What the day is really measuring
Centres exist so an employer can compare dozens of candidates for a few places on level terms. Trained assessors observe you across the exercises and mark named behaviours - communication, analysis, collaboration, resilience, commercial awareness. No column on their sheet says loudest person at the table, and treating volume as strategy is the most common way to waste the day.
The scoring is a grid. Every exercise feeds two or three competencies, each scored on a scale, and the offer goes to the totals rather than to any single moment. Once you see that, you stop performing and start supplying evidence. Start with the job pack: its competency list is the marking scheme in plain sight, and rehearsing one example per competency shows quickly that the hard part is trimming, not remembering.
The shape of a typical day
A full UK centre runs roughly 9am to 4pm on a rhythm that varies little between employers. This is the sequence candidates report most often.
- 09:00 - Arrival and briefing. Informal in tone; observed from the first handshake.
- 09:30 - Introductions. A minute on yourself, which you will repeat several times before lunch.
- 10:00 - Group exercise. Six of you, one brief, 30 to 45 minutes, assessors writing throughout.
- 11:15 - Tests. Numerical, verbal and logical reasoning, or a role-specific task.
- 12:15 - Lunch with staff. Rarely on the score sheet; never entirely off it.
- 13:00 - Presentation. Either prepared in advance or built on the day in 20 to 30 minutes.
- 14:00 - Competency interview. Around 45 minutes, usually a line manager plus HR.
- 15:15 - Case study or in-tray. An inbox, a budget and a clock - prioritisation under pressure.
- 16:00 - Wrap-up and questions. Ask something researched; a vague question suggests vague interest.
Go through the invitation line by line and mark every named exercise, because employers disclose more than candidates expect. A sentence like “you will give a five-minute talk on a business challenge” is an instruction about what to rehearse. Give the brief a live run in Sharp AI Interview and the day contains one fewer first attempt.
Group exercises: what earns ticks and what does not
Six candidates share a brief - split a charity budget, cut eight bidders to three - and the instinct is that the biggest talker wins. Assessors are trained against that instinct. Their sheet asks whether the group reached a decision and what you contributed to getting it there.
Four behaviours that reliably earn marks
- Opening the door for others. “Priya has been closest to the numbers and we have not heard from her.”
- Extending someone's point. “Building on Tom's cost argument” is proof of listening, the scarcest behaviour in the room.
- Holding the clock. “Twelve minutes left and requirement two is untouched” is leadership at conversational volume.
- Conceding visibly. “That case is stronger than mine” goes down as flexibility, not surrender.
Three habits that quietly leak marks
Interrupting is the obvious one, and completing another candidate's sentence costs almost as much. Falling silent because the discussion accelerated is the second, since unspoken contributions cannot be scored. The third is claiming the chair in the first half-minute. Everyone defaults to one of the three under pressure; a single mock round tells you which is yours.
Three sentences do most of the heavy lifting. “Shall we agree a structure before we start?” opens the discussion. “Can I check we all read requirement two the same way?” recovers a drifting one. “Five minutes left - shall we settle the recommendation?” closes it. Say each aloud a few times beforehand so they remain available at a raised pulse.
If speaking up at all is the barrier, the fix is rehearsal volume rather than resolve - see handling interview nerves for the physiology. Arriving with a prepared opening line is what gets you into the discussion inside the first ninety seconds, which is why the first line deserves the most repetitions.
The presentation: structure carries it
Five minutes of speech is around 650 words, which caps you at three points. Give the recommendation inside the first fifteen seconds, support it with three reasons that each carry a number, name the largest risk, then take questions. The marks fall on clarity, evidence and timing - and timing is the easiest of the three to bank in advance.
Rehearse standing, out loud, twice, with a clock in view; silent reading reveals nothing about pace. If the brief lands on the day, spend the first five of your thirty minutes choosing the recommendation and the rest supporting it. A second run of the same material shows exactly where the sentences bolt.
The questions afterwards are scored with the talk, not after it. When you lack an answer, describe how you would get it and when: “I would pull two quarters of returns data and confirm by tomorrow” is complete. Drill that reply until it comes without hesitation, because hesitation there is read as exposure.
The timed tests
Numerical papers run around 20 to 25 questions in 25 minutes - school-level arithmetic at an uncomfortable tempo, dominated by percentage change and ratios. Verbal reasoning asks true, false or cannot say from the passage alone, and the trap is importing what you already know. Practise from the passage, nothing else.
Where the invitation names the publisher, practise that publisher - SHL, Cappfinity, Talent Q and Korn Ferry each have a distinct feel, and three timed sets teach more than forty leisurely ones. Situational judgement tests score you against the employer's published values, not your boldest instincts, so read the values page the night before. Between test blocks, sixty seconds of spoken warm-up shifts the brain back from numbers to language before the interview.
The interview inside the day
Forty-five minutes, five or six questions, most beginning “tell me about a time when”. Answer in STAR and let Action dominate: roughly a fifth on situation, a tenth on task, half on action, the rest on result. Scene-setting is the drift to watch for, and it is precisely what a practice partner pulls you back from.
Six stories serve better than twenty. A strong one bends - the rescued deadline answers resilience, planning, influencing or ambiguity depending on which part you lean on. Test each story against four different question stems and you learn whether it genuinely bends or whether you are re-telling it in a different order.
Probing is standard. “What would you change next time?” is coming, and an honest reply is often the difference between the top two bands. If the structure itself is wobbly, work through the STAR interview technique first, then have Sharp AI Interview fire the probes at random until they stop landing as surprises.
Case studies and the in-tray
The in-tray gives you eighteen emails, a diary, a budget and forty minutes, and it is not measuring inbox speed. It is measuring ranking. Deal with the top three items, delegate two with a named owner and a deadline, and spend one line explaining why the remainder waits. Case studies reward the same move: speak your assumptions - “I am taking 12 per cent churn as holding through Q4” - because reasoning left unsaid earns nothing. Practising the narration aloud beforehand makes it fluent when it counts.
Lunch, corridors and the unscored hours
The formal marking pauses between exercises; the impressions do not. Assessors compare notes in corridors, and a line like “she asked our graduate about the rotation and used his name” finds its way into the wash-up meeting. Posture signals engagement before any words do - twenty minutes with the guide to confident body language covers what the room sees. Recording your sixty-second introduction and watching it back is uncomfortable in exactly the useful way.
Remote assessment centres
Roughly half of UK centres now run online, with the same exercises and some added failure modes. Breakout rooms eat their first half-minute, so speak early and attach your name: “Ali here - shall we set a running order?” Video also flattens presence, which means verbal signposting has to do what posture normally would. Rehearsing on the identical laptop, microphone and clock is the closest available simulation of that.
Check the platform the day before. Light in front, camera at eye level, notifications silenced. If the connection fails, rejoin and summarise what you missed in one sentence - recovery is a scored behaviour, and assessors have watched it happen many times.
Seven days out
Spread over a week this plan is light. Compressed into the final evening it is impossible.
- Day 7: Extract the competency list from the job pack; attach one true story to each entry.
- Day 6: Study the employer - the latest annual report, three recent announcements, your target division.
- Day 5: Two timed test sets, marked honestly, with the wrong answers read back.
- Day 4: Six STAR stories spoken against ten question stems, not six.
- Day 3: Build and time a five-minute talk on anything; the structure transfers even when the topic does not.
- Day 2: Two mock group discussions, plus three written lines for inviting quieter voices in.
- Day 1: Logistics only - route, clothes, documents, early night. Seven hours of decisions need a rested brain.
If a full week is not available, the checklist in preparing for an interview condenses into two evenings. The one step that never condenses is speaking aloud - no candidate has become fluent in the STAR method by reading about it, and that hour is the one Sharp AI Interview exists to provide.
Questions candidates ask about assessment centres
How long does an assessment centre last?
The standard UK version fills a working day, about 9am to 4pm; half-day and remote editions of three to four hours are increasingly common. Expect a group exercise, a solo presentation, timed tests, a competency interview and a lunch with current staff. The timetable is usually shared beforehand, so build your preparation around the exercises your own invitation lists.
What do assessors look for in a group exercise?
Defined behaviours, not volume. Marks go to drawing quieter candidates in, building visibly on another person's idea, watching the time and the brief, and conceding when a better argument appears. Marks come off for interrupting, dominating and prolonged silence. Because the group must land a decision, moving it toward one scores more than winning any individual point.
How do I stand out at an assessment centre?
Through specificity rather than volume. Attach numbers to your examples, say the assumptions under your recommendations out loud, and credit other candidates by name when you extend their ideas. Bring one researched question about the team you would join. The scoring is a competency matrix, and the candidate who feeds it measurable evidence in every exercise beats the loudest voice.
Can you fail one exercise and still get the job?
Yes. Each exercise is marked on its own and the totals decide, so a flat group discussion can be outweighed by a strong presentation or interview. The competency interview often carries the heaviest weighting. A visible recovery after a poor exercise even works in your favour, since resilience is usually one of the behaviours on the matrix.