L.E.K. Interview 2026: Seven Case Steps, Two Fit Words

Seven steps for the case, two words for the behavioral interview

The interview preparation page is short, about 1,137 characters, and almost all of it is a numbered sequence for working a case.

Six of seven carry an instruction, which is 86 percent, since 6 divided by 7 is 0.857. The one that does not is the only one about the person rather than the reasoning. A candidate reading this page learns exactly how L.E.K. wants a problem approached and learns nothing at all about how it wants a story told.

Check the preparation advice on the same page and the pattern holds. Four suggestions: attend campus case preparation workshops, read the business press regularly, practice with people currently working in strategy consulting and know the firm's work, and "Review the case studies and other content available throughout our website". Four suggestions, zero about assembling personal examples.

That is a real absence and it has a real consequence. If you prepare from L.E.K.'s own page, you will arrive with a well-drilled case method and an unprepared answer to the first question you are asked.

A behavioral interview question inside the case has a clock

One document tells you, and it comes from an employee rather than from the firm's editorial voice. A first-year associate in the London office wrote a post on the firm's careers blog, and the useful sentence is this: "The in-person interviews were case-based, focusing on problem-solving and fit questions."

Read it carefully. The interviews are described as case-based, and fit questions are inside them. So there may be no separate slot at all, which would explain why the preparation page treats the personal half as one line at the end of a case sequence.

L.E.K.'s roles page confirms the title he held: Associate is the undergraduate entry level and Consultant sits at post-MBA. The same post gives the shape of the process, again as one person's recollection: "The interview process for an associate role in London consists of multiple rounds, including online testing and face-to-face interviews. The online testing included psychometric and math exams." One office, one title, one candidate, published by his employer. The firm itself publishes no round count, no duration and no test.

The practical consequence is a length constraint that nobody states. If a fit question arrives inside an interview whose main business is a case, the answer has to be short enough that the interviewer can return to the case without visibly steering. Three minutes of story inside a case interview is a decision about the interviewer's clock that the interviewer did not make.

The same post carries the single most useful sentence about standards in our whole L.E.K. reading: "The key was not necessarily getting the right answer but demonstrating structured problem-solving skills." That is an associate's reading of his own interview rather than a published criterion, and it is still the closest thing available.

Stick to your approach, then give up conclusions

Two of the seven steps pull against each other, three places apart.

Step three: "Outline how you are thinking and move methodically to address each issue, point-by-point." Step six: "Commit yourself to a course of action, but show flexibility and a willingness to rethink in the face of new data."

This is the real tension of any interview, and the firm writes it well enough to be quoted rather than smoothed over. Hold your structure so the conversation has a spine. Abandon a branch when the data kills it. The resolution is a distinction the page never makes: you keep the structure and you give up conclusions.

It also sets L.E.K. apart from a firm that publishes the opposite emphasis. Bain's case interview page says it is looking at "how you think, structure problems, and build on ideas as the conversation evolves". Building on ideas as they evolve is nearer to step six than to step three. Two firms, two published postures, both citable.

For a behavioral answer the same tension appears in a smaller form. Stick to the story you started. Give up the framing when the interviewer's follow-up shows the framing was wrong. A candidate who defends a framing under pressure is doing step three where step six was wanted.

One behavioral interview answer sized to fit inside a case

The example below is ours. It is built to run about sixty seconds, because that is what fits when the question arrives inside a case. Treat it as a demonstration rather than as an answer to reuse.

Question form. Something like "tell me about a time you had to get a group to commit to something".

The answer. "I ordered kit for a cycling club of 140 members. The supplier needed a minimum order of 100 and I had 62 members presold at 85 dollars, so I was 38 short. I offered a 70 dollar early tier to the 60 members who had not decided, and 44 took it, which put the order at 106. The trade-off was that about 18 of those 44 had already told me they would pay 85, so I gave away roughly 270 dollars, since 18 times 15 is 270. I took that because failing the minimum meant no kit at all that season. What I got wrong was not asking the supplier whether 90 units was negotiable before I discounted anything."

Recompute: 62 plus 44 is 106, which clears the minimum of 100. 18 times 15 is 270.

Sixty seconds, one decision, one number, one admitted error. The last sentence is the one that earns the follow-up you want, and it is also the one step seven of L.E.K.'s list is gesturing at without saying so.

The unit check a behavioral interview needs, from L.E.K.'s blog

The London post contains a failure worth more than most success stories. The author describes a market sizing question about weddings where he sized the market as though every person married separately, rather than two people to a wedding. The interviewer credited his method and pointed out the error.

That error has a shape, and the shape transfers to behavioral answers as much as to sizing. It is a unit error: counting people when the event is the unit. Take a hypothetical to see the size of it. If 4.6 million people marry in a year, the number of weddings is 2.3 million, because 4.6 divided by 2 is 2.3. Sizing per person overstates the number of events by a factor of two, and the numbers are illustrative rather than from any source.

The behavioral version is the same slip. Saying "we grew revenue by 30 percent" when you owned one of four regions attributes the whole event to yourself. Three checks fix it before an interviewer does.

1. Name the unit out loud. Per site, per member, per order, per region. L.E.K. step one already tells you to state assumptions; this is that instruction applied to your own history.
2. Say which share was yours before you say the total, so the follow-up has nothing to catch.
3. Give the denominator, always. A 30 percent gain on 40,000 dollars and on 40 million dollars are different stories told with the same sentence.

One firm nearby removes the question entirely by folding the fit half into the case: Strategy& shares a single clock between the two.

What L.E.K. never publishes, and how we score it

The firm publishes no behavioral question, no criterion, no duration, no threshold and no weight, and its only round count is one associate's phrase "multiple rounds". The weights below are ours, applied by our own grader and read from the code that scores, and no firm publishes a marking scheme for a behavioral answer.