Candidates usually arrive at the McKinsey Personal Experience Interview with a finished-sounding story. Biggest project on the resume. Stakes inflated just enough. A neat lesson at the end.
Plenty of those stories still fail.
The interviewer is not scoring how impressive your CV line sounds. They are scoring whether you can operate a messy situation the way a consultant has to: clock running, facts incomplete, people who do not report to you.
I have been on both sides of that table. The miss is usually the same.
This page is the teaching essay. The mistake. A bad vs good walkthrough. What good looks like. Failure modes. A practice assignment. For dimensions and a four-week plan, use the McKinsey PEI guide. For prompts, use McKinsey PEI questions. For rehearsal structure, use how to practice PEI.
CoachNed is independent practice material, not affiliated with McKinsey.
The mistake
People treat the PEI interview like a highlight reel.
Prompt lands: tell me about a time you led a team through something hard.
Candidate starts at the ending. We shipped. Revenue moved. Manager was happy. Then they invent a plot that makes them look decisive from minute one.
That is not the ask.
Listen for three checks:
- Did you own a real decision, or were you merely present?
- Did you change someone else's behavior without a title that forced compliance?
- Can you name tradeoffs in plain language, without buzzwords?
Skip any of those and the story reads like a LinkedIn post. Interviewers have heard a thousand of those.
A worked walkthrough (bad vs less bad)
Prompt: tell me about a time you influenced a decision when you did not have formal authority.
The version that usually dies
"In my internship I was on a pricing project for a retail client. The team wanted to raise prices across the board. I built a model that showed elasticity by category. I presented it to the partner. They used my analysis and we recommended a selective increase. The client liked it and I got great feedback."
Why it dies:
- Nobody disagreed, so there was no conflict.
- You made a spreadsheet. That is not personal risk.
- The partner already wanted analysis. That is not influence.
- "Selective increase" is the answer, not a choice you fought for.
An interviewer hears: capable intern. Move on.
A version that survives
Same internship. Different spine.
"Week three of a six-week pricing case. Client CFO wanted a flat 4 percent lift on the whole assortment because comps looked soft. Our engagement manager was leaning the same way. It would close the gap on the steering committee slide.
I owned the private-label category deep dive. My numbers said private label would lose enough volume in three regions to erase the margin gain. National brand could take 3 to 5 percent with almost no volume loss.
Timing was the problem. The flat lift was already in Thursday's draft pack. I did not own the pack. I did not own the client relationship.
What I actually did: pulled the three worst private-label SKUs, ran a simple before-after at plus 4 percent using the client's own elasticity ranges from last year, and spent twelve minutes with the EM before the internal dry run. Not a full model. Three SKUs. One chart. One ask: put private label in a test cell, or we burn volume to make a slide look clean.
EM pushed back. CFO already liked the flat story. I said then put the risk on the page. Hide it, watch volume fall in month one, and we own that too.
We kept the flat lift on national brand. Private label went to a staged test in two regions. Client still got a near-term revenue move. We did not pretend the risk was zero.
What I would still change: I waited until Wednesday to escalate. Should have flagged it Monday, the day the draft locked the flat number."
Difference in one line: a named opposing view, a risk you took, a concrete move only you made, and a leftover scar (timing) you can still defend. The scar is useful. It proves you are not performing sainthood.
Do not copy this internship. Steal the shape.
What good looks like in the room
I coach PEI stories to hit five beats in under three minutes.
Situation in two sentences. Who wanted what. Why it mattered that week, not in abstract career language.
Your stake. What broke if you stayed quiet. Career risk is fine. Client risk is better. Team risk is best.
The move. One or two actions you personally took. Not "we aligned stakeholders." Name the email, the meeting, the analysis, the hallway conversation.
The pushback. Someone said no, or not now, or this will upset X. Quote them if you can. No pushback means it was not a hard situation.
The outcome plus the leftover mess. What changed. What stayed ugly. What you would redo.
Cannot hit those five? Pick a smaller story. A smaller true story beats a big hollow one every time.
Common failure modes I keep seeing
Hero with no other people. You and a spreadsheet is not influence.
Committee win. "We all agreed after a workshop." Facilitation, maybe. Not the leadership muscle PEI is testing.
Fake number. "Engagement went up 40 percent." If you cannot say how you measured it in one clause, drop the number. Fake precision gets poked.
Resume tour. One story asked for, three mini stories delivered. Pick one and stay inside it. McKinsey will not let you switch once they lock on. That depth is the point of a PEI interview.
Moral at the end. "And that taught me the importance of communication." Cut it. Show the lesson inside the tradeoff.
Over polishing. Every sentence too clean and you sound coached. Leave one rough edge: an exact quote from a skeptical VP, a moment you almost backed down. That is what makes it human.
How firms actually score this
Labels differ. McKinsey PEI. BCG behavioral. Bain experience interview. The underlying ask does not.
They map you to a few dimensions: personal impact, entrepreneurial drive, leadership, courage and grit, inclusive collaboration. Do not say the dimension name out loud. Make it obvious from the plot.
Leadership prompt: stop spending two minutes on how smart the analysis was. Spend them on how you moved a reluctant person.
Personal impact prompt: show a decision that would have gone the other way without you. Specific enough that the interviewer could retell it to a colleague without losing the plot.
Public names on the McKinsey side shifted in 2025 (Leadership, Connection, Drive, Growth). The verification did not. Details live on the PEI guide. Cross-firm behavioral muscle: consulting behavioral interview.
Follow-ups they will ask (prepare these)
After the story, they poke. These four show up constantly.
What would you do differently? Empty answer means you do not reflect. Give one concrete redo, not a personality upgrade.
Who disagreed with you? Name a person and their incentive. "The client" is not a person.
How did you know it worked? One signal you actually had at the time. Not a number you invented later.
What did your manager say? If they rubber-stamped everything, say so and explain why you still pushed. If they blocked you, say how you handled the block without throwing them under the bus.
Practice those four as thirty-second bullet answers. That is often where the score moves after a decent delivery. Drill them via PEI practice or against the question bank. Live simulator: behavioral interview practice.
A practice assignment for today
Take one PEI prompt you already prepared. Rewrite it in this ugly outline. No paragraphs yet.
- Opposing person / view:
- What I risked by speaking up:
- The one concrete move I made:
- What still went wrong:
- One sentence outcome:
Say it out loud once into your phone. Three minutes max. If you cannot finish without reading, the story is still a document, not a story.
Do that for two prompts before your next mock. Then throw away the pretty version you had.
Worksheet version: PEI / fit workbook.
Last bit
Firms are not hunting for the most impressive project. They are hunting for evidence you can operate when the org chart does not protect you.
Bring friction. Bring a story where you almost lost. Bring something you can defend under follow-ups.
If you want to pressure-test the spoken version, the first scored case on CoachNed is free and the debrief is blunt about whether your thinking lands: try CoachNed free. Trusted by 2,147+ candidates across 30k+ sessions. Same muscle PEI needs: clear ownership, named tradeoffs, no buzzword fog.
The case still decides most loops. PEI decides whether they trust you in the room while you do the case. Pair this with how to do a case interview when you run both in the same afternoon.
Whole game, compressed: one real situation, one hard person, one move you owned.