A PhD arrives at a case interview with five years of training in the hardest skill on the scoring sheet and almost none in the three easiest. The firm knows this. It built a separate pipeline to find you, then scores you on the same sheet it uses for a 28-year-old MBA.
Here is the thesis. Doctoral candidates rarely fail a case on rigor, structure, or thinking under uncertainty. They fail on points that cost twenty hours to fix and that business-school candidates picked up for free: forty words of business vocabulary, arithmetic said aloud at conversational speed, and reading a chart for its one decision-relevant number instead of its axes. From the interviewer's chair, the debrief sentence I heard most often about PhD candidates was some version of "clearly bright, but I don't know what they would do in front of a client on Monday." That sentence is not about intelligence. It is the business-judgment and communication lines coming in at a 2 while structure sits at a 4.
By the end of this piece you will know what the three big firms say publicly about hiring PhDs, which lines on the sheet you will win and lose, and how to spend a short preparation budget in the opposite proportion to an MBA's.
The pipeline is separate. The scoring sheet is not
McKinsey, BCG, and Bain each run a dedicated route for what McKinsey calls advanced professional degree (APD) candidates and the other two call advanced degree candidates. The routes differ in shape but agree on the point that matters for preparation: you enter at the same level as an MBA hire and sit the same interviews.
| Firm, as of September 2026 | PhDs join as | Program for PhDs and postdocs | Published dates |
|---|---|---|---|
| McKinsey | Associate ("APDs typically join us as full-time associates") | Insight workshop: 2.5 days in person, with a mock case | Class of 2027 applications July 1 to August 11, 2026; rolling for graduates by October 2026 |
| BCG | Consultant | Bridge to BCG: a full-time application "paired with a workshop to experience BCG's culture and prepare for final interviews" | 2026 workshop May 19 to 20 in Boston and Chicago, closed; September deadline for next-year starts |
| Bain | Consultant in the Americas; Associate Consultant in London and select Asia and Australia offices | ADvantage: one week, US and Toronto only, one day of training then four on a live case team | 2026 Texas program November 30 to December 4; apply by September 27, 2026 |
The sources are the firms' own pages: McKinsey's APD page, BCG's advanced degree page and Bridge to BCG page, and Bain's ADvantage page. Dates move every cycle; check before you plan a summer around one.
Two details on those pages tell you how the firms think. BCG says MBA students and holders are not eligible for Bridge because it "is designed for applicants with little to no business experience." Bain gives ADvantage participants one day of training before four days on a client team. Both are saying: we know you have not seen a P&L, we will teach you, and we will still interview you the way we interview everyone else. The programs source you and acclimatize you. They do not grade you on a curve.
None of the three publishes its scoring sheet, and I have no inside knowledge of anyone's current form. What BCG does publish is what the case interviewer watches for: how you structure the approach, the questions you ask, how you analyze data and "perform quick calculations," whether you identify the most important factors, and how clearly you communicate. There is no line for depth of expertise. In my years at a Big Four strategy practice, I never saw a column for the candidate's degree either.
You are not a novelty, either. US universities awarded 57,862 research doctorates in 2023, and 47% of the 22,462 recipients with a definite job lined up were headed to industry or business, against 34% to academia, down from 54% in 2003. Your interviewer has seen PhDs before, and the pattern in their head is the debrief sentence above.
Where a PhD's points actually go
Below is the six-line sheet I used, in plain words. Firms name the lines differently and weight them privately, but every form I have seen maps onto these. The last two columns are my estimate from watching doctoral candidates in first practice cases, not a published statistic.
| Line on the sheet | What the interviewer is listening for | Typical PhD, first case | Hours to move it one point |
|---|---|---|---|
| Structure | A tree that covers the question and says where you will start | Strong, often the best in the round | 5 to 10 |
| Prioritization | A stated guess, a reason, and a willingness to drop it | Sound logic, slow to commit | 10 |
| Quantitative | Set-up first, round numbers, the answer said aloud with its unit | Accurate but silent and slow | 8 |
| Business judgment | Knowing what a margin or a fixed cost implies without being told | Weak: the vocabulary is missing, not the reasoning | 6 to 10 |
| Synthesis | The answer first, then two reasons, inside a minute | Ends with "more work is needed" | 6 |
| Presence | Would I put this person in front of a client this week | Reads as a seminar speaker | Moves when the others move |
Add up the right-hand column and you get 35 to 45 hours from "clearly bright" to an offer, most of them on lines three to five. The MBA next to you spends most of their preparation on lines one and two, because those are what school did not give them. Copy their plan and you polish your surplus and leave your deficit alone.
Cheap point one: forty words of business
Business judgment in a case is not intuition. It is vocabulary plus arithmetic. If you know that gross margin is price minus the cost of one unit, divided by price, then a falling margin on flat volume tells you the story is price or unit cost, and you look sharp for saying so in ten seconds. If you do not know the word, you ask a clarifying question, lose ninety seconds, and lose the point.
You need about forty terms, each with a one-sentence definition you could say aloud: revenue, gross margin, contribution, fixed and variable cost, operating profit, EBITDA, break-even, market share, segment, channel, customer acquisition cost, churn, capex and opex, working capital, payback, and the rest. Write them on one page. It takes an evening, and it is the highest-return evening in your preparation.
The other half of vocabulary is translation. Your training gave you excellent concepts under the wrong names.
| You would say | What the interviewer hears | Say instead |
|---|---|---|
| Dependent variable | Jargon | The number the client is paid on |
| Treatment effect | Jargon | The lift from the change |
| Confound | Jargon | The other thing that could explain it |
| Statistical power | Jargon | Whether the pilot is big enough to show a difference |
| Prior literature | A seminar is starting | What competitors have already tried |
| Robustness check | More delay | The one thing that would change my answer |
Cheap point two: arithmetic you can say out loud
Nobody doubts you can multiply. The problem is that you have not done arithmetic without a tool, in front of another person, on a clock, since your second year as an undergraduate. The sheet does not score accuracy alone. It scores whether the number arrives while the conversation is still alive, whether you rounded sensibly, and whether you said the unit.
Three habits fix most of it. State the set-up before you compute: "sixty thousand units times a hundred dollars of lost price is the price effect." Round to one or two significant figures and say so. Give the answer its unit and its meaning in the same breath: "six million dollars, so about half the decline." Ten minutes a day for two weeks moves this line more than any other; start with the rounding and factoring shortcuts, and the quick-math drill is free for as long as you want it.
Ned's rule. Spend preparation time in inverse proportion to your pride. The exercise you would be embarrassed to be seen doing, saying "sixty thousand times a hundred is six million" to a wall, is the one standing between you and the offer.
Cheap point three: read the exhibit like a client, not a referee
Doctoral training teaches you to read a figure the way a reviewer does: axes, units, sample, error bars, then the claim. In a case, the interviewer hands you a chart and starts a mental stopwatch. The point goes to whoever says the one number that changes the answer within about fifteen seconds, and then says what it changes.
Reverse your order. Title first, then the biggest gap or the odd bar, then one sentence on what it means for the recommendation, and only then, if it genuinely matters, the caveat about the sample. The caveat is not banned. It is fourth, not first. The exhibit guide has the full routine, and the charts drill times you on it.
Kestrel Labs: one margin case, scored two ways
Kestrel Labs is an invented maker of benchtop laboratory centrifuges. Revenue was $240 million two years ago with an operating margin of about 14%. This year the margin is about 9%. The client wants to know why and what to do. When you ask, the interviewer gives you four numbers.
| Kestrel Labs | Two years ago | This year |
|---|---|---|
| Units sold | 60,000 | 60,000 |
| Average price | $4,000 | $3,900 |
| Variable cost per unit | $2,600 | $2,700 |
| Fixed costs | $50 million | $50 million |
The arithmetic fits on a napkin. Revenue fell from $240 million to $234 million; variable cost rose from $156 million to $162 million; contribution fell from $84 million to $72 million; operating profit fell from $34 million to $22 million, and margin from 14.2% to 9.4%. Volume is flat, so the whole $12 million decline is unit economics: $100 of price times 60,000 units is $6 million, and $100 of unit cost times 60,000 units is the other $6 million. Half price, half cost, nothing else.
Candidate A, the doctoral default, asks whether the $3,900 is a like-for-like cut or a mix shift toward cheaper models, whether the cost increase is one supplier or many, and whether two data points make a trend. All three are good questions. All three arrive before any statement of what the four numbers already show.
Candidate B says: "Profit is down $12 million on flat volume, so this is a unit-margin problem, not a demand problem. It splits evenly: $6 million from a $100 price decline and $6 million from a $100 cost increase. I would start with price, because it is the lever the client controls this quarter, and my first question is whether the $100 is a list-price cut or a mix shift, because the fix is different." Then B asks A's questions.
Same brain, same numbers. Here is how they land on the sheet.
| Line on the sheet | Candidate A | Candidate B | Why |
|---|---|---|---|
| Structure | 4 | 4 | Both split profit into price, volume, and cost |
| Prioritization | 2 | 4 | B named a starting point and a reason |
| Quantitative | 3 | 4 | Same arithmetic; B said it aloud with units and a meaning |
| Business judgment | 2 | 4 | B knew flat volume rules out demand and named the controllable lever |
| Synthesis | 2 | 4 | B's first sentence was the answer |
| Total of five lines | 13 | 20 | The gap is sequencing, not knowledge |
B did not know anything A did not. B said what the data meant before asking for more of it. That is the whole technique, and you can rehearse it on every exhibit you see for the next month.
The expertise trap
Sometimes the case lands in your field. A chemist gets a specialty-materials pricing case; an epidemiologist gets a hospital throughput case. In my experience this makes the case worse, not better. The candidate corrects the case's simplifications, answers the question they know rather than the one asked, and skips structure because the answer is obviously X.
The simplifications are deliberate. The interviewer usually knows less about your field than you do and does not care, because the sheet has no line for it. Use your expertise for exactly one thing: a sharper question. "Hospitals in this market buy through purchasing groups, so is the price decline at list or at contract?" That is expertise doing work. Three sentences on the history of the sector is expertise doing damage.
The line you win without trying
There is one place your training pays off directly, and most PhDs waste it. A good close ends with next steps and risks, and the interviewer wants a specific analysis, not "gather more data." You are the only candidate in the round who has designed a real experiment. Use it.
"To confirm price is the driver, I would pull eight quarters of invoice-level pricing for the top twenty accounts and separate list changes from mix. If it is list, restoring the $100 across the base recovers the $6 million. If it is mix, the fix is the product line, not the sales team." Sample, duration, decision rule. That is a pilot design, and it earns a point on the synthesis line that MBA candidates rarely collect.
The same asset works in the fit interview. Every PhD owns one story nobody else in the round has: the experiment or chapter you killed, on what evidence, and how you told your advisor. Told in three minutes with the disagreement left in, it covers judgment, personal impact, and resilience at once. Told as your thesis abstract, it covers nothing.
Practice this today
One assignment. Take the Kestrel numbers above and record yourself giving the Candidate B answer in under sixty seconds. Play it back and count the seconds before a sentence contains both a number and its meaning. If it is more than fifteen, do it again.
Then run one typed rep at /start: three turns of a case with instant scores, no account needed, to find your low line. Spend the week on it: quick math for arithmetic, charts for exhibits, synthesis for the close. When it moves, sit the live voice case with Ned; typing hides the seminar voice and speaking does not. Everything is open for seven days, no card; then $120 for a recruiting season or $49 a month.
CoachNed is independent and has no affiliation with McKinsey, BCG, Bain, or any other firm named here.
Frequently asked questions
Do McKinsey, BCG, and Bain hire PhDs without an MBA?
Yes, through dedicated routes. As of September 2026, McKinsey's APD page says APDs "typically join us as full-time associates," BCG lists full-time Consultant roles for PhDs, postdocs, MDs, and JDs, and Bain hires advanced degree candidates as Consultants in the Americas and Associate Consultants in London and select Asia and Australia offices.
Is the case interview different for PhD candidates?
Not in any way the firms describe publicly. Bridge to BCG and Bain's ADvantage source candidates and teach business basics before interviews; the case itself is the same generalist problem-solving interview given to MBA candidates.
How much does a consultant with a PhD earn compared with a postdoc?
Below is the six-line sheet I used, in plain words. Firms name the lines differently and weight them privately, but every form I have seen maps onto these. The last two columns are my estimate from watching doctoral candidates in first practice cases, not a published statistic.
The firms do not publish PhD-specific pay, but they say PhDs enter at the MBA level, and Harvard Business School reports a [$190,000 median base salary](https://www.hbs.edu/recruiting/employment-data/function/Pages/details.aspx?func=Consulting&view=career&year=2025) for its class of 2025 consulting hires, plus a $30,000 median signing bonus. The NIH NRSA stipend for a first-year postdoc is [$63,480 for fiscal 2026](https://www.vanderbilt.edu/postdoc/current-postdocs/stipends-salaries/), released in March 2026. Roughly three times, before bonus, tax, and the hours.?
Doctoral training teaches you to read a figure the way a reviewer does: axes, units, sample, error bars, then the claim. In a case, the interviewer hands you a chart and starts a mental stopwatch. The point goes to whoever says the one number that changes the answer within about fifteen seconds, and then says what it changes.
How many hours should a PhD spend preparing for case interviews?
Ned's estimate is 35 to 45 hours, spent in the proportions in the table above: heavy on vocabulary, arithmetic aloud, and the close, light on frameworks. The folklore figure of a hundred-plus hours comes from candidates who practiced their strengths.
Sources
- Advanced Professional Degree Candidates, All Schools — McKinsey & Company — APDs join as full-time Associates; 2026 and 2027 application windows. Checked 2026-09-24.
- 2026 McKinsey Recruiting Program for APD Candidates — Vanderbilt BRET Career Development — Insight as a 2.5-day in-person workshop with a mock case. Checked 2026-09-24.
- Advanced Degree Candidates — BCG Careers — eligible degrees, Consultant role, September deadline. Checked 2026-09-24.
- Bridge to BCG — BCG Careers — 2026 workshop dates, purpose, MBA ineligibility wording. Checked 2026-09-24.
- Case Interview Preparation — BCG Careers — what BCG says the case interview assesses. Checked 2026-09-24.
- ADvantage — Bain & Company — format, eligibility, 2026 Texas dates and deadline. Checked 2026-09-24.
- Advanced degree career opportunities — Bain & Company — Consultant in the Americas, Associate Consultant in London and select Asia and Australia offices. Checked 2026-09-24.
- Doctorate Recipients from U.S. Universities: 2023, Postgraduation trends — NSF NCSES — 57,862 doctorates; 47% of 22,462 employment commitments to industry or business; academia 54% to 34%. Checked 2026-09-24.
- Stipends and Salaries — Vanderbilt University Office of Postdoctoral Affairs — NIH NRSA postdoctoral stipend table; year-0 level of $63,480 for FY 2026, released March 18, 2026. Checked 2026-09-24.
- Employment Data, Consulting, Class of 2025 — Harvard Business School — $190,000 median base, $30,000 median signing bonus. Checked 2026-09-24.