Most people who reach a consulting final round can already do the case well enough to be hired. In the decision meetings I have sat in, the case scores of the candidate who got the offer and the candidate who did not were usually close; the argument in the room was about something else. The final round is not a harder first round. It is a less structured one, judged by people with more discretion and less patience, and that changes what is worth practicing.
The instinct after a first-round pass is to do more cases. That treats the final as a difficulty problem. It is a variance problem: fewer interviewers, less script, more room for one partner's impression to decide the outcome. The research on unstructured interviews says that impression is noisy. Your job in the last week is to make yourself hard to misread.
Below: what changes when the interviewer is a partner, the three moments a partner remembers, how to take an interruption with numbers in hand, and a seven-day taper.
What changes when the interviewer is a partner
Firms publish less about final rounds than prep sites imply, so separate what they state from what I have observed. BCG's own interview brochure, dated November 2018, describes the second round as "similar to the first round, involving case studies and interviewing with more senior members of the firm," with a decision "within a few days"; BCG's current process page lists stages rather than numbered rounds. McKinsey's interviewing page says most client-facing candidates get a personal experience interview followed by a problem-solving interview, without distinguishing rounds. Bain's interviewing page says it gives "all candidates the same questions for each role to reduce bias."
The table marks which rows are firm-stated and which are my observation.
| Dimension | First round | Final round | Basis |
|---|---|---|---|
| Interviewer | Consultants and managers | Partners, principals, directors | Firm-stated (BCG brochure) |
| The case | A trained case with a scoring guide | Often the partner's own case, from memory | My observation |
| Interruptions | Few; you run your structure | Frequent; the interruption is the test | My observation |
| Experience questions | Standard questions, 10 to 15 minutes | Same slot, much deeper probing | Allocation firm-stated; depth my observation |
| The question being decided | Can this person do the job | Would I put this person in front of my client on Monday | My observation |
| Speed of decision | Days to weeks | Often within days | Firm-stated (BCG brochure) |
Why the final round is noisier, and why that helps you
A first-round interviewer usually has a script, a scoring guide, and a calibration session behind them. A partner usually has a case they have run for years and a strong sense of what a good consultant looks like. In the language of selection research, round one leans structured and the final leans unstructured.
Schmidt and Hunter's meta-analysis put the predictive validity of structured interviews at .51 against .38 for unstructured ones. A 2016 update by Schmidt, Oh, and Shaffer revised both figures to .58 using a different statistical correction, but the same paper reports the number that matters for you: when different interviewers rate the same applicants on different days, unstructured ratings correlate at only about .37 to .40. Two partners, two days, the same candidate, loosely related verdicts.
You cannot fix the partner's half of that noise. You can fix yours. A candidate whose opening is crisp on Tuesday and rambling on Thursday is adding variance to an already noisy instrument. The prep that pays in the final week makes you the same person in every room: the same 60-second opening, the same way of taking an interruption, the same shape of close. Bain's university hiring team said as much in March 2026: "build consistency rather than cramming cases at the last minute."
The three moments a partner remembers
An unstructured interviewer does not add up points. They form a summary impression, and in my experience it is built from three moments and back-filled from everything else.
The first 90 seconds of the case. The partner has heard the prompt many times; what is new is you. Restate the objective in one sentence, give a structure with two or three branches and a reason for the order, and say which branch you would open first. Under a minute, out loud, no preamble.
The first interruption. Partners interrupt because they are bored, because they disagree, or to see what happens, and you cannot tell which. Treat all three the same way: answer the exact question in one sentence, with a number if you have one, then ask whether they want the rest of your line of thought. A candidate who finishes their sentence before answering has told me they were reciting.
The last five minutes. The close is said to a client, not an examiner: recommendation first, the two numbers that carry it, the risk that would change it, the next step. Then your questions to the interviewer, which in my experience also get written down. The questions worth asking are the ones that show you listened to the case.
Ned's rule. When a partner interrupts, the interruption is the case. Answer it in one sentence with a number, then offer the rest. I have never failed a candidate for stopping mid-structure to answer me; I have failed several for finishing the structure first.
A worked example: taking the interruption with numbers
The partner opens with their own case. Harbor Fitness runs 40 gyms in one region. Each averages 2,400 members paying $45 a month. The owner wants to know whether to launch a premium tier at $75 a month. Go.
Your structure names three things: how many members would trade up, what the tier costs to deliver, and what it does to members who do not trade up. You open with uptake because it drives the rest. The partner gives you a figure: 15% of existing members move to premium, no new members, no churn effect.
| Step | Calculation | Result |
|---|---|---|
| Members today | 40 clubs × 2,400 | 96,000 |
| Revenue today | 96,000 × $45 × 12 | about $51.8M |
| Members trading up | 96,000 × 15% | 14,400 |
| Incremental revenue | 14,400 × $30 uplift × 12 | about $5.2M, a 10% lift |
You are halfway through "a 10% revenue lift" when the partner cuts in: "The premium tier needs new equipment and a refit, roughly $150,000 a club. Still a yes?"
The weak answer finishes the previous thought, then says the refit "would need to be considered." The strong answer takes the number and runs it: 40 clubs at $150,000 is $6 million of capex. Assume 70% of the incremental revenue survives as contribution after the extra classes and staff; that is about $3.6 million a year, so the refit pays back in roughly 20 months. "Still a yes, with a 20-month payback."
The partner pushes again: "Your 15% is a guess. Make it 10%." Now 9,600 members trade up, incremental revenue is about $3.5 million, contribution about $2.4 million, and payback stretches to about 30 months. The sentence that gets written down: "At 10% the payback is 30 months, so I would pilot in five clubs for a quarter, risking $750,000 instead of $6 million, and roll out only if uptake clears 12%."
Two interruptions told the partner whether you can hold a number while being talked over, update proportionally rather than collapse, and think like someone spending the client's money. None of that appears on a first-round scoring guide.
Stories that survive the tenth minute
The experience half is where finals are most often lost, and the half candidates practice least. McKinsey's interviewing page asks for two examples in each of four areas; the labels for those areas have changed over the years, so prepare for the substance rather than the label. Bain's recruiters put the substance plainly: "Show us what changed because you were involved."
A first-round interviewer asks two or three follow-ups and moves on. A partner stays on one story for ten minutes and climbs a ladder of probes like this one:
| Rung | The probe | A ready story produces | An unready story produces |
|---|---|---|---|
| 1 | "What exactly was your role?" | A title and the decision that was yours | "I was heavily involved" |
| 2 | "What did the other person actually say?" | A quoted sentence | A summary of their attitude |
| 3 | "What did you try first that did not work?" | A specific failed attempt | "I stayed persistent" |
| 4 | "What did it cost you?" | Time, a relationship, a grade, a bonus | "It was challenging" |
| 5 | "What would you do differently?" | One concrete change | "Nothing, it worked out" |
| 6 | "How do you know it worked?" | A number or a named consequence | "Everyone was happy" |
A story is ready when the sixth rung still produces a fact rather than an adjective. Most candidates arrive with stories that die at rung three. Bring three stories at full depth, none of which you used in round one if the same office is interviewing you, and rehearse them aloud with someone who keeps asking.
The seven days before, and the day itself
The week before a final is a taper, not a training block.
| Day | Do | Do not |
|---|---|---|
| 7 | One full live case, deliberately interrupted; record and time your opening | Start a new casebook |
| 6 | Three stories, six rungs each, with a friend who probes | Write new stories from scratch |
| 5 | One messy case with no clean prompt; invent the structure | Drill frameworks |
| 4 | Read your interviewers' public profiles; write three questions | Prepare a speech about the firm |
| 3 | Opening and close only, three times in a row | Chase a higher score on anything |
| 2 | Stories again, out loud, standing up | Rehearse silently |
| 1 | Logistics, sleep, one 60-second opening said aloud | Any new material |
Bain's what-to-expect page says it "always looks great when candidates know the backgrounds of the people they'll talk to"; that is day four, and it takes an hour.
On the day, each room is a fresh start. Do not recap interview one to interviewer two; they will compare notes without your help. A lunch or coffee with consultants is an interview with a low bar for disqualification and no bar for a hire, so stay in headline mode without going stiff. If you think you bombed a room, you probably did not; partners give nothing away. Reset with water and the next 60-second opening. And do not change your opening style on day three because a forum said a partner is "friendly"; a new style is exactly the variance you are trying to remove.
Practice this today
One assignment. Sit a full voice case with Ned at /interview and, after the first branch of your structure, stop and state the recommendation you would give if the case ended now. Then continue. The seven-score debrief shows whether your opening and close hold when you are forced to commit early, which is what an interruption does.
Then run one story through /behavioral, which asks four follow-ups on a single story. If the fourth follow-up produces an adjective, the story is not ready.
Everything on CoachNed is open for seven days, no card; after that it is $120 for a recruiting season or $49 a month. The synthesis drill is a four-minute daily warm-up for the close.
CoachNed is independent and not affiliated with McKinsey, BCG, Bain, or any other firm named here.
Frequently asked questions
How is a final round case interview different from the first round?
On paper the format is usually the same: a case plus experience questions in each interview. What changes is the interviewer, typically a partner or principal, and the style, which is less scripted, more conversational, and more likely to interrupt. The seniority shift is firm-stated; the change in style is my observation.
How many interviews are in a consulting final round?
It varies by office and role. Two to three interviews with senior people is the common pattern; some offices add a written case or a group exercise, and your recruiter should tell you in advance.
Should I do more practice cases before a final round?
Fewer, better ones. Two or three live cases in the final week, at least one with deliberate interruptions and one with a messy prompt, do more than ten from a casebook. Spend the saved time on stories.
How long after a final round do you hear back?
BCG's brochure says it notifies candidates within a few days of the second round, and in my experience that is typical, though holidays and partner travel stretch it. A week of silence is not a signal either way.
Sources
- BCG Recruiting Interview Preparation Brochure, November 2018 (PDF) — second round "similar to the first round" with "more senior members of the firm"; decision "within a few days." Checked 2026-09-24.
- BCG Consulting Interview Process — current interview stages. Checked 2026-09-24.
- Interviewing at McKinsey — PEI then problem-solving interview; two examples in each of four areas. Checked 2026-09-24.
- Bain: Interviewing — same questions for each role to reduce bias. Checked 2026-09-24.
- Bain: What to Expect — research your interviewers' backgrounds. Checked 2026-09-24.
- Bain: How to stand out in Bain's consulting recruiting process, March 11, 2026 — "build consistency rather than cramming"; "show us what changed because you were involved." Checked 2026-09-24.
- Schmidt and Hunter, Psychological Bulletin, 1998 (PDF) — structured interview validity .51 versus .38 unstructured. Checked 2026-09-24.
- Schmidt, Oh, and Shaffer, working paper, October 2016 (PDF) — revised validity of .58 for both interview types; unstructured interview reliability of .37 and .40 across interviewers and occasions. Checked 2026-09-24.