TL;DR. Three distinct grids in 2026: PEI Connection at McKinsey (~1% offers), Casey Chatbot + Online Case at BCG (1-3% offers), Bootcamp at Bain with fit weighted 30-40%. An AI layer on top: Lilli has joined McKinsey interviews since January 2026. "Case in Point + 30 cases" no longer cuts it.
You are prepping McKinsey, BCG or Bain from HEC, X, ESCP, EM Lyon or Dauphine. You've read Case in Point. You drill cases every night.
Except the grid moved in 2026: PEI rebranded, Casey mandatory, Lilli walking into the McKinsey interview.
Simple question: are you still preparing the 2022 version of the MBB interview, or the 2026 one?
MBB France process 2026: why the grid changed in 18 months
The strategy consulting market accelerated. BCG posted $14.4 billion in 2025 revenue, up +7%, with 33,500 staff (Consultor, 2026).
McKinsey now stands at roughly 40,000 employees worldwide — versus 45,100 in 2023 (Consultor, 2026). The firm trimmed headcount while remaining the most selective of the three.
Mix of work has tilted hard: 40%+ of BCG's revenue comes from tech and AI projects in 2025 (Consultor, 2026). The profiles in demand shifted accordingly: AI engineers, data scientists, hybrid business/tech candidates.
On the HR side, McKinsey rolled out Lilli, its internal AI assistant. Documented result: a 25% reduction in time spent on performance reviews and adjacent HR tasks (Consultor / Bloomberg, 2026). Faster processes, more standardised evaluation, far less tolerance for candidates who aren't quite ready.
McKinsey France: PEI Connection, Lilli and the post-2024 grid
First structural shift: in summer 2025, McKinsey rebranded its Personal Experience Interview (PEI) dimensions. Out: Personal Impact. In: Connection. The full set: Drive, Growth, Leadership, Connection (WashU McKelvey Career Hub, 2025; StrategyCase, 2026).
Coaches call it a "facelift" — same substance, new vocabulary. If you're prepping with a pre-2025 manual, you'll walk in with the wrong wording.
- Drive — what sets you in motion, what you defend under pressure.
- Growth — your learning trajectory and how you handle feedback.
- Leadership — how you move a group without formal authority.
- Connection (formerly Personal Impact) — your ability to influence in direct one-on-ones.
Format: 1 case + PEI per round, across 2 rounds — 4 to 6 interviews total. End result: ~1% of applicants get an offer (PrepLounge, 2025). Raw selectivity.
New for 2026 — Lilli in the interview. Since January 2026, McKinsey has been piloting a format where the junior candidate has to use Lilli to analyse a case. The evaluation is no longer just about the structure of your thinking — it covers the quality of your prompting and your critical thinking when reading the AI's answers (Consultor, 2026).
Concrete implication: if you've never trained a structured analytical prompt on ChatGPT or Claude, you're starting at a disadvantage. The candidate who interrogates the AI beats the one who only receives its answer.
On the PEI side: prepare 8 to 12 timed stories (10 min each), covering the four dimensions, with a clean situation → action → measurable impact arc.
BCG France: Casey Chatbot, Online Case and the 1-3% filter
At BCG, the filter is now called Casey Chatbot: 30 min, 8-10 analytical questions plus a final video recommendation, no break (StrategyCase, 2026).
It has become an almost mandatory screening step for consultant candidates. And it's a hard filter: only 1 to 3% of BCG applicants eventually land an offer (StrategyCase, 2026).
Three classic traps on Casey:
- Underestimating the video recommendation. Many candidates burn time on the analytical questions and rush the video. Flip it: the recommendation is what evaluators read first.
- Failing to structure MECE. Casey isn't forgiving on rigour. A MECE answer is scored as such immediately.
- Mismanaging time. No break, no going back. Disciplined timing beats a perfect answer on question 4 and a blank on question 9.
In face-to-face rounds, BCG sticks to a candidate-led format, freer than McKinsey's, with strong focus on quantified hypotheses and a "so what" check at every step.
Bain France: Bootcamp, fit at 30-40%, and the "teamwork" culture
Bain has its own signature. First-round format: case 30-40 min + fit 15-20 min, with frequent mental math and market sizing (Duke University Career Hub, 2026; Consulting Bootcamp, 2025).
The differentiator: fit weighs 30 to 40% of the overall evaluation. By a wide margin, Bain is the MBB firm where authenticity and culture matter most.
- ✓Candidate-led: you drive the sequence
- ✓Regular mental math (3 digits, %)
- ✓Frequent market sizing
- ✓Hypotheses spoken out loud
- ✓≈ 60-70% of the first-round evaluation
- ✗Authenticity over recited script
- ✗Interview treated as teamwork, not a courtroom
- ✗MECE as a frame, not a straitjacket
- ✗Sharp 'why Bain' stories
- ✗≈ 30-40% of the first-round evaluation
Spencer Anderson, ex-senior manager at Bain Houston (9 years), published his hiring criteria in September 2025 (Consultor, 2025):
- Authenticity first. A candidate reciting their stories sounds fake. Bain wants to meet a person, not a script.
- MECE as a frame, not a straitjacket. You structure, but you adapt to the actual question.
- Treat the interview as teamwork, not a courtroom. The interviewer isn't there to trap you. Ask questions, check your assumptions, reason out loud.
In France, Bain has owned this strong-fit logic for decades. The Paris office, whose official hiring process gives fit explicit weight alongside the case, favours stable, well-rounded profiles with strong interpersonal capital — the opposite of transactional hiring (Bain & Company France, 2026).
Practical consequence: if you prep Bain like McKinsey, you over-invest the case and under-invest the fit. Flip it: two hours of fit for every hour of case in your final prep weeks.
MBB and French candidates: target schools and 2026 profiles
Core target schools for MBB France: HEC, X (Polytechnique), ESSEC, ESCP, Sciences Po, Centrale, EM Lyon, EDHEC, Dauphine (selective M2). Plus a few specialised engineering schools and the top international MS / MBA programmes.
What changed is the profile being chased inside those schools. With 40%+ of BCG's revenue from tech/AI (Consultor, 2026), firms are actively courting AI engineers, data scientists, and hybrid business/tech profiles.
Concrete implication if you come from a "pure business" school:
- A file with no data flavour, no quantitative project, no tech exposure is less competitive than in 2022.
- A dual engineering degree, a data MS, a credible open-source project, or a tech product role reshapes your positioning.
- Conversely, an X / Centrale candidate who tells a clean commercial story is now an ultra-coveted profile.
Outside target schools, it's harder, not impossible. You need a sharp differentiator: a prestigious PhD, serious international experience, or a recognised AI engineering profile.
How to prepare concretely for the three grids in 2026
Your most likely prep mistake: treating MBB as one single prep. It isn't. There are three distinct grids.
Differentiated game plan:
- McKinsey: timed PEI stories (4 dimensions, 8-12 stories), interviewer-led case, trained prompting on Lilli / ChatGPT / Claude — now evaluated (Consultor, 2026).
- BCG: Casey Chatbot speed-runs (5+ timed simulations), video recommendation polished as a deliverable, candidate-led case.
- Bain: deep fit bootcamp (15-20 min drilled in a mirror), daily mental math, market sizing with hypotheses spoken out loud, candidate-led case 30-40 min.
- 30-40 live cases, mix interviewer-led (McKinsey) + candidate-led (BCG, Bain).
- 8-12 PEI stories timed (10 min each), audio-recorded so you can listen back.
- 5+ Casey Chatbot simulations under realistic conditions, including the video reco.
- Daily mental math: 10 min/day, three-digit calculations, conversions, percentages.
- 2 sessions/week of market sizing said out loud.
- AI prompting: Claude / ChatGPT as a McKinsey interviewer, 2 sessions/week.
8-12 week routine:
- 30-40 live cases (mix of interviewer-led and candidate-led).
- 8-12 PEI stories drilled, timed, audio-recorded so you can listen back.
- 5+ Casey simulations under realistic conditions.
- Daily mental math: 10 min/day, three-digit calculations, conversions, percentages.
- 2 sessions/week of market sizing said out loud.
Bake AI into your training. Not just to revise — to train your prompting. Ask Claude or ChatGPT to play the McKinsey interviewer, force it to challenge you, critique its answers out loud. That is exactly what Lilli is scoring.
Want a structured framework? Our AI interview platform simulates an MBB interview (PEI + case + fit) with quantified feedback on your structuring, timing and AI prompting. And before you apply, have your CV audited against a target-school grid with tech/AI signals.
Frequently asked questions about MBB interviews in France
What is the acceptance rate at McKinsey, BCG and Bain in 2026?
About 1% at McKinsey (PrepLounge, 2025), 1-3% at BCG (StrategyCase, 2026). Bain doesn't publish an official figure but sits in the same range, with fit weighted at 30-40% of the final evaluation.
Do you need to come from HEC or Polytechnique to land an MBB interview?
No, but target schools concentrate most offers. Outside that pool you need a differentiating profile: PhD, AI/data engineering background, or a strong international track record.
What is McKinsey's PEI Connection?
The PEI rebranded in summer 2025: Drive, Growth, Leadership and Connection (formerly Personal Impact). Same substance, new label. Three 10-minute timed stories to prepare.
Is the BCG Casey Chatbot a hard filter?
Yes. 30 minutes, 8-10 analytical questions, plus a final video recommendation with no break. Failing Casey closes the door on face-to-face interviews.
How long should you prepare for an MBB interview?
8-12 weeks part-time for a target-school profile: 30-40 live cases, 8-12 PEI stories, 5+ Casey simulations.
Is the Bain fit really heavier than McKinsey or BCG?
Yes: 15-20 minutes out of a 45-60 minute interview, i.e. 30-40% of the evaluation (Consulting Bootcamp, 2025).
Should you train with ChatGPT or Claude for the MBB interview?
Yes, and it is now evaluated. Since January 2026 McKinsey has been piloting a format where the candidate uses Lilli to analyse a case.
McKinsey, BCG or Bain: which firm targets tech/AI profiles most?
All three, but BCG leads: 40%+ of global revenue from tech/AI projects in 2025.
What's the difference between a McKinsey case and a Bain case?
McKinsey: interviewer-led, tightly structured, math embedded. Bain: candidate-led, 30-40 minutes, frequent mental math and market sizing.
How many interviews are there in a complete MBB France process?
Usually 4 to 6 interviews across 2 rounds, plus the BCG Casey or an equivalent test upstream.
Key takeaways
- Three distinct grids: PEI Connection (McKinsey), Casey + Online Case (BCG), Bootcamp + heavy fit (Bain). Generic prep no longer works.
- Offer rates 1-3%: firm-by-firm strategy is now critical to clear the filter.
- McKinsey scores AI prompting via Lilli since January 2026 — train with ChatGPT or Claude.
- BCG is betting on tech/AI: 40%+ of revenue, AI engineers and data scientists as priority hires.
- Bain weighs fit at 30-40%: authenticity, flexible MECE, interview as teamwork.
- French target schools concentrate offers, but an AI engineering profile from outside stays competitive.
- Realistic prep plan: 8-12 weeks, 30-40 live cases, 8-12 PEI stories, 5+ Casey simulations.
