
It is not monitoring. The same McKinsey piece warns that real-time feedback tools can leave employees feeling watched. Continuous management is about more frequent conversation, not more surveillance of activity.
It is not a weekly form. If every one-to-one requires a template with ten fields, managers will either stop doing them or stop writing anything useful. Two or three lines per conversation are usually enough.
What the evidence says
The direction of travel. In The Performance Management Revolution (HBR, October 2016), Peter Cappelli and Anna Tavis wrote that more than a third of US companies had abandoned traditional appraisals, and that the emphasis was moving from accountability for the past towards learning and improvement. The CIPD’s 2026 factsheet describes a similar shift in the UK: less weight on annual appraisals and forced distributions, more on regular, coaching-style conversations, and it states there is good evidence some of these changes are worthwhile rather than fashion.
Why the annual form struggles. McKinsey (2016) observed that many companies track 15 to 20 individual indicators per person, and that it regularly saw KPIs accounting for less than 5% of an overall rating. A once-a-year judgement across that many measures is hard to make accurately. Shorter goal sets reviewed more often are easier to score and easier to discuss.
Feedback can backfire. Avraham Kluger and Angelo DeNisi’s meta-analysis in Psychological Bulletin (1996) covered 607 effect sizes and 23,663 observations. Feedback raised performance on average (d = 0.41), but over a third of interventions reduced it. Their explanation is that feedback helps when it keeps attention on the task and hurts when it turns attention to the person’s self-image. Continuous performance management multiplies the number of feedback moments, so it multiplies this risk too, unless managers know how to keep feedback specific to the work.
Training changes the conversations. The CIPD’s Strengths-based performance conversations: an organisational field trial (Gifford, Urwin and Cerqua, November 2017) tested manager workshops in HMRC, HM Prison and Probation Service and the Valuation Office Agency against control groups. Across the organisations, employees became 9.7% more likely to agree their one-to-ones helped them learn and develop, and 7.4% more likely to agree they helped them perform better. Employees in the intervention group were also 9.1% more likely to say they received regular feedback, although the authors note this mostly reflected avoiding a decline. The fuller intervention, which added HR policy changes and guidance to training, was the one linked to more frequent performance conversations.
Annual appraisal vs continuous performance management
| Element | Annual appraisal | Continuous performance management | What the evidence suggests |
|---|---|---|---|
| Goals | Set once a year, often out of date by mid-year | Set per quarter or project; updated when priorities move | CIPD 2026: objectives should evolve with business priorities |
| Feedback | Concentrated in one meeting | Short, task-focused feedback in regular one-to-ones | Kluger and DeNisi 1996: feedback helps on average but harms in over a third of cases when it targets the person |
| Evidence | Written from memory at year end | Dated notes captured through the period | Pulakos et al. 2015: link the system to day-to-day behaviour |
| Formal review | Long form, many indicators | Short summary of the record, quarterly or six-monthly | McKinsey 2016: 15 to 20 KPIs per person dilute focus |
| Ratings, pay and promotion | Decided in the same meeting as feedback | Decided after calibration, usually in a separate step | McKinsey 2016: most firms still score and pay on scores; differentiate clearly at the ends |
| Manager role | Assessor once a year | Coach through the year, assessor at the decision point | CIPD 2017 field trial: training improved one-to-ones; frequency needed policy support too |
If you are moving from a single annual meeting, the pillar page on alternatives to annual performance reviews sets out the options that combine into a continuous model: quarterly scorecards, calibration, feedback in one-to-ones and six-monthly promotions. For the timing of each layer, see how often performance reviews should be conducted.
Is it right for a company of 100 to 1,500 people?
Usually yes, but the reasons change with size. Below about 100 people, a founder or People lead can often see most of the work directly, and an annual cycle with good informal habits may be enough. Between roughly 100 and 1,500, managers multiply, priorities shift by quarter, and the annual cycle starts to depend on dozens of managers remembering a year of work. These signals suggest it is time:
- Goals set in January no longer describe what teams are doing by spring.
- Employees are surprised by their rating, which suggests feedback is not happening between reviews.
- Calibration turns into a debate about who remembers what, rather than a comparison of evidence. The calibration meeting guide shows what an evidence-led session looks like.
- The review cycle consumes weeks of manager time; price it with the performance review cost calculator before deciding.
- Review data lives in spreadsheets that someone has to merge by hand. The guide to switching from spreadsheets to performance management software covers when that becomes the constraint.
It is a poorer fit where managers have very large spans and no time protected for one-to-ones, or where roles are so standardised that output measures already give continuous feedback. In those cases, fix spans or measures first.
Where it goes wrong
- Frequency without skill. More conversations run by untrained managers produce more of the person-focused feedback that Kluger and DeNisi found can lower performance. Build manager capability first; the manager training guide has a starting curriculum.
- Training without policy. The CIPD 2017 trial suggests workshops alone improve quality but not necessarily frequency. Change the policy, the calendar and the expectations at the same time.
- No decision point. If continuous feedback replaces the review entirely, pay and promotion decisions drift back to opinion. Keep a periodic, calibrated decision.
- Admin multiplied. Running a long annual form four times a year is not continuous management; it is four annual reviews. See more review cycles, same managers.
- Feeling watched. Logging every interaction or ranking people on activity data will undermine trust. Record outcomes and feedback, not keystrokes.
How to know it is working
- One-to-one completion rate against the agreed rhythm, by team.
- Evidence depth: number of dated notes behind each rating at the decision point.
- Surprise rate: share of employees who say their formal review contained news to them.
- Usefulness: ask the two questions the CIPD trial used, whether one-to-ones help people learn and develop and whether they help them perform better, before and after the change.
- Calibration time per department, compared with the last annual cycle.
For a quick baseline across these areas, run the free performance health check. If you would like a second view of your current set-up, send a short brief through the People Performance Playbook intake.
Our perspective
Most People leaders I speak to already believe in continuous performance management. What stops them is not the idea but the plumbing: notes in one tool, goals in a slide deck, reviews in a spreadsheet, and calibration in a meeting room where nobody can see the evidence. Managers then experience “continuous” as extra admin, and the initiative quietly fades by the second quarter.
So I would start with the record, not the rhythm. Agree what a good two-line note looks like, where it lives, and how it turns up at review time. Once managers see their own notes pre-filling the review, the weekly habit sells itself. Then train them on keeping feedback about the work, because that is where the research is clearest about what can go wrong.
— Ash Taherian, Product Sales Executive, Revolut People
How Revolut People supports continuous performance management
I work at Revolut People, so read this section with that in mind. The relevant question for any system is whether the four components share one record. In Revolut People, goals, one-to-one notes, review cycles and calibration sit together, and reviews pre-fill with goals and past-cycle data. Once three or more one-to-ones in a cycle have been summarised, the AI assistant can draft feedback from them, with a quality check that flags drafts which are not specific, evidence-based or actionable. Ebury’s published story reports departmental calibration falling from 4 hours to 1 after moving off disconnected spreadsheets (Ebury customer story). What the software does not do is train your managers, set your policy or choose your cadence. For the manager experience in detail, see Revolut People for line managers.
If you want to check whether your current cycle could run continuously without adding manager admin, book a 15-minute review. Bring your review calendar and one anonymised one-to-one template.
Step plan: one quarter to switch on
- Week 1–2: Define the four components. Write down the goal period, one-to-one rhythm, note format and decision point. Keep each to a sentence.
- Week 2–3: Cut the formal form. Remove every field not used in calibration or pay. Aim for a handful of goals and a short scorecard per role.
- Week 3–4: Change the policy, not just the training. Publish the one-to-one expectation and the review dates, and brief managers on task-focused feedback.
- Week 5–10: Pilot in two teams. Run one-to-ones with two-line notes and a mid-quarter goal check. Track completion and ask the two CIPD usefulness questions at the start.
- Week 11–12: Run the decision point. Hold a short quarterly review from the notes, then calibrate. Compare manager hours and calibration time with the last cycle using the review-cycle checklist.
- Week 13: Decide and scale. Re-ask the usefulness questions, fix what the pilot flagged, and extend to the next set of teams.
Bottom line
What is continuous performance management? It is performance management run as a cycle of living goals, regular one-to-ones, a running evidence record and a periodic calibrated decision, rather than a single annual event. The CIPD’s 2026 guidance supports the continuous model, HBR (2016) and McKinsey (2016) document the shift and its limits, and the research from Kluger and DeNisi (1996) and the CIPD’s 2017 field trial shows the outcome depends on how well managers give feedback and whether policy backs them. Start with the record and the manager skills, then raise the frequency.
FAQ
Does continuous performance management replace annual reviews?
Not usually. It replaces the annual review as the only moment for feedback and goal-setting. Most organisations keep a formal review and calibration, often quarterly or six-monthly, because pay and promotion still need a decision point. McKinsey (2016) reported that nearly nine in ten companies still used performance scores to inform pay.
How often should managers give feedback?
Often enough that the formal review contains no surprises, which for most teams means a short one-to-one every week or two. Frequency matters less than focus: Kluger and DeNisi (1996) found feedback helps most when it stays on the task and can hurt when it becomes about the person.
How do you give constructive feedback at work?
Describe a specific piece of work, explain its effect, and agree one next step, close to the event and in private. Keep it about the work rather than the person’s character. The CIPD’s 2017 field trial found that training managers in strengths-based conversations made one-to-ones more useful for both performance and development.
How often should 360 feedback be used?
Less often than one-to-one feedback. Many organisations use multi-rater feedback once or twice a year for development, alongside weekly or fortnightly manager conversations. Our post on whether 360 feedback should be anonymous covers rater numbers and debriefs.
Sources
- CIPD: Performance management: an introduction (factsheet, 29 January 2026) - performance management as a continuous cycle; objectives evolve with priorities; regular, timely feedback; shift away from annual appraisal and forced ranking towards regular, coaching-style conversations
- Peter Cappelli and Anna Tavis: The Performance Management Revolution, Harvard Business Review (October 2016) - more than a third of US companies had abandoned traditional appraisals; shift from accountability to learning
- Boris Ewenstein, Bryan Hancock and Asmus Komm, McKinsey: Ahead of the curve: the future of performance management (16 May 2016) - nearly nine in ten companies still generate scores and use them for pay; 15 to 20 KPIs per person; KPIs under 5% of a rating; risk of employees feeling watched
- Avraham N. Kluger and Angelo DeNisi: The effects of feedback interventions on performance, Psychological Bulletin 119(2), 254-284 (1996) - 607 effect sizes, 23,663 observations; average d = 0.41; over a third of interventions decreased performance
- Elaine D. Pulakos, Rose Mueller Hanson, Sharon Arad and Neta Moye: Performance management can be fixed, Industrial and Organizational Psychology 8(1), 51-76 (2015) - formal systems disconnected from day-to-day behaviours that drive performance
- Jonny Gifford, Peter Urwin and Augusto Cerqua, CIPD: Strengths-based performance conversations: an organisational field trial (November 2017) - group randomised trial in HMRC, HMPPS and VOA; 9.7% and 7.4% impacts on usefulness of one-to-ones; 9.1% on regular feedback; frequency gains needed the fuller intervention