Simon-Kucher Case Interview 2026: Where AI Skill Counts

What the firm says about AI, and what McKinsey says

The wording sits on the firm's becoming a consultant page, inside the passage on analytical confidence:

as AI increasingly becomes an important part of consulting, we also value your ability to work effectively with AI-supported tools. Rather than replacing analytical thinking, these technologies make it even more important to ask the right questions, critically evaluate outputs, and translate information into practical recommendations for clients.

Read what it asks for. Asking the right question. Judging an output that might be wrong. Turning the result into something a client can act on. Those are three skills, and none of them is prompting.

Now the other side. McKinsey's interviewing page carries a table of encouraged and prohibited uses. Polishing your resume is encouraged. Generating answers in real time during an interview is prohibited, as is using AI during an assessment without explicit permission. The page closes the point by saying it wants to get to know the candidate rather than the candidate's AI tool.

Here is where a reader can go wrong, so it is worth being blunt. Simon-Kucher has not given anyone permission to run a model during a case. It has described a professional skill it wants you to already have. Nothing on its pages says the skill is tested, observed during an exercise, or asked about out loud. That absence is real and we are not filling it.

What you can do with it is concrete. If your own work has involved checking a generated output and finding it wrong, that is a story this firm has told you it cares about. Say what the tool produced, say how you tested it, say what you changed. That answer belongs in the behavioral half, and the Simon-Kucher behavioral questions page covers the shape of those.

The five skills it says it assesses every time

The methods vary. The firm says so plainly: behavioral interviews, business case interviews, and for some positions cognitive assessments including numerical and reasoning tests, chosen by role and level. What does not vary is the list of skills it says it looks for consistently.

Analytical confidence is the entry worth pausing on, because the firm's own definition of it includes the critical evaluation of AI-generated output. Confidence here means being willing to say a figure looks implausible, and then showing the check that made you say so.

The firm describes its exercises as reflecting the kind of thinking consulting uses: identifying patterns, structuring data, making well-founded decisions. That phrasing tells you what the cognitive tests are for. They are not trivia. They are the same three moves under a clock.

One line about experienced hiring, since it changes who this page is for. The firm writes that for most experienced hire roles it typically looks for candidates with prior consulting experience. If you are coming from industry, that sentence is the one to plan around, and a pricing or commercial background is the closest adjacent thing you can bring.

Frameworks: two firms, two published instructions

Michaela Scherhag, who leads junior recruiting in Frankfurt, is quoted on the firm's page saying that "a solid understanding of economic principles and frameworks can also be beneficial, especially in our core focus areas such as pricing, marketing, digitalization, and sales."

Kearney publishes the reverse. Its case page tells candidates to avoid forcing a solution into a standard framework, and names the only three things a solution has to fit. We wrote that page up in full in the Kearney guide.

Two firms, two published instructions, pointing opposite ways. The resolution is in the word Scherhag used, which is principles.

A framework is a named box diagram someone else drew for a category of problem. An economic principle is a mechanism that holds whether or not anyone drew it: demand falls as price rises, and how fast is an empirical question. Fixed costs are paid out of contribution margin, so revenue alone settles nothing. Willingness to pay differs across customers, which is why segmentation exists. A price a customer accepts depends on what they think the alternative costs.

Those four sentences are worth more in a Simon-Kucher case than any four-box grid, because they generate branches instead of supplying them. Ask where the margin actually sits and the structure follows. Announce a named framework and you have told the interviewer you brought a shape.

So the practical rule for this firm is narrow. Know the economics of pricing cold. Do not open by naming a template. The difference shows up in the first ninety seconds of the case.

Why a Simon-Kucher case is a pricing case

The careers page describes the work as helping clients optimize every lever of their commercial strategy, naming product, price, innovation, marketing and sales, based on deep insight into what customers want and value. The firm says it has more than forty years behind it, offices in over thirty countries, and more than 2,200 people, across five sectors: consumer and retail, financial services, healthcare and life sciences, industrials, and technology media and telecoms.

That shapes the case you get. A generic profitability question has a fork in it: cut cost or raise revenue. At a pricing firm the cost branch is usually the boring one, and the interesting question is what a customer will pay and why.

Three habits follow from it.

Name the unit before the total. Price per what? Per seat, per month, per ton, per procedure. Cases go wrong in the first minute when a candidate computes a total without deciding what is being sold.

Separate volume from mix. Average price can fall while every single list price rises, because customers moved to the cheaper tier. A candidate who does not check mix will report the wrong cause.

Ask who pays. In healthcare the payer is often not the user. In B2B the buyer is often not the budget holder. Willingness to pay is a property of whoever signs.

The pricing strategy cases page collects more of these. The one below is written for this firm, and it carries the arithmetic in full.

A price rise case, in six moves

The prompt: a subscription software company charges $40 a month and has 50,000 subscribers. Marketing wants a 15% increase. Should it go ahead?

1. Play back and fix the unit. "So we sell one subscription tier at $40 a month, to 50,000 customers, and we are considering $46. The decision is whether annual contribution goes up or down. Is $40 the same for everyone, or is there a discounted annual plan?"
2. Get the cost to serve. Suppose it is $8 per subscriber per month, mostly hosting and support. Contribution per subscriber is $32 today, and $38 after the rise. Monthly contribution today is 50,000 times $32, so $1.6 million, or $19.2 million a year.
3. Compute the break-even churn, because that is the whole case. To hold $1.6 million a month at $38, we need $1,600,000 divided by $38, which is 42,105 subscribers. We can lose 7,895, which is 15.8% of the base. Now round out loud, and say that you are rounding: call it 16%, or roughly one subscriber in six. One in six is 16.7%, so the spoken round is slightly kinder than the arithmetic. Announcing that gap takes two seconds and is the difference between a rounded answer and a wrong one.
4. Test the number against something real. If monthly churn is already 2%, losing 15.8% at once is close to eight months of normal churn arriving in one billing cycle. That is a large move, but a 15% rise is also large, so the comparison is worth making rather than assuming.
5. Segment before deciding. Annual plan holders have already committed and cannot react until renewal. If 30% of the base is annual, only 35,000 subscribers can leave this quarter, and the effective near-term risk is smaller than the headline. Heavy users face a higher switching cost than light ones, so the loss will not be spread evenly.
6. Recommend, first sentence first. "Raise the price, but only for monthly subscribers and only on new signups first, because break-even churn is about 16% and a staged rise lets us measure the real number before exposing the whole base. Evidence: contribution per subscriber moves from $32 to $38, so we can lose about 7,900 of 50,000 and still be level. The risk is that light users churn at several times the base rate, so I would watch the lowest usage decile for one quarter before extending it."

Notice what move three did. It converted a pricing question into a single threshold, and a threshold is something a client can test.

How we score the arithmetic on a pricing case

Simon-Kucher publishes no marking scheme, and none of the pages read for this guide names one. So the weights below are ours and carry our name: the dimensions our own grader measures on a math answer, and the weight each carries out of a hundred. Nothing here is the firm's scheme, because the firm has not published one.

The four preparation habits the firm publishes

The becoming a consultant page ends with advice, and unusually for this genre it is about method rather than content. Four habits, with the part candidates skip.

The last one is the cheapest advantage available here, because Simon-Kucher publishes heavily on pricing and commercialization. Reading two of its own pieces before an interview gives you the vocabulary its consultants use and a specific thing to ask about.

On the case itself, the firm lists five things it looks at: whether you understand complex challenges, structure information logically, solve quantitatively, solve creatively, and interpret and communicate results. Four of those five are visible in the six moves above. The fifth, creativity, is the one to plan for deliberately, and in pricing it usually means proposing a mechanism rather than a level: a tier, a bundle, a usage charge, a commitment discount.

Timed practice against those five is what the drill library exists for, and the dimension you lose repeatedly is more informative than the score.

The interview runs both ways, and that is on the record

The firm writes that it sees the interview as a two-way conversation. On its own that sentence would not be worth quoting. This one has published what it means, in the words of the people running the process.

Scherhag writes that candidates "are welcome to ask thoughtful questions, challenge perspectives, and actively contribute to the discussion," and that the firm particularly values genuine interest, reflectiveness, and a willingness to engage on equal footing.

Challenge perspectives is a strong phrase to publish. Taken at face value it means a candidate who pushes back on a premise, with a reason, is doing what the firm asked for. It does not mean arguing for sport. The difference is whether you name the evidence that would settle it.

Courtney Bromfield, recruiting manager in London, writes on the same page: "Think carefully about the experiences and achievements that best demonstrate what you would bring to the role and what genuinely motivates you to join Simon-Kucher." That is an instruction to select. The alternative is bringing your whole history and leaving the interviewer to sort it.

From the what we look for page, Ines Cubillo in Madrid frames entry-level hiring as a strategic investment in the future of the firm, looking for people who are motivated, curious, and eager to grow alongside strong analytical capability and academic results. Read together with the section above, the picture is consistent: the firm is hiring a way of thinking and expecting to train the content.

So prepare two or three questions that could only be asked of this firm. A question about a pricing piece it published lands differently from a question about culture, because only one of them proves you read anything.

What Simon-Kucher does not publish

How many rounds, and how long each runs. Nothing on any page we read. The firm says the combination of methods depends on the role and the level, which is a warning against any confident general description you encounter.

Which vendor supplies the cognitive tests, and for which roles. The pages say cognitive assessments such as numerical and reasoning tests apply to some positions. That is the whole of it. No test name, no publisher, no duration.

The entry job title after a bachelor's degree. No page read for this guide names one, in any country or any practice. The Madrid quote describes entry-level hiring without attaching a title to it.

Any threshold, weighting or academic cutoff. The Madrid quote mentions excellent academic results without attaching a number, and no page turns any of the five skills into a score.

What working with AI tools means inside an assessment. The firm says it values the ability. No page says whether it is observed in an exercise, asked about in an interview, or simply noted on a resume. Treat it as a stated preference until a page says otherwise.

That list is deliberately long, and it is the part of this page with the shortest shelf life. A careers page can change overnight. If your invitation email contradicts anything above, the invitation is the authority, and we would like to hear what it said so we can date the change.

What changed in this Simon-Kucher guide

29 September 2026. The version published on 24 September 2026 stated that the graduate track opens at Consultant, and built its advice around that seat. No Simon-Kucher page carries that title, and we could not find it on any careers page in any of the countries we read. The claim is gone, and the list above now names the entry title as one of the things the firm does not publish.

The same version was rebuilt around the firm's own wording on AI, which nothing on the site had been quoted for before. Five sources were added, each with the date it was read, including the two Wayback snapshots used for McKinsey and Kearney because those sites refuse our connections.

One arithmetic line was tightened the same day. The break-even churn in the pricing case is 15.8%, and the page had rounded straight to one subscriber in six, which is 16.7%. Both numbers are now on the page, with the direction of the round said out loud, because a rounded threshold that is never announced as rounded reads as a precise one.