Bain cases are candidate-led. The interviewer sets the prompt; you propose the structure, choose where to dig, run the math, and drive to a recommendation. First rounds are often more standardized. Partner rounds are conversational and will punish you for waiting to be handed the next chart.
This page is the Bain room. It is not the internship pay and programs page and it is not a McKinsey interviewer-led drill.
Last full pass: 24 August 2026. CoachNed is independent practice material, not affiliated with Bain.
What "candidate-led" means in the first two minutes
You open. A usable opening at Bain has four beats:
- Objective in one sentence, including the decision (buy / do not buy, enter / do not enter, fix margin / do not cut price).
- A structure you will actually steer, not a poster of every consulting topic.
- A hypothesis and where you want data first.
- A request: "I would start with contribution by segment — do we have that?"
If you sit in silence after the tree, you have handed the case back. McKinsey interviewers will often take it. Bain interviewers will let you drown, then write "did not drive."
Some offices add SOVA (verbal, numerical, logical, situational judgment) before live rounds. Confirm the invite. The SOVA simulator is transfer practice, not a leaked test.
Fit is not a separate sport the way McKinsey PEI is. Bain still wants stories with ownership, grit, and whether a team would want you back. Prepare them; do not spend the case proving you are nice.
First round vs partner round
| First round | Partner round | |
|---|---|---|
| Who | Often more junior interviewers; cases can feel standardized | Partners; conversational, more pushback |
| Your job | Clean structure, clean math, a recommendation with a number | Judgment, commercial instinct, being easy to staff |
| PE flavor | Possible, especially if the office does a lot of due diligence | Common: "would you put our client's money in this?" |
| Written case | Some offices / some loops | Confirm the invitation; do not assume |
| How you lose | Passive, template tree, arithmetic with no units | Brilliant analysis, no point of view |
Bain's private-equity work is a real interview signal, not trivia. You do not need a PE internship. You do need to think like an owner: cash, not just EBITDA; what is one-time vs repeatable; what breaks if volume is 10 percent light.
Full-time US campus pay, if you are choosing among offers, is on Bain salary: Associate Consultant base $112,000, MBA Consultant base $192,000. Intern monthly bases Bain actually posted are on Bain internships.
Worked example: three HVAC companies, one roll-up
Prompt. A mid-market PE fund is evaluating a $240 million purchase of a regional HVAC platform: three family-owned contractors in adjacent states. Combined revenue $180 million. Combined EBITDA $18 million. The thesis is density (shared dispatch, shared purchasing) plus a service-contract attach rate the owners never pushed. Should the fund proceed, and at what price?
Your drive (this is the Bain skill). Do not tour Porter. Name three tests and start:
- Quality of earnings. How much of the $18 million is owner compensation add-back, one-time storm work, or under-maintained trucks?
- Density math. What is the overlap in zip codes, and what is a realistic purchasing save?
- Attach. What share of installs currently convert to annual maintenance, and what is contribution per contract?
You ask for the quality-of-earnings bridge first.
Data they give once you ask. Reported EBITDA $18 million. Add-backs: $4 million of owner salaries that a professional CEO would still cost $1.5 million to replace, so only $2.5 million is real. Storm work that will not repeat: $3 million. Maintenance capex understated by $2 million a year (the fleet is old). Adjusted EBITDA: $18 − $1.5 − $3 − $2 = $11.5 million. Entry multiple on reported: $240 / $18 = 13.3x. On adjusted: $240 / $11.5 ≈ 20.9x.
You keep driving. Purchasing: combined spend on equipment $54 million. A national account could save 8 percent = $4.3 million, but only if the three brands accept one SKU list. Service attach today 22 percent; comparable platforms run 45 percent. Each incremental contract is $280 contribution. Installed base 90,000 systems. Moving attach 22 → 35 percent is 0.13 × 90,000 × $280 ≈ $3.3 million of contribution — if the sales motion exists. It does not, yet.
Recommendation. "Not at $240 million. Adjusted EBITDA is $11.5 million, so you are paying ~21x for a family business with a hope. I would need a price that is closer to 11–12x adjusted — call it $125–$140 million — before I underwrite density and attach as upside, not as the price. Near-term, I would diligence the fleet and the storm-year. The risk I am most worried about is that the 8 percent purchasing save requires killing two of the three equipment brands, which is how these companies won local share."
That is a Bain close: price, adjusted earnings, what is in the base case vs the upside, and a risk an owner would recognize.
Practise driving a messy commercial case on CoachNed. If the opening tree is the weak part, isolate it on structure drills.
How Bain differs from the other two rooms (without making that the whole page)
McKinsey will hand you the next question. BCG will often hand you a thicker exhibit pack and expect you to choose. Bain wants you to ask for the exhibit that tests your hypothesis, then move. Practising only interviewer-led cases creates the characteristic Bain fail: you wait, they wait, the clock dies.
Digital / AI roles under the Vector brand are a different loop (technical screen plus a build-flavored case). That is Bain Vector, not this page.
Failure modes that are specific to Bain
- Waiting to be led. You practised McKinsey. In this room that looks like low energy.
- EBITDA as cash. PE cases die when you ignore add-backs, maintenance capex, and one-time volume.
- A structure you never use. If you drew four buckets and then followed the interviewer's chat, the tree was decoration.
- No price. "It is interesting" is not a recommendation. Name a number or a walk-away.
- Fit as a TED talk. Bain listens for whether people would want you on a Thursday night in a hotel. Specific stories, not slogans. Cross-firm behavioral: behavioral interview consulting.
- Ignoring SOVA when it is on the invite. It is a real gate in some markets. Prep the sections, then come back to cases.
A three-week candidate-led plan
Week 1. Five openings only: prompt, two-minute tree, "I want this data first." Record yourself. Kill "I would like to look at…"
Week 2. Full cases where you request exhibits. One PE-style (quality of earnings, cash, price). One operations or growth. Score yourself on whether you drove.
Week 3. Partner-style: interviewer is warmer and more interrupting. Keep the thread. Close in 90 seconds with a number. If your office uses SOVA, one timed SOVA session so it does not steal the week of cases.
Bain varies by office. The posting and the invitation beat this guide. Use the method; confirm the logistics.