PartnerRound- 01PauseAsk for 20 to 30 seconds
- 02SplitPick one MECE axis
- 03FillTwo or three ideas per bucket
- 04PrioritiseImpact and feasibility
By the end you can
- Answer any brainstorming question with three or four MECE buckets in under 30 seconds
- Choose the right split for the question: internal and external, supply and demand, customer journey, value chain or time
- Make ideas specific enough to size, and size the big ones in a line of maths
- Prioritise ideas on an impact and feasibility matrix and say where to start
On this page
01What a brainstorm tests
Most cases include at least one open question: 'What could be causing this?', 'How could they grow?', 'What risks should they consider?'. It sounds like a test of creativity. It is mainly a test of structure.
Interviewers listen for three things: whether your ideas are organised so nothing big is missing, whether they fit this client rather than any client, and whether you can say which matter most. Twelve unconnected ideas score worse than six well-organised ones.
02The routine: pause, split, fill, prioritise
Ask for a moment. Twenty to thirty seconds is enough, and no interviewer minds. Use it to choose a split, not to think of ideas.
Pick one axis to split on, so the buckets do not overlap and together cover the whole space. Name the buckets before any ideas. It shows the interviewer the shape of your answer, and it makes gaps visible to you.
Fill each bucket with two or three specific ideas. Then say which you would pursue first and why. That last step is the one most candidates forget, and it is where the business judgement marks are.
03Five splits that cover most questions
You need a handful of splits you can reach for without thinking. Each one is MECE by construction, so it does the hard part for you.
The revenue and cost equations work as splits too. Revenue is volume × price, and volume is customers × purchases per customer. When the question is about a number, split the number. It is usually the cleanest start.
Notice that the four buckets multiply out to sales: visitors × conversion × basket size × repeat rate. That is why the split is complete. Any cause has to work through one of them.
04How to present ideas
- Signpost: say how many buckets there are and name them before going into any.
- Go bucket by bucket, two or three ideas each, one line per idea.
- Be specific: 'sell personal training as a monthly add-on' beats 'upsell'.
- Tie ideas to the case facts: if the client has spare off-peak capacity, use it.
- Stop at three or four buckets and about ten ideas, then prioritise.
Weak
They could raise prices, do more marketing, maybe offer personal training, partner with companies, open more gyms, add a cafe, sell supplements, run social media campaigns and improve the app.
Strong
I'd group ideas into three areas: more members, more revenue per member and new revenue streams. For more members: corporate memberships, and an off-peak tier to fill quiet hours. For revenue per member: a small price rise, and personal training sold as a monthly add-on. For new streams: renting studio space to physios, and paid online classes. I'd prioritise personal training, which looks like the biggest lever, and the price rise, which is the quickest.
Why: The weak list has decent ideas but no shape, so the interviewer cannot see what is missing or what matters most. The strong answer names its buckets, makes each idea specific and ends with a priority.
05Prioritise on impact and feasibility
A list of ideas is half an answer. The other half is which to do first. Score each idea on two axes: impact, meaning how far it moves the client's objective, and feasibility, meaning cost, time, risk and whether the client can actually do it.
Size impact with one line of maths: customers × uptake × price. You do not need precision. You need to know whether an idea is worth $1m or $10m.
Judge feasibility on four questions: how much upfront investment it needs, how long until it pays back, whether the client has the capability, and what could go wrong, such as cannibalising existing sales or provoking a competitor.
06Worked example: a gym chain
Worked example
Gym chain: find $10m of extra annual revenue
Prompt: 'Our client runs 50 gyms with 100,000 members, each paying $40 a month. It wants $10m more revenue a year within two years. What could it do?'
Baseline first: 100,000 × $40 × 12 = $48m a year. The target is about 21% growth, so no single small idea will get there.
Buckets and ideas
- More members: corporate memberships for local employers; an off-peak tier at a lower price to fill quiet hours.
- More revenue per member: a $2 monthly price rise; personal training sold as a monthly add-on.
- New revenue streams: renting studio space to physios; paid online classes for non-members.
Size the biggest ideas
Sense-check the risks before committing. If the price rise made 2% of members leave, it would still add about $1.4m: 98,000 × $42 × 12 is about $49.4m, against $48m today. The off-peak tier could tempt existing members to downgrade, so set its price and access hours to limit that.
Prioritise
- Quick win: the price rise. Low effort, about $2.4m, and it can start next quarter.
- Big bet: personal training. The largest at about $6m, but it needs trainers and floor space, so pilot it in five gyms first.
- Fill-ins: corporate memberships and the off-peak tier, about $3m together, worth doing if the sales effort is small.
- Drop for now: online classes and studio rental. Hard to size and small against the target.
Close: 'The four sized ideas add up to about $11m, just above the $10m target. I'd start the price rise now and pilot personal training in five gyms, since together they are worth about $8m. The main risk is members leaving, so I'd track cancellations closely after the price change.'
07Pitfalls and checks
- Generic ideas that would fit any company. Use the facts of the case.
- Overlapping buckets, such as 'marketing' and 'new customers'. Pick one axis and stick to it.
- Running dry after the first bucket. Go broad first, then deep.
- Missing a hint. If the interviewer asks 'Anything else on the cost side?', there is something else on the cost side.
Check yourself
A client's profit has fallen and you are asked for possible causes. Which opening split is cleanest?
Check yourself
An idea would add $8m of profit a year but needs a new factory that takes three years to build. Where does it sit on the impact and feasibility matrix?
Cheat sheet
The lines worth rereading the night before.
- 1A brainstorm is a structure test first and a creativity test second.
- 2Pause, split, fill, prioritise. Name the buckets before any ideas.
- 3Five splits: internal and external, supply and demand, customer journey, value chain, short and long term.
- 4When the question is about a number, split the number: volume × price, customers × frequency.
- 5Two or three specific ideas per bucket, tied to the case facts.
- 6Size the big ideas in one line of maths: you need $1m or $10m, not decimals.
- 7Prioritise on impact and feasibility, and always end by naming where to start.
Now drill it
Reading builds the model; drills make it automatic. Start with the first topic, easiest drills first.
- Growth ideas7 drills · Brainstorm ways to grow revenue in clear buckets, then say which ideas you would pursue first.
- Cost & operations ideas7 drills · Find cost levers and operational fixes in a structure the interviewer can follow, then point to where the money is.
- Risks & next steps7 drills · Name what could sink your recommendation, how to de-risk it and what you would analyse next.
Questions candidates ask
How many ideas should I give in a case interview brainstorm?
Around eight to twelve, in three or four buckets. Fewer looks thin; many more usually means you have stopped organising and started listing. Breadth across buckets matters more than depth in one.
How long can I take before answering a brainstorming question?
Twenty to thirty seconds is normal. Ask for it: 'Let me take a few seconds to organise my thoughts.' Then name your buckets before any ideas, which shows structure straight away.
What if the interviewer keeps asking for more ideas?
Treat it as a hint that a bucket is missing or thin. Check your split against the case, add the missing bucket or go deeper in the weakest one. Do not repeat earlier ideas in new words.
Is a brainstorm the same as a case framework?
Not quite. A framework structures the whole case, while a brainstorm structures one open question inside it. The same MECE thinking applies, but brainstorm buckets should be quick to name and specific to the question.