PEI Connection dimension

Start with the litmus test

Coaches on PrepLounge use one test to sort Connection stories from everything else: take the other person's change of mind out of the story. If your result still stands, it is not Connection. A model adopted because it was right is a problem-solving story. A decision that went your way because you outranked someone is an authority story. Connection is the one dimension where the outcome runs through another person's head.

What a top Connection story contains, on our anchors: one named person who started opposed; the reason behind their position, which you found out before you argued; a deliberate move suited to that person, such as a reframe, involving them in the work, or evidence in the form they trusted; and a change of position that stuck. Two things disqualify a story on fit: the persuasion was the analysis itself, or your title did the work.

The neighbor to watch is Leadership. Leadership is a group you delivered through by adapting to several people; Connection is one person. A story with a whole team in it can still be Connection if one named individual is the pivot. If you cannot name the person, you are on the wrong page.

If your story is about a group rather than one person, read the Leadership guideinstead.

What Connection means, and what Personal Impact was

Connection was formerly Personal Impact. Coaches and candidates report that McKinsey renamed all four dimensions in summer 2025; a coach on PrepLounge listed Connection as the former Personal Impact that August, and PrepLounge's PEI article calls the rename "more of a facelift than a change in substance". McKinsey's localized careers pages still show the old name and the old prompt, so both will turn up in your preparation.

That old prompt is worth reading, because the current question is a close cousin. McKinsey's careers page, in its localized versions, says: "Explain a challenging situation you encountered when working with someone with an opposing opinion to yours." Coaching sites quote McKinsey's older framing of the dimension as "Interacting effectively with people is key to creating positive, enduring change." The word that matters is enduring. Interviewers want the person to have stayed changed.

They also want it to have happened at pace. In one PrepLounge write-up, an Implementation candidate rejected in round one had told a story about persuading someone "over several months"; the interviewer judged it "not compelling even after a second story" and advised showing influence in weeks. A senior interviewer on the same forum says constraints like "with strong disagreement" get added to filter out stories where nobody pushed back.

An example Connection opening (an illustration)

Written for this guide, not taken from a candidate. About 160 words, the shorter end of what coaches recommend for an opener.

"Last spring I was the analyst on a pricing project for a regional logistics client. The operations director, Marta, had run the depot network for twelve years and had told my manager that our proposed zone-based pricing would 'break the drivers' routes'. My manager asked me to prepare a rebuttal deck. I did not send it. Instead I asked Marta for forty minutes at the depot and walked a morning shift with her dispatcher. Her objection was not about price; it was that zones would be drawn by postcode, which cut across the routes her team had tuned by hand. So I redrew the zones on her route map and showed her the revenue difference was under two percent. She asked for one change, a buffer corridor between zones, and then presented the pricing to the board herself. The policy went live in eight weeks, and Marta is the person who now trains new dispatchers on it."

What this sets up: a named person, her objection in her words, the move you did not make (the rebuttal deck), the move you did make (the shift, the route map), and two proofs that the change stuck. It also hands the interviewer the thing to test: was it the two-percent number that won, or the map? That answer decides whether the story is Connection or analysis.

Connection opening questions candidates report

Two former final-round interviewers on PrepLounge stress that there is no fixed question bank; follow-ups are built from your story. Openers do repeat. These ten are paraphrased from candidate reports and coaching sites, not from any McKinsey list.

Prepare two. McKinsey's careers page says "We suggest coming prepared with two personal examples" for each area, and interviewers can ask you to switch mid-answer if the first story does not show the trait. Coaches add their own conventions: within the last two to three years, you were the one deciding, and you remember the conversations in detail.

• Tell me about a challenging situation with someone who held the opposite view to yours.
• Describe convincing someone to change their mind on something that mattered to them.
• Tell me about influencing a senior leader or stakeholder when you had no authority over them.
• Describe getting buy-in from a reluctant stakeholder.
• Tell me about resolving a disagreement with a teammate.
• Describe aligning two groups with conflicting priorities.
• Tell me about delivering feedback that was hard to hear.
• Describe handling a politically sensitive situation.
• Tell me about influencing an external partner or client.
• Describe a time your communication skills got you out of a tough spot.

All four dimensions' reported openers and the twenty reported probes are in the McKinsey PEI questions guide.

The probe chain on a Connection story

An ex-Bain coach writing on PrepLounge reports 8–10 follow-ups on one story; coaching sites put the range at 10–25. A senior interviewer on the same forum estimates roughly 70% judgment probes to 30% factual checks, factual first. For a Connection story the chain usually runs like this:

Answer each in one to three sentences and do not restart the story. If a probe asks for words you do not remember, say so; coaches say admitting a gap beats inventing one, and the late re-ask of an early fact exists to catch invention.

• Fit check: "Who specifically was on the other side of this, and what did they say they objected to?"
• Factual: "How senior was she to you, and how long did the disagreement run?"
• Ownership: "You said 'we revised the plan'. What did you personally change?"
• Ownership: "What did you actually say to her in that first conversation?"
• Judgment: "Why go one-to-one instead of raising it with the whole team?"
• Judgment: "What did she say back, and why do you think she felt that way?"
• Counterfactual: "If she had still said no, what then?"
• Evidence: "How do you know she was genuinely on board and not just conceding?"

Hear this chain on your own story in a free Connection mockbefore the real one.

Strong, weak, and the three mistakes

Strong Connection stories share the markers coaches agree on for any PEI story: quoted dialogue, your decisions with the reasoning, a real obstacle, a verifiable result, honest reflection, and consistency under re-asking. A weak one sounds like "I presented the data and they agreed." Three Connection-specific mistakes recur in coach write-ups and rejection threads.

A problem-solving story in disguise. "I built a model to prove my point" is the failure coaches name for this dimension. If the analysis did the persuading, dimension fit is 0 on our scale, however good the analysis.

No understanding of the other side. If you cannot answer "What was their argument, in their words?" or "Why did she resist at first?", you argued at the person rather than with them. A final-round rejection discussed on PrepLounge turned on exactly this: the case was very good, but the conflict story showed no empathy and no lasting influence on the person.

Influence that never finished. Months of lobbying with no decision, or a person who conceded in the meeting and reverted afterward. Show the change, the timeframe in weeks, and the later action that proves it held.

How the simulator scores Connection, and what the free mock gives

McKinsey's scale is private; coaches say the criteria live in an internal interviewer guide. Our simulator scores six criteria from 0 to 4 on our scale, not McKinsey's: Ownership, Specificity, Obstacle, Impact, Reflection and Dimension fit. On each, 0 is absent, 1 asserted, 2 described generically, 3 specific and credible, 4 specific, quantified and intact after probing. For Connection, Dimension fit scores 4 when one specific person moved from no to yes because you understood their view, not through analysis alone; 1 for "I presented the data and they agreed"; 0 when the analysis or your authority did the work. The overall score is the lower of Dimension fit and the mean of the other five, because coaches say an off-dimension story fails whatever its quality.

The first full mock is free, with an account and no card. The AI interviewer takes one Connection story at full depth: the opening question, 6–12 follow-ups, up to 14 interviewer turns. The debrief scores the six criteria with quotes from your own answers, rewrites your weakest moments, and saves a story outline to your story bank. Starter adds the full debrief on every mock, unlimited single-dimension drills and the story bank; Elite adds unlimited full mocks, tough-interviewer mode, transcript export and analytics. Current prices are on the pricing page. One-time payments, 7-day money-back. PEIMock is independent and not affiliated with McKinsey & Company.

For where Connection sits inside the full interview, start with the McKinsey PEI guide.