1. Pause and clarify
Restate the goal, ask for a moment, think in silence.
- Ask
- "Short term or long term?"
- Pause
- About a minute, in our experience
Case interview brainstorming, buckets before ideas
A brainstorming question asks for many ideas at once: ways to grow revenue, causes of a problem, risks of a plan. The strong answer takes a short pause, splits the question into two to four buckets that don't overlap, fills each with specific ideas and ends by naming the one or two worth doing first. In our experience interviewers weigh the structure and the judgment, not the length of the list. Soreno, an AI voice interview practice app for consulting, investment banking and other finance candidates, runs a free growth strategy case with no sign-up, where you list and rank ways to grow, and full mock cases with scored feedback.
Updated
- The method
- Pause, bucket, fill, rank
- Ours, built on what firms say they look for
- Buckets
- 2 to 4, no overlaps
- MECE: nothing counted twice, nothing big missing
- Worked case
- 4 ideas sized, 2 picked
- Invented by us, every number shown
- Free practice
- No sign-up
- A full growth strategy case with a timer
How do you brainstorm in a case interview?
A brainstorming question usually comes mid-case, after the opening structure: "what could explain this?", "what else could they do?", "what would worry customers?". It is narrower than the opening and wants ideas, not a plan for the whole case. Our method has four steps, and the third card is where most answers fall short.
2. Name the buckets
Two to four categories that split the question with no overlap.
- Say
- "Three areas: ..."
- Test
- Could an idea fit two buckets?
3. Fill with specifics
Two or three concrete ideas per bucket, not labels.
- Weak
- "Marketing"
- Strong
- "Text past customers after rain"
4. Rank and pick
Say which one or two you would test first, and why.
- By
- Size, cost, speed, risk
- Then
- Tie it back to the case goal
Worked example: Brightline Car Wash (ours)6 steps
Brightline Car Wash, an invented chain of 40 car washes, wants to grow revenue 15% next year. The interviewer asks: what are all the ways it could do that? This case and every number in it are invented by Soreno for practice.
Size the goal
$2.7 million of new revenue
200,000 customers wash 6 times a year at $15, 1,200,000 washes and $18 million of revenue. 15% more is $2.7 million. Knowing the gap tells you later which ideas are big enough to matter.
Bucket
Customers, visits, spend
Revenue is customers times washes per customer times spend per wash. Those three factors are the buckets: they don't overlap and together they cover all revenue. Say them before any idea.
MECE, with examplesFill
Specific ideas in each
More customers: fleet deals with local businesses, a site near the new office park, a ride-share driver discount. More washes per customer: an unlimited membership, text reminders after rain, a loyalty card. More spend per wash: a wax add-on, interior cleaning, a premium tier. Nine ideas, each one a thing the client could start.
Size the top ideas
Four ideas, four numbers
Membership: 5% of customers join at $30 a month, 10,000 members paying $3.6 million a year, minus the $900,000 they spent before, net $2.7 million. Wax: 20% of 1,200,000 washes add $5, $1.2 million. Fleet: 50 businesses with 20 vehicles washed 24 times at $10, $240,000. Rain reminders: 60,000 customers wash 0.25 times more, $225,000.
Rank
Unlimited membership and wax add-on first
In order: unlimited membership $2.7 million, wax add-on $1.2 million, fleet deals $240,000, rain reminders $225,000. Membership alone is worth 100% of the goal, and the top two together reach $3.9 million, 144% of it. Fleet deals and reminders are cheap, but too small to lead. One check before trusting the membership number: the first to join an unlimited plan are usually the frequent washers, so the pay-per-wash revenue it replaces is likely more than $900,000.
Say it
Strong answer in about a minute
"I'd split revenue into three buckets: more customers, more washes per customer and more spend per wash. For customers: fleet deals, a new site, a driver discount. For washes: an unlimited membership, rain reminders, a loyalty card. For spend: a wax add-on, interiors, a premium tier. If I had to pick, I'd start with the membership and the wax add-on. Rough numbers say the two bring in about $3.9 million, more than the $2.7 million goal, and both run on the sites we already have. The risk is members washing so often that the queue puts off pay-per-wash customers, so I'd test it at a few sites first."
Weak vs strong: the same questionOurs
Both answers to the Brightline prompt, written by us. The ideas overlap a lot. The structure doesn't.
Weak
A list in the order it came to mind
"They could do more marketing, maybe social media, or discounts, or memberships, or open more locations, or partner with gas stations, or improve the customer experience, or offer more services." Eight ideas, no buckets, no pick. Marketing and social media overlap, and customer experience is a label, not an idea.
Strong
Buckets, specifics, a pick
The answer in the worked example above: three named buckets that add up to revenue, three concrete ideas in each, then two picks with a rough size and one risk. The interviewer can follow it, push on any bucket and see the judgment at the end.
Bucket sets by question type (ours)6 types
Starting splits we use in practice. Adapt the words to the client. A memorized list read out word for word is easy to spot.
Grow revenue
Customers, frequency, spend
Or existing customers vs new customers, then products and channels inside each.
Cut costs
Fixed vs variable
Or by step of the value chain: buy, make, move, sell, support.
Why did a metric fall?
Internal vs external
Inside: the product, price, operations, people. Outside: customers, competitors, the market, regulation.
Risks of a plan
Financial, operational, market, people
Or by stakeholder: customers, employees, partners, regulators.
Customer reaction
The customer's journey
Before buying, while buying, after buying. McKinsey's Beautify practice case asks this kind of question: what would make a customer switch to a mostly virtual sales experience?
What to investigate
Should we, can we, how
Is it attractive, can the client do it, and what it takes. McKinsey's Diconsa case opens with a question like this.
When you run dry, or hear "what else?"4 moves
In our experience "what else?" usually means a bucket is missing, not that a bucket needs more ideas.
Check the split
Look for the missing bucket
If all your ideas are inside the company, add the outside. If all are about price, add volume.
Flip it
Ask the opposite
"What would make revenue fall?" Each answer reversed is an idea to grow it.
Borrow
Use another industry
Gyms sell memberships, airlines sell add-ons, coffee shops sell loyalty. Ask which of those fits this client.
Walk it
Follow the customer or the product
Step through the customer's day or the product's path from supplier to buyer and ask what could change at each step.
What do interviewers listen for in a brainstorm?
Five things we listen for in practice cases. A long list hits none of them by length alone.
Buckets said first
"Three areas" before any idea lets the interviewer follow and spot a gap.
No overlaps
"Retention" and "more customers" overlap, since keeping a customer adds to the total. "Existing" and "new" customers don't.
Ideas the client could start
"Sell fleet washes to local plumbers" beats "partnerships". If an idea is a single word, it is probably a label.
A pick with a reason
End on the one or two you would do first, with a rough size or a reason. That is the judgment part.
Calm on "what else?"
Add a bucket, not filler. Saying how you found the new area shows the structure is working.
7 brainstorming questions to practice
Each asks for many ideas. Under each is what a strong answer does.
Grow
Ways to make more money.
A movie theater wants to grow revenue. What could it do?
Splits revenue into tickets and everything else, then visitors and spend inside each, and picks the idea that uses empty seats on weekdays. Practice it, see a model answer
A gym chain wants to double its revenue in five years. How would you approach it?
Buckets by more members, more spend per member and new formats, sizes each against the doubling goal and drops the ones too small to matter. Practice it, see a model answer
A consumer goods company is thinking about a premium pet food line. What should it consider?
Uses should we, can we and how as the buckets, with specific questions under each, then names the deciding one. Practice it, see a model answer
Cut or fix
Causes and cures.
A retailer needs to cut costs by 10%. Where could the savings come from?
Splits costs fixed and variable, or along the value chain, lists concrete cuts and checks they add up to 10%. Practice it, see a model answer
A social app's likes dropped 10% overnight. What could have caused it?
Starts with internal vs external, then names checks like a release, a bug in counting or an outage before any user-behavior idea. Practice it, see a model answer
A streaming service's churn has doubled. What could explain it?
Buckets by product, price and competition, with a specific cause in each, then says which one the data should test first. Practice it, see a model answer
Public sector
No profit, same method.
A city wants to cut traffic deaths in half. What could it do?
Buckets by road, vehicle and behavior, with specific levers in each, then ranks them by lives saved per dollar. Practice it, see a model answer
Brainstorming mistakes we hear most
From practice cases. BU's deck names the overlap one too.
Talking before thinking
Starting to list at once usually means no buckets. Ask for a moment. McKinsey's practice case hints tell you to take time to organize your thoughts.
Overlapping buckets
"Marketing" and "customer acquisition" will share half their ideas. Pick splits that can't.
Labels instead of ideas
"Operations", "technology" and "partnerships" are buckets, not answers. Each needs something the client could do on Monday.
Reading a memorized list
The same twelve ideas for every client sound recited. Tie each idea to something you learned in this case.
No pick at the end
A brainstorm with no priority leaves the interviewer to do the judgment for you. Close with the first one or two.
What firms and career offices publish
The firm pages below don't use the word brainstorming. They and two career offices do say what the answer to an open question should show: an organized approach, creativity and judgment over a single right answer. Read on October 31, 2026.
McKinsey
- Hint
- Take time to organize your thoughts
- Also
- Overall approach before details
McKinsey's Diconsa practice case asks what the team should investigate, and its hints say to take time to organize your thoughts and develop an overall approach before diving into details. Its Beautify case asks what factors the client should consider, and what would make a customer switch to a virtual experience.
BCG
- Do
- Think structurally
- Don't
- Rush your response
BCG says you are expected to think logically and show both business intuition and creativity, and that there isn't always a single right answer. Its list of things not to do includes rushing your response and overcomplicating solutions.
Bain
- Rewards
- Precision and creativity
Bain says its case interviews reward precision but also creativity, that there isn't necessarily a right answer, and asks candidates to show they can think in a logical and structured way.
Harvard and Boston University career offices
- Harvard
- Organize your thoughts first
- BU
- Exhaustive, no overlaps
Harvard's career office says not to dive headfirst into immediate solutions, because interviewers want the analysis. BU's 2021 workshop deck, citing MyConsultingCoach, warns that in a "what else?" question your list should be exhaustive and without overlaps.
Bucket it, fill it, pick one, out loud
Soreno's AI interviewer runs a full case, asks "what else?" when your list runs short, checks your buckets for overlap and scores the session. Start your free week.
Questions people ask
What is a brainstorming question in a case interview?
An open question that asks for many ideas at once, usually mid-case: ways to grow, ways to cut costs, possible causes, risks or customer reactions. The McKinsey, BCG and Bain pages we read don't use the word, but McKinsey's practice cases include questions of this kind.
How many ideas should I give?
No firm publishes a number. In our experience two to four buckets with two or three specific ideas each, so roughly six to twelve, is enough when it ends with a clear pick.
Can I ask for time to think?
Yes. McKinsey's practice case hints say to take time to organize your thoughts before answering. In our experience a pause of about a minute is normal.
How is brainstorming different from the opening framework?
The opening structure plans the whole case. A brainstorm answers one narrower question inside it and wants concrete ideas, not areas to analyze.
What if I run out of ideas?
Look for a missing bucket, flip the question (what would make it worse?), borrow from another industry or walk through the customer's journey. Adding a bucket beats padding one.
Do I need industry knowledge to brainstorm well?
It helps, but structure carries most of the answer. BCG and Bain say there isn't always a single right answer, so clear buckets and a reasoned pick matter more than knowing the industry.
Should brainstorming ideas be MECE?
The buckets should be: no overlap and nothing big missing. BU's workshop deck makes the same point about "what else?" questions. The ideas inside a bucket just need to be specific.
Do I have to prioritize at the end?
We would. Ending on the one or two ideas to start with, and why, turns a list into a recommendation and shows judgment.