The 3Cs, the 4Ps, Porter’s five forces, the BCG matrix: most candidates arrive at a case interview with a toolbox learnt from preparation guides. The problem is that the interviewer knows it too. When a candidate announces “I am going to analyse this problem with the 3Cs”, the partner across the table mostly hears that they are reciting. Firms say so themselves on their careers websites: they prefer a structure built for the problem at hand to a generic framework. This article explains where these frameworks come from, why they penalise candidates, and how to build a case-specific structure in two minutes, with a before and after on the same prompt.
Where frameworks come from and why they bore interviewers
The classic frameworks were not born for interviews. The five forces come from strategy research, the marketing mix from marketing textbooks, the growth-share matrix from a consulting firm’s work in the 1970s. They were taken up by preparation guides, then turned into grids to roll out: for a market entry case, this list; for a merger case, that one.
This mechanism raises three problems.
It shows immediately. An interviewer runs dozens of interviews per season. They recognise the guidebook grids from the first sentence, and they know the candidate is not yet thinking about the problem: they are putting the problem in a box.
It produces structures that are too broad. A generic framework covers everything that might be relevant to a family of problems. Applied to a specific case, it contains useless branches and often misses the question that really matters.
It does not prove the skill being tested. What the firm tests is your ability to break down a problem you have never seen, as you will do with a client. McKinsey puts it on its interviewing page 🔗: there is not one logical solution, and the candidate should listen carefully and share tailored solutions rather than standard frameworks. BCG goes in the same direction on its recruitment process page 🔗, warning that standardised concepts learnt during studies or interview preparation may seem relevant but may not stand up to closer scrutiny.
What a partner calls a tailored structure
A “tailored” structure is not a structure invented from scratch. It is a breakdown that starts from the question asked, uses the vocabulary of the client and its industry, and is limited to the branches that can change the answer.
It can be recognised by four characteristics:
- it restates the client’s objective as a decision criterion (“the client wants to know whether this acquisition creates value within three years”);
- its branches are specific: you can read the industry in them, not only abstract categories (“volume of patients referred by community doctors” rather than “customers”);
- it is prioritised: the candidate says where they start and why;
- it is MECE: the branches do not overlap and, together, cover the problem.
A partner will never blame you for having a “market” branch in a market entry case. They will blame you for having the same “market” branch, worded the same way, in all your cases.
Start from the question, not the textbook
The safest method is to work in three steps before drawing anything.
1. Restate the question and the objective. Repeat the prompt in your own words and ask what the client wants to achieve: a level of profitability, an investment decision, a market share, a deadline. This step seems scholarly; it is what prevents you from solving the wrong problem.
2. Ask two or three clarifying questions. They concern what changes the structure: the geographic scope, the client’s business model, the time horizon. “Does the client sell to restaurants or to households?” completely changes the breakdown of a growth case.
3. Ask yourself what must be true for the answer to be yes. This is the most useful reflex. For a case “should the client launch this product?”, the answer is yes if the market is large enough, if the client can take a sufficient position in it, if the product is profitable at that volume and if the launch is feasible with its resources. These four conditions already form the branches of your tree, and they are specific to the case.
Building a MECE tree in two minutes
Once the conditions are set, the tree is built by going down one or two levels, no more.
- First level: the three or four conditions or levers that answer the question.
- Second level: for each branch, the two or three concrete analyses you would carry out, worded in the language of the industry.
- Priority: one sentence saying which branch you start with, and why it is the one most likely to settle the question.
To check that the tree is MECE, ask yourself two questions. Could the same analysis fit in two branches? If so, there is overlap. Does an important factor find no place anywhere? If so, there is a gap. Fix it before presenting.
Silent thinking time is generally accepted if you announce it: “I will take a minute to structure.” Draw the tree legibly on your sheet, turn it towards the interviewer if the interview is in person, and present it in under a minute.
Before and after on the same prompt
Take the following prompt: “Our client is a regional network of medical testing laboratories. It is considering buying a competing network in a neighbouring region. Should it do so?”
Before, the memorised structure:
“I will structure this with a four-part acquisition framework: the market, the target, synergies and risks. For the market, I will look at size, growth and competition. For the target, its profitability. For synergies, revenue and costs. And for risks, regulatory aspects and integration.”
Nothing is wrong. But the interviewer learns nothing about your understanding of the case: the same sentence would apply to buying a chain of bakeries.
After, the structure built for the case:
“The objective is to know whether this acquisition creates value for the client at a reasonable price. For the answer to be yes, four things must be true. One: the target’s business must be solid, meaning a stable volume of tests, which depends mainly on prescriptions from community doctors and healthcare facilities in the region, and a controlled exposure to changes in the pricing of tests. Two: combining the networks must produce real savings, first on the technical platforms, which could be consolidated, then on the logistics of samples between sites and on reagent purchasing. Three: the asking price must remain below the value of the target plus these synergies. Four: the deal must be feasible, particularly with regard to the rules on laboratory ownership and the ability to retain the biologists. I would start with the technical platform synergies, because that is probably what justifies buying in a neighbouring region rather than growing organically.”
The second version contains the same categories as the first, but each branch is worded for this industry and this decision, and the candidate says where they start. That is what the interviewer calls a structure.
Keeping a few reflexes without reciting
Giving up recited frameworks does not mean reinventing everything for each case. Some reflexes remain useful, provided you adapt them:
- profit = revenue − costs for any profitability problem, as a starting point rather than a complete answer, as our article on the profitability case explains in detail;
- supply and demand for market sizing or a pricing question;
- internal and external to separate what depends on the client from what depends on its environment;
- short term and long term for any recommendation that commits investment.
These reflexes are checking tools: once your tree is built, they let you make sure that no major dimension has been forgotten. They do not replace thinking about the problem.
Experienced candidates often have an advantage here: they know an industry and naturally word specific branches. Our article on moving into consulting mid-career explains how to make the most of it in interview.
This demand for structure also applies to your CV: a firm’s recruiter expects bullets that state a problem, a method and a result, not a list of tasks. The strategy consultant CV template gives a concrete example. And so that form does not cost you preparation time, the ConseilCV editor takes care of the layout and produces a PDF that firms’ screening software can read.
Frequently asked questions
Should you still learn the classic frameworks?
Yes, as general knowledge and as a checklist. Knowing them helps you forget nothing; reciting them penalises you. The difference lies in how you word the branches.
How many branches should a good structure have?
Three or four at the first level, rarely more. Beyond that, the structure becomes hard to present and the interviewer doubts your ability to prioritise.
What if the interviewer explicitly asks for a framework?
That is rare. If they do, propose your structure while explaining how it relates to the framework they mention. Most of the time, they want to know whether you understand the logic, not whether you know the name.
How can you practise building tailored structures?
By starting from real prompts, for example firms’ sample cases or business news, and forcing yourself to word each branch with the vocabulary of the industry. Having your structures reviewed by a partner who plays the interviewer is the fastest way to progress.
Sources
- McKinsey & Company, Interviewing at McKinsey 🔗
- Boston Consulting Group, Bewerbung und Interviews 🔗
- Boston Consulting Group, Case interview preparation 🔗