Start with four people, two recurring time slots and a two-week trial. Agree that everyone will prepare cases to give, protect the debrief and leave each session with one action to practise. That is enough structure to turn a group chat into useful preparation.
I would keep the core group small and add a few occasional partners for variety. The main challenge is usually coordination and session quality. Adding more names does little if nobody knows who is interviewing whom next Tuesday.
Recruit for compatible schedules and useful habits
State the interview language, time zone, likely meeting times and current level. Mention the firms and roles you are preparing for, but prioritise availability and willingness to give thoughtful feedback.
Turn interest into specific appointments. A group of willing candidates becomes useful when each person knows whom they will meet, when, and what they need to prepare.
A useful invitation is: “We're forming a four-person group for two reciprocal sessions each week, Tuesday and Saturday at 19:00 IST. Each person prepares a case to give and records one repair after receiving feedback. Would you like to try the first two weeks?”
If your schedules do not overlap, keep the person as an occasional partner. A group built around a time nobody can consistently attend will spend its energy rescheduling.
Agree how each meeting works
Use a short agreement rather than a long rulebook. Confirm the appointment by the previous evening, read and solve the case beforehand, keep solutions hidden and give notice of cancellations.
Choose whether the session is for learning or simulation. Beginners can pause and discuss a difficult step. For a realistic mock, save coaching for the debrief. Mixing those modes without agreement makes feedback hard to interpret.
Reserve enough time for the material. A 90-minute reciprocal meeting can accommodate two shorter cases and their reviews. If each full case needs 40 minutes, extend the appointment or split giving and receiving across different days. Do not let the second person's feedback disappear because the first case ran long.
Quick Answer Repair Drill
In a practice mock, the interviewer plays a ferry operator's finance director who wants to understand this year's weaker results.
- Annual profit
- down 15% on last year
- Passengers carried
- flat on last year
- Candidate's opening
- revenue minus cost, starting with passenger volume
What is the main weakness in the candidate's opening plan?
Rotate the pairs and who goes first
With four members, this three-session rotation gives everyone a meeting with each of the others:
| Session | First pair | Second pair |
|---|---|---|
| 1 | A and B | C and D |
| 2 | A and C | B and D |
| 3 | A and D | B and C |
Alternate who receives a case first. Familiarity with one partner can hide unclear explanations, while always going second can become a comfortable warm-up routine. Rotation gives the group a wider view of each person's performance.
With three people, use candidate, interviewer and observer roles. Give the observer a clear job, such as recording where the calculation became difficult to follow. They should not become a second interviewer introducing unrelated questions.
Make interviewer preparation part of the commitment
Before the meeting, the interviewer should know the objective, the information available on request, the calculations and the intended decision. They should also recognise plausible alternative approaches.
Know the case well enough to give it without continually searching the answer sheet. That preparation lets you listen to your partner's reasoning.
If the case contains an error or missing information, resolve it before reuse. During a session, state the missing assumption clearly and continue where possible. Do not penalise a candidate for failing to infer information the material never supplied.
Keep a case list and a correction log
The case list records who has seen each prompt and solution. That prevents someone accidentally receiving a “fresh” mock they watched last week. Include the source and the member who has prepared it as interviewer.
The correction log is personal. Record the observed mistake, why it mattered, the repair and the next test. One line is enough: “Used a simple average for margins despite different revenues; practise weighting by revenue; retest on a new product-mix case.”
Keep those records separate. The group's inventory helps you coordinate; your correction log helps you learn. There is little benefit in turning everyone's scores into a public league table.
Quick Case Math Drill
An independent bike shop is applying for a bank loan, and the lender asks for last year's results by product line.
- Lender's minimum
- 15% overall profit margin
| Product line | Revenue (£) | Profit margin |
|---|---|---|
| Bike sales | 80,000 | 10% |
| Servicing | 20,000 | 40% |
What was the shop's overall profit margin across both product lines last year?
Plan for cancellations before they happen
If one person cancels, the remaining three can run a case with an observer and rotate the favour next time. If nobody else is available, use the reserved block for a specific spoken drill and its review.
Address recurring problems directly: “Could you read the case before the next session and keep ten minutes for feedback?” If the arrangement still does not work, change the roster. Someone whose interview is over or whose deadline differs may fit better as an occasional participant.
Review the group after two weeks
Ask whether meetings happened, interviewers prepared and feedback changed subsequent attempts. Keep what works and simplify what does not. You may need fewer sessions with better preparation, or occasional input from a more experienced reviewer.
Use the first-partner guide for new members and the mistake-repair guide when the group keeps observing the same problem. A dependable group makes it easier to keep practising and harder to leave useful feedback unused.
Frequently asked questions
Setting up the group
How big should a case practice group be?
Start small, with four people, two recurring time slots and a two-week trial. Keep a small core group and add a few occasional partners for variety. The usual challenge is coordination and session quality, so adding more names does little if nobody is actually meeting or preparing properly.
What should I look for when recruiting members?
Prioritise availability and willingness to give thoughtful feedback over matching target firms. When inviting people, state the interview language, time zone, likely meeting times and current level, and mention the firms and roles you are preparing for. Compatible schedules largely decide whether sessions happen at all.
How do I stop our group chat from going quiet?
Turn interest into specific appointments. A group of willing candidates becomes useful when each person knows whom they will meet, when and what they need to prepare. Put named pairings in the calendar from the outset rather than waiting for someone to volunteer for each session.
Do we need formal rules for the group?
A short agreement works better than a long rulebook. Confirm appointments by the previous evening, read and solve the case beforehand, keep solutions hidden and give notice of cancellations. Those few commitments cover the basics of a useful session without making the group feel bureaucratic.
Running sessions well
Should our sessions be learning sessions or realistic mocks?
Decide before each session and make sure both people know. In learning mode, beginners can pause and discuss a difficult step. In a realistic mock, the interviewer saves coaching for the debrief. Mixing the two without agreement makes feedback hard to interpret, because nobody is sure which help was intended.
How should four people rotate partners?
Use a three-session rotation: A with B and C with D, then A with C and B with D, then A with D and B with C. After three sessions everyone has practised with each of the others. Within each pair, also alternate who goes first.
How much should the interviewer prepare?
Enough to give the case without constantly searching the answer sheet. Know the objective, the information available on request, the calculations, the intended decision and plausible alternative approaches. That preparation lets you listen to your partner's reasoning, and fix any errors or missing information in the case before it is reused.
How do we avoid giving someone a case they've already seen?
Keep a shared case list recording who has seen each prompt and solution, plus the source and which member has prepared it as interviewer. Without it, someone can easily receive a supposedly fresh mock they watched the week before, which wastes a realistic attempt.
Feedback that changes the next attempt
What is a correction log and how detailed should it be?
It is a personal record of mistakes and repairs. Each entry notes the observed mistake, why it mattered, the repair and the next test, for example using a simple average for margins despite different revenues. One line per mistake is enough, as the aim is checking whether your next attempt actually changes.
How do we make sure everyone means the same thing by good feedback?
Once a week, give everyone the same short answer to review and have each person write a diagnosis before discussing it. Differences in what members call good maths or clear communication then become visible, and the group can agree a shared standard rather than giving inconsistent feedback.
What's wrong with averaging the two margins in the calibration example?
It ignores the products' different sizes. Product A makes £10,000 profit on £100,000 revenue and Product B £120,000 on £400,000, so the combined margin is £130,000 on £500,000, or 26%, not 20%. Margins must be weighted by revenue, which a simple average fails to do.
Keeping the group on track
What should we do when someone cancels?
Plan for it in advance. If one person cancels, the remaining three can run a case with an observer and rotate the favour next time. If nobody else is available, use the reserved block for a specific spoken drill and review it, so the time still produces practice.
How do we handle a member who is repeatedly unprepared?
Raise it directly and specifically, for example by asking them to read the case before the next session and keep ten minutes for feedback. If the arrangement still does not work after that, change it rather than letting one recurring problem weaken everyone's sessions.
How do we know whether the group is working?
After two weeks, ask whether meetings happened, whether interviewers prepared and whether feedback changed later attempts. Keep what works and simplify what does not. You may need fewer sessions with better preparation, or occasional input from a more experienced reviewer.