A development centre is a structured session in which people work through simulated professional situations while trained assessors observe what they actually do. The output is not a score used to decide anything about them. The output is a description of how they work, and a plan built from it.
Quick Overview
The method is widely confused with its older sibling, and the confusion is not academic: it changes what participants disclose, what assessors write down, and whether anything happens afterwards. This article treats the development centre as an instrument of growth and works through it end to end:
- What the method is — and the single design decision that separates it from selection
- The exercise set — what each format is actually built to surface
- Designing the session — choosing competencies, briefing participants, preparing assessors
- Inside the session — observation, assessor integration, and the feedback conversation
- Who carries the result — the division of responsibility that decides whether it lands
- From diagnosis to plan — turning observation into an individual development plan
- What the organisation gets back — beyond the individual participant
This is the opposite question to the one a salary benchmark answers: a benchmark tells you what the market currently pays for a capability somebody already has, while a development centre tells you which capability a person could plausibly hold next. One prices proven work; the other locates unproven potential.
What a Development Center Is, and What It Is Not
The mechanics look familiar. Participants work through simulations that mirror real professional situations — a full inbox, a disputed decision, a difficult conversation, an analytical problem with incomplete data. Assessors watch and record behaviour against defined competencies rather than impressions. The U.S. Office of Personnel Management describes this family of methods under Assessment Centers, and the observational apparatus is essentially shared.
What differs is the purpose, and the purpose changes everything downstream. A selection centre exists to support a decision about a person: hire, promote, do neither. A development centre exists to support a decision by a person about their own next step. The results are confidential to the participant, they are not routed into a personnel file, and they carry no consequence for status. That is not softness — it is the precondition for the data being any good. A participant who suspects the observation feeds a promotion decision will perform the behaviour they believe is wanted, and the assessors will diligently record a performance rather than a person.
The professional literature on selection keeps these purposes apart for the same reason. The Chartered Institute of Personnel and Development treats structured simulation under Selection methods when the aim is a hiring judgement, and the framing shifts entirely once the aim becomes development. Organisations that blur the two get the worst of both: the cost of a selection centre and the candour of an appraisal. Naming the competencies precisely before anything else is what keeps the boundary visible, which is where a well-built competency matrix earns its place.
The Exercise Set and What Each One Surfaces
Exercises are not interchangeable. Each format has a narrow band of behaviour it reliably surfaces, and choosing on the basis of what looks impressive rather than what needs observing is the most common design error.
An in-tray exercise — a queue of competing demands under time pressure — surfaces prioritisation, delegation and the willingness to decide with incomplete information. It is unusually good at showing what somebody does when they cannot do everything, which is the ordinary condition of managerial work and almost never visible in an interview.
A group discussion, whether roles are assigned or not, surfaces the interpersonal register: how a person builds an argument, whether they make space for others, what happens to their position under disagreement. Assigned roles let assessors separate the person’s natural style from the behaviour a role demands of them, which is exactly the distinction a development conversation needs.
A case analysis surfaces reasoning. In a selection setting the interest is whether the conclusion is right. In a development setting the interest is the route to it — what was considered, what was discarded, what assumption went unexamined. The written output matters less than the reconstruction afterwards.
A role play — a performance conversation, an escalating client, a peer negotiation — surfaces behaviour under interpersonal pressure, which is the hardest thing to self-report accurately. Participants routinely describe themselves as calm and direct, then discover on playback that they conceded early or hardened too fast.
A presentation surfaces structure, adaptation to an audience and composure. Diagnostic questionnaires on working style or preference are often added alongside, not as verdicts but as a second lens to test observed behaviour against self-perception. Where the two disagree is usually where the useful conversation starts.
The mix matters as much as the individual choices. Each target competency should appear in more than one format, because a single observation is an anecdote and a pattern across formats is evidence. A person who dominates a group discussion but delegates well in the in-tray is telling you something specific about how they respond to an audience, and that finding is only available because the two exercises sat in the same session.
It is equally important to be honest about what the format cannot see. A development centre observes behaviour under artificial conditions over a short window. It does not observe persistence, the ability to rebuild a relationship damaged months earlier, or how someone behaves when genuinely exhausted. Treating the session as a complete picture of a person is the fastest way to discredit it — participants know what was not visible, and a report that overclaims loses the reader on the first page.
Designing the Session: Competencies, Assessors, Briefing
Design begins with the competency set, and it has to be short. A session that tries to observe everything observes nothing well, because each competency needs several independent occasions to appear before an assessor can claim to have seen it rather than glimpsed it. Definitions must be behavioural: not “leadership” but the specific observable actions the organisation means by it. Public competency frameworks such as the OPM Competencies catalogue are useful as a starting vocabulary, but a borrowed list that nobody in the organisation recognises produces feedback nobody acts on.
Assessor preparation is the second decision, and it is the one most often underfunded. Assessors in a development centre carry a double role: they must observe accurately, and they must later hold a conversation that a participant leaves motivated rather than diminished. Those are different skills. Observation training is about separating behaviour from interpretation — recording what was said and done before deciding what it meant. Feedback training is about delivering an uncomfortable observation in a way that survives being heard.
Participant briefing is the third, and the cheapest to get right. People arrive at a development centre suspicious, because the format resembles an examination and organisations have a poor record of explaining otherwise. The briefing has to state plainly what the results will and will not be used for, who sees them, and what the participant will walk away with.
Scheduling is the constraint that quietly shapes everything else. Assessors cannot observe many participants at once without the quality of the record collapsing, which sets the practical ceiling on group size and makes the ratio of assessors to participants a design decision rather than a budget line. The session also has to leave room for the integration work and the individual conversations that follow; programmes that compress the observation into a full day and then push feedback into a fortnight later reliably lose the participant’s memory of what they did, and with it most of the value.
Remote delivery changes the exercise set rather than ruling it out. Group discussion and presentation transfer to video with acceptable fidelity; anything that depends on reading a room, physical materials or informal interaction between formal exercises does not. The honest approach is to redesign for the medium, and to say which competencies the remote version can no longer observe. Where the organisation already maps team capability, showing participants how their skill matrix assessment connects to the session removes most of the residual anxiety.
Inside the Session: Observation, Integration, Feedback
During the exercises the assessors work quietly and separately. Each records behavioural evidence against the competencies without discussing it, because early discussion contaminates independent observation — once assessors know each other’s readings, they converge on them.
The discipline that makes this work is recording behaviour before interpreting it. “Interrupted the speaker, then restated their point accurately” is evidence. “Poor listener” is a conclusion wearing the clothes of an observation, and once written down it cannot be re-examined. Assessors trained to keep the two apart produce notes that a participant can recognise; assessors who conflate them produce notes that a participant can only accept or reject.
The predictable errors are worth naming because they recur. Halo lets a strong showing in the first exercise colour everything afterwards. Recency does the reverse, weighting the last thing observed. Similarity bias rewards participants whose working style resembles the assessor’s own. None of these disappear through good intentions; they are held down by structure — independent recording, evidence-first integration, and a competency definition specific enough that an assessor has to point at something.
The integration session follows. Assessors bring their evidence together, competency by competency and participant by participant, and reconcile discrepancies by returning to what was recorded rather than by deferring to whoever speaks most confidently. Disagreement here is informative rather than awkward: a person who reads as decisive to one observer and dismissive to another is usually both, depending on who they are talking to, and that is a finding worth carrying into the feedback conversation.
The feedback session is the point of the whole exercise, and everything before it is instrumentation. It is a long, individual conversation in which the assessor works through observed behaviour with the participant — describing what happened, asking what the participant was trying to achieve, and letting the gap between intention and effect emerge rather than announcing it. The conversation aims at recognition, not agreement. A participant who accepts a verdict politely has learned nothing; a participant who says “I do that, and I hadn’t noticed” has learned the thing the session exists to teach. Organisations that treat this stage as a report handover waste the entire investment, which is why the feedback conversation belongs inside a wider approach to strategic talent development rather than sitting as an isolated event.
Who Carries the Result: Participant, Manager, Organisation
Responsibility for what happens next is divided, and the division is where most programmes quietly fail.
The participant owns the plan. Nobody else can do the noticing, the practising or the asking for opportunities. The most reliable predictor of whether a development centre changes anything is whether the participant left with a concrete next action they had chosen rather than been assigned.
The manager owns the conditions. A plan that requires the participant to practise a behaviour they are never given the chance to use is a plan that dies within weeks. The manager’s job is to hand over work that demands the target behaviour, to notice attempts rather than only outcomes, and to keep the follow-up conversation on the calendar when everything else is louder. Managers who were not involved in the process tend to regard the plan as HR’s document rather than theirs, which is why involving them at the design stage pays for itself.
What happens when the manager is absent from the process is worth stating plainly, because it is the ordinary case rather than the exception. Participants return from a session with a plan their manager has not seen, into work that has not changed, and the plan becomes a document rather than an intention. Within a quarter the participant concludes the exercise was theatre, and the organisation concludes development centres do not work. Neither conclusion is about the method. The remedy is unglamorous: involve managers when the competencies are defined, brief them on what they will and will not be told, and put the first follow-up conversation in the calendar before the session runs.
The organisation owns the environment. It has to make the confidentiality promise credible, fund the development activities the plans call for, and connect development-centre findings to succession and talent processes without letting them leak into appraisal. That last balance is delicate: talent management practice, as the CIPD sets out under Talent management, depends on organisations knowing where capability sits, while the method described here depends on participants believing the session is not being used against them. Both are satisfiable, but only deliberately. Managers who take this seriously usually need their own preparation, which is what training in coaching and supporting employee development provides.
From Diagnosis to an Individual Development Plan
An individual development plan is not a list of courses. It is a small number of behavioural objectives, each with the situations in which the behaviour will be practised, the evidence that will count as progress, and a date on which it will be reviewed.
The weighting matters more than the content. Most durable development happens in the work itself — a stretch assignment, a project outside the comfort zone, a stand-in for an absent manager. A smaller portion comes through other people: coaching, mentoring, structured feedback from peers. Formal training is the smallest share by volume and the highest by leverage, because it is what makes the on-the-job practice legible instead of accidental. Plans that invert this ratio — a stack of courses and no changed work — produce certificates and no change.
Objectives must also be small enough to fail visibly. “Improve communication” cannot be reviewed; “open every project meeting by stating the decision we need to reach” can. Review conversations then have something concrete to work with, and modification becomes possible: a plan that survives unmodified for a year was almost certainly not being used.
Cadence decides the rest. Plans reviewed on a fixed rhythm — short conversations, close together, rather than an annual retrospective — keep the behaviour in working memory and let the participant abandon an objective that turned out to be the wrong one. The review is not a progress report to a supervisor; it is a working session in which the participant describes what they attempted, what happened, and what they will try differently. Managers who run it as an assessment convert the plan back into an appraisal, and the confidentiality that made the diagnosis possible quietly evaporates.
Plans die in recognisable ways. They die when they are too large to start. They die when the practice opportunity never arrives, because nobody changed the work. They die when the participant is the only person who remembers the plan exists. And they die when the objectives were written to please the assessor rather than to address something the participant actually wants to change — which is why the participant, not the assessor, has to hold the pen.
What the Organisation Gets Back
The individual benefits are the visible half. Participants report sharper self-knowledge, and self-knowledge is what converts effort into direction — people stop working on the wrong thing.
The organisational return is less obvious and more valuable. A development centre produces a behavioural picture of a population, not just of individuals: which capabilities are thin across a cohort, which are assumed present and are not, where the leadership bench actually stops. That picture is difficult to obtain any other way, because appraisal data reflects performance in the current role and says almost nothing about readiness for a different one.
Run repeatedly, the method also changes the culture around feedback. Organisations that put cohorts through development centres tend to find that feedback becomes ordinary rather than exceptional, and that people ask for it. That shift is slow, hard to attribute and worth more than any single participant’s plan. It is also the reason the method sits naturally alongside deliberate work on talent management rather than as a standalone HR product.
Frequently Asked Questions
How is a development centre different from an assessment centre?
The apparatus is similar; the purpose is not. An assessment centre supports an organisational decision about a person — selection or promotion — and its results are consequential. A development centre supports the person’s own decisions about their growth, its results are confidential to them, and no personnel decision follows from it.
Who benefits most from taking part?
People facing a change in the scope of their work: specialists moving towards management, managers preparing for a wider remit, and anyone whose role is about to change shape. The method is least useful for people whose work and level are stable, because there is no forthcoming decision for the insight to inform.
What happens to the results afterwards?
They belong to the participant. With their consent, the manager sees enough to support the plan, and the organisation may see aggregated patterns across a cohort. If results circulate beyond that, the method stops working — participants adjust their behaviour accordingly and the next cohort produces performances rather than data.
How do we know whether it worked?
Not by satisfaction scores collected on the day. The honest measures are behavioural and delayed: whether the planned behaviours appear in ordinary work, whether the plans were reviewed and modified, and whether internal readiness for wider roles improved across the cohort.