McKinsey internships are a shortened version of the full-time loop, not a casual summer job. Undergraduates (and some non-MBA master's) recruit for Summer Business Analyst. MBA, JD, MD, PhD, and other advanced-degree candidates recruit for Summer Associate. Both still run Solve, interviewer-led cases, and a Personal Experience Interview. A strong summer is the cleanest path to a return offer in the same office. McKinsey does not publish a universal intern monthly rate the way Bain has posted ACI and Summer Associate bases — do not copy a forum number into your spreadsheet and call it official.
Last full pass: 24 August 2026. CoachNed is independent practice material, not affiliated with McKinsey.
Which track you are actually in
| Summer Business Analyst | Summer Associate | |
|---|---|---|
| Typical candidate | Junior undergrad; some first-year master's | MBA intern summer; other advanced degrees |
| Full-time analog | Business Analyst (~$112K US base in the 2026 campus conversation) | Associate (~$192K US base; year-1 totals discussed on the salary page) |
| Interview shape | Solve (confirm modules) + interviewer-led case + PEI | Same bar, more ownership expected in the case |
| What "win" means | Return as BA (or later as Associate after a degree) | Return as Associate without repeating the circus |
| Prep mistake | Treating it as "easier McKinsey" | Bringing only PEI polish and a weak case |
Title trap: McKinsey Associate is post-MBA. BCG Associate is undergrad. McKinsey salary and BCG salary exist so you do not compare the wrong jobs.
Deadlines are school- and office-specific. How to get into MBB notes undergrad McKinsey windows often cluster around mid-August for some full-time cycles; internships can sit earlier. Use McKinsey's careers site and your school's schedule, not a friend's.
Hours and culture if you convert: working at McKinsey — 60–80 hour weeks as the honest range, obligation to dissent, up-or-out. Internships are a test of that rhythm, not a vacation.
The intern interview is not a discount case
You still have to:
- Open a structure that is specific to the prompt.
- Stay active in an interviewer-led room (case guide).
- Close with a number.
- Survive 10–20 minutes on one PEI story with no switch.
Solve is still commonly Redrock ~35 min + Sea Wolf ~30 min. Do not discover that the week you wanted to case. Redrock, Sea Wolf.
Intern fit questions include "why this office" and "what will you do in ten weeks." Vague "learning" answers sound like you want a tour. Specific: a practice, a city, a staffing model you actually asked a McKinsey person about.
Worked example: the ten-week utilization case they love giving interns
Prompt. Your intern class is 12. The office has 9 active studies. Each study wants "an intern for staffing calls and a deck." Historically, intern utilization above 90 percent for four weeks straight predicts burnout and a weak return-offer case because the intern never owns a module. Utilization below 60 percent predicts "we didn't see enough." Design a staffing rule for the 10-week summer. You have 12 × 10 = 120 intern-weeks.
Structure. (1) Capacity: intern-weeks. (2) Demand: studies × weeks they actually need a body. (3) Quality: at least one owned module per intern, which requires contiguous weeks, not scattershot staffing calls.
Exhibit (interviewer hands you). Nine studies: 4 "surge" studies want 1 intern × 8 weeks (32 intern-weeks). 5 "steady" studies want 0.5 intern-equivalent × 10 weeks (25). Total demand 57 intern-weeks against 120 supply — you are over-supplied, not scarce. The real constraint is contiguous ownership, not hours.
Math. If you spread 12 people across 9 studies as floaters, everyone is at ~57/120 = 48 percent billed-to-intern-week and nobody owns a workstream. If you staff 8 interns as primary on the 4 surge + 4 of the steady studies (pairing two steadies), and keep 4 as a flex pool for spikes, primary intern-weeks ≈ 8 × 8 = 64 on surge-heavy plus steady coverage — you can land ~70 percent utilization on primaries with a visible module, and flex stays lower by design.
Synthesis. "Do not maximize utilization. Demand is only 57 of 120 intern-weeks. Staff eight people as module owners on the studies that last eight weeks; hold four as flex so surge weeks do not raid the owners. Target ~70 percent for owners, not 90. The return-offer file should show a piece of analysis they can name, not 90 percent of staffing calls. Risk: a ninth study goes critical in week 3 — I would pre-agree which flex intern moves, so we do not cannibalize an owner."
That is an intern-flavored McKinsey case: operations, people, a number, a risk. Practise the motion on try and drills. More interviewer-led volume: practice cases.
Failure modes intern candidates repeat
- Assuming intern = easy. Same PEI depth, same interviewer-led habits.
- Inventing pay. If McKinsey did not post it, do not. For a posted intern monthly comparison at another MBB, Bain lists ~$9k ACI and ~$16k Summer Associate on Bain internships.
- Inventing return-offer rates. Offices differ. Perform; do not quote a blog percentage.
- PEI unprepared because "it's just a summer." They still will not let you switch stories.
- No office story. National brand worship is not a staffing reason.
- Solve last. It consumes the exact week you needed cases.
A compact prep plan
Applications. Resume with owned results. McKinsey interviewing for official cases. Cover letter only if asked.
Weeks before first round. Two interviewer-led cases a week. Eight PEI stories, probes on the weakest. Solve transfer if invited.
During the summer. Ask for a module with a metric. Document what you did in I-language for the return conversation. Read working at McKinsey so the hours do not surprise you after you convert.
Other firms' intern pages: BCG Summer Associate, Bain internships.