Bain Case Interview 2026: Two Tests, Two Different Rule Sets

What the Bain case interview process actually is

The hiring process page is vaguer than most summaries of it. It says the team reviews your application, that you might have a call with a recruiter, and that you will have several rounds of interviews. If you apply for a role such as consultant, you may also have a case interview. Everything after the application review is conditional or uncounted.

Several rounds means several. Your invitation names how many and in what order, and it is the only document that does. A tidy two-step pipeline of assessment then cases is a summary someone added, and this page used to carry one.

Two promises on that page are unusual enough to plan around. The firm says any feedback you receive will be based on an objective scoring system, and that after each interview you will learn exactly how it went. Most firms tell you nothing until a decision. If that holds in your process, an early round is a diagnostic you can act on before the next one, which changes how you should treat a first interview that goes badly.

One boundary worth knowing before you prepare the wrong thing: the firm states that case interviews are used for consulting roles only. If your application is not for a consulting seat, the case preparation industry is not aimed at you.

Sova or Test Gorilla, and why the difference matters

This is the part that costs candidates their application, and it is stated plainly on the digital assessment page. After you submit, an email tells you whether you must complete Sova or Test Gorilla. Requirements vary by role and by office. The two are not interchangeable.

Read the sittings row twice, because it is the line that costs an application. Splitting the assessment across sittings is allowed on Sova and forbidden on Test Gorilla. A candidate who has read a general rule somewhere, is sent Test Gorilla, and stops halfway has lost the application before anyone reads a word of it.

The rules that apply to both are strict. Sit it alone, in a quiet place, on a laptop or desktop with a stable connection. The firm's words on AI leave no room: their use is strictly prohibited, and that explicitly includes recording, screenshots, and any software that transcribes or gives feedback during the assessment. A meeting assistant you installed months ago counts.

Three more facts from the same page, each of which changes a decision:

• It takes roughly 30 to 40 minutes. Short enough to sit properly rather than squeeze in.
• Miss the deadline and you are not considered. The firm writes that applications will not be considered if the assessment is not completed in time, which is a harder consequence than a lower score.
• If you have already completed it, you do not retake it unless told otherwise. That matters if you are reapplying or have applied to more than one office.

In some markets the link arrives within a week of applying, though timing varies by office and region. The deadline itself lives in that email and on no careers page.

What Sova actually is, since Bain does not say

Bain publishes nothing about what its assessment contains. Neither the section names nor the question counts circulating online come from the firm, and that includes the three-section description this page used to carry.

What can be sourced comes from the vendor Bain points you to. Sova's own candidate hub lists a dozen assessment types under its own names, from personality and motivation questionnaires to "Situational Judgement", "Numerical Reasoning", "Logical Reasoning", "Verbal Reasoning", "Checking & Accuracy" and "Learnability" tests. The important line is how they combine: you see a set of activities selected by the employer. There is no fixed Sova test to memorize, because the employer builds it.

That has one clean consequence for you. Any list of the sections in your Bain assessment is a guess, including a confident one. Sova offers candidates a free practice test covering "Situational Judgement", a video interview and three types of ability question. Take the vendor's own practice test rather than anyone's description of the format, and treat the composition you meet as the one Bain chose rather than the one you read about.

Our Sova sampler practices two of the likely ability types under a clock. It practices the question types under time. It is no replica of your sitting, and it says so on the page.

The tools rule, and what it tells you to practice

The firm's wording is precise. Pen, paper and calculator are allowed for the digital assessment, and it says calculators are allowed only there. The interviews that follow are run without one.

A calculator being allowed tells you what is being measured. It is reading a dense exhibit fast and pulling the right figures out of it. Mental math drills train a different muscle, and the assessment never asks for it. Practice reading tables under time, calculator in hand, exactly as you will sit it.

The reverse holds in the interview, where the arithmetic is mental, spoken and deliberately roundable. A candidate who prepared both halves with the same drills has mis-prepared one of them. Bain's own in-person tip list makes the paper half explicit: bring pen and paper, write clearly, and be ready to walk the interviewer through your notes. Your working is part of what is read.

What Bain says a case interview is

The case interview page calls it a structured conversation around a real business problem of the kind its teams tackle every day. Then it says what it wants: how you think, how you structure problems, and how you build on ideas as the conversation evolves. It adds that a single right answer is not what it is looking for.

That last phrase is the one to read twice. It describes a case where the conversation moves and you move with it. You pick up what was just said rather than delivering the plan you arrived with.

In practice it asks for three behaviors. Say what you are doing before you do it, so the interviewer can follow or redirect you. When they hand you a fact, visibly use it rather than continuing down your prepared branch. And when a branch closes, say it is closed and name what you want to look at next, rather than drifting.

The firm publishes nineteen tips across three lists, and most of them are ordinary. Four are not, because they name a behavior rather than an attitude: clarify the objective before diving in, think aloud even when you are unsure, emphasize the key numbers so they are easy to refer back to, and pause to gather your thoughts when you need to. That last one is permission, printed by the firm, to stop talking for a few seconds. Candidates rarely take it.

A retail profitability case, worked through

Five moves, with a number carried through.

1. Play back the objective. "A specialty retailer has seen profit fall by a fifth over eighteen months while store count held flat, and we want to know why and what to do. Success is profit back to its prior level, over what horizon?"
2. Ask two or three things that change the structure. Did revenue move, or only profit? Is this the whole category or just this retailer? Two answers, and everything after them is different.
3. Structure from the arithmetic. "If revenue held and profit fell, the movement is in gross margin or in operating cost. I'd split gross margin into price, product mix and cost of goods, then look at operating cost by store and by function, and check whether the decline is the whole estate or a subset."
4. Run the number aloud, without a calculator. Say 120 stores at $2.4m is $288m of revenue. If gross margin moved from 42% to 39%, that is three points on $288m, or roughly $8.6m. Round hard: three points on nearly $300m is about $9m. If total profit fell by $14m, margin covers about three fifths of it and operating cost carries the rest.
5. Recommend, first sentence. "Address gross margin first, because it covers about three fifths of the decline, and the most likely driver is mix rather than price. Evidence: three points of margin on a $288m base is roughly $9m of a $14m fall. The risk is that the mix shift is customer-led and reversing it costs volume, so I'd look at units by category before touching the range."

That close is four sentences: the call, the figure, the risk, and what happens next. One line in step four is where cases quietly go wrong. Roughly $8.6m against $14m is 61%, and the mouth wants to say two thirds, because two thirds sounds like a finding. Round the inputs hard and then read the ratio as it lands.

The two cases Bain gives away, and the marking hidden in them

Bain publishes two complete cases on its own site, and they are the closest thing to a Bain case that exists in public. They sit three clicks inside the careers section, which is further than most preparation ever goes.

Coffee Shop Co. is a sizing question followed by a break-even. A friend wants to open a coffee shop in Cambridge. The firm hands you 100,000 residents drinking a cup a day, then a price of £3, £245,610 to open, £163,740 a year to run, and £1 of coffee per cup. Its published working is one line: 3Q minus 409,350 minus Q equals zero, so Q is 204,675 cups. The case sets that against a market of 7 million cups, which is about 3% share. Watch the units there. One cup a day across 100,000 residents comes to 36.5 million cups a year, so the 7 million rests on assumptions the case makes inside an image we could not read. We quote the figure as Bain's and we do not rebuild it.

FashionCo. is five years of falling revenue in women's fashion, resolving into a choice between what the case calls an awards program and three months of discounting a year. The published answer picks the discounting, at roughly $50m more in the year.

Work them aloud and timed before reading on. Then read them again for the thing they are actually teaching, which the firm never labels.

Both cases mark every answer twice, as a good answer and a better answer. That gap is Bain's scoring logic printed in the open, and it is the same gap every time:

Read down that right-hand column and it is one instruction. A good answer answers the question. A better answer adds the trade-off, the horizon, the risk and the next step. Nothing in it requires more analysis than the good answer did. That makes it the cheapest upgrade on the board, and it is the part an answer drops first when the clock is loud.

The coffee case carries one more line worth keeping. On sizing, the firm writes that there is no right or wrong answer and that you should base estimates on reasonable assumptions. Time spent hunting for the correct population figure goes on a problem the firm has said does not exist.

The entry title, and the bar it sets

Bain's management consulting page lists its ladder: intern, associate consultant, senior associate consultant, consultant. It lists associate consultant first among the full-time seats and talks about coming in at the entry level. The word Associate alone carries a different level elsewhere. McKinsey's MBA page points its students at Associate roles, and our McKinsey guide works through that firm's titles.

The consequence sits in what counts as evidence on the fit half, and here we are reading rather than quoting, because Bain publishes no standard by level. At the entry seat you are read on potential: a structure that holds under time, arithmetic you run without being asked, and a recommendation you defend when pushed. A university society you actually ran is strong evidence here, where against a graduate-school hire the same story lands thin.

Bain also has a large private equity practice, and commercial due diligence asks a narrower question than a strategy case. The question is whether our client should pay this price for this business. If that work is why you are applying, say so, because it is a reason that does not transfer to another firm's application. Our why Bain answer builds that reason on a figure the firm publishes, work with about 80 percent of the largest private equity firms in the world.

What the three things Bain names are worth

Bain names three things it wants to see: how you think, how you structure, and how you build on ideas. It attaches no weight to any of them, and no firm publishes a marking scheme. The weights below are ours, measured by our own grader, each out of a hundred.

Where Bain candidates lose it

1. Assuming you know which assessment you are getting. Sova and Test Gorilla have different rules on preparation and on sittings. The email that names yours arrives after you apply, and reading it is the whole defense.
2. Stopping halfway through Test Gorilla. It must be done in one sitting. Only Sova may be split.
3. Preparing the assessment with mental arithmetic. A calculator is allowed there, so what it measures is exhibit reading under time.
4. Practicing the case with a calculator. None is allowed there, and the habit shows in the first minute.
5. Letting the assessment deadline pass. The firm's wording is that the application will not be considered, which is a harder outcome than a lower score.
6. Answering only the question asked. Bain's own cases mark that as good rather than better. The trade-off, the horizon, the risk and the next step are the upgrade, and they cost no extra analysis.
7. Following the conversation into one branch and never coming back. Coverage is 45 of 100, and this is the failure mode of a collaborative case specifically.
8. Recapping before recommending. Answer first is 35 of 100.

The first five are specific to how Bain runs its process and are avoidable by reading the invitation and the digital assessment page carefully. The last three recur at every firm, which is why a drill diagnosis is worth more than firm-specific preparation once the rules are known.

A preparation order that fits this process

No fixed calendar, because the right length depends on which dimension you lose, and anyone selling a fixed calendar has not asked.

1. Diagnose, in twenty minutes. One short drill of each type: math, structuring, brainstorming, exhibits, synthesis, sizing. Read the dimension each rubric names as the first correction. That list is the plan.
2. Work the losing dimension until it changes. The structuring drill and the math drill are open without an account.
3. Read the email that names your assessment before preparing for it. Everything about how you sit it follows from whether it is Sova or Test Gorilla.
4. Split your assessment practice from your case practice. Exhibit reading with a calculator for one, spoken arithmetic without one for the other. The tools rule says so explicitly.
5. Take the vendor's own practice test, then the Sova sampler for two of the ability types under a clock.
6. Work Bain's two published cases aloud, then mark your own answer against its good and better columns. The gap is the fastest feedback available without a partner.
7. Say a ninety-second close out loud, ten times. Recommendation, two supports, one risk, one next step. That is the better answer, delivered.

What Bain does not publish

What is in your assessment. Neither the sections nor the question counts. Sova assembles each assessment from types the employer selects, and Bain does not say which it selects. Any list you read, including a confident one, is a guess.

Any pass threshold. Not for either assessment. Figures quoted elsewhere are unsourced, and the firm says only that the result is considered as one part of the review alongside the rest of your application.

The number of rounds and cases. The page says several rounds and gives no figure for either.

Which assessment you will be sent. The firm says it depends on the role and the office, and that the email tells you after you apply. Nobody can tell you in advance.

Whether the tools rule varies by region. We read it on the global digital assessment page. If your invitation says otherwise, the invitation wins.

What share of first-year work is private equity diligence. The practice is large and publicly described. Staffing decides where you land, a careers page cannot, and the people interviewing you can answer it.

Everything above about Bain was read on the page it is credited to. The one comparison with another firm carries that firm's own page as its source, and the scoring weights are ours, as the page says where they appear.

What changed on this Bain page

Each entry carries the date of the change.

29 September 2026. The sittings rule. An earlier version said the assessment could be split across sittings, with no condition attached. That is true of Sova and false of Test Gorilla, which the firm says must be completed in one sitting. A reader sent Test Gorilla who stopped halfway would have lost the application on our advice. The table above now sets the two rule sets side by side, and the difference has its own paragraph.

29 September 2026. Three invented sections. The page asserted that Sova contains numerical reasoning, logical reasoning and a personality questionnaire. Bain publishes nothing about the content of its assessment, and the vendor builds each one from types the employer picks. The claim has gone, replaced by what the vendor's own candidate hub says.

29 September 2026. A two-step pipeline. The page compressed the hiring process into an assessment followed by cases. The firm's page describes an application review, a possible recruiter call, several rounds of interviews, and a case interview for consulting roles. That summary was ours and it is gone.

29 September 2026. A ratio rounded the wrong way. The worked case called $8.6m of a $14m decline roughly two thirds. It is about three fifths, and the step now says so.