The recruiting funnel

#The five gates

Across McKinsey, Bain, and BCG, the funnel runs the same five stages, with firm-specific detail inside each: resume/GPA screen → online assessment → first-round interviews (typically two, each pairing a case with a fit component) → final-round interviews (typically two to three, often with partners, higher-stakes) → offer. [Directional] — this five-stage shape is consistent across every firm-specific guide surveyed for this course, but no firm publishes it as a formal named sequence; it's a synthesis of consistent secondary reporting, not a quoted firm document.

#Stage 1 — the resume/GPA screen

This stage is fast and heuristic, not holistic: recruiters report spending well under a minute on a first pass, with school, GPA, major, and standout achievements read as signals almost simultaneously rather than weighed one at a time. [Directional] None of the three firms publishes a formal minimum GPA. The practically observed floor at US target schools clusters around 3.5–3.6/4.0, but a lower GPA does not automatically disqualify a candidate — a markedly quantitative major, an unusually rigorous course load, or an exceptional achievement elsewhere (a nationally competitive venture, a technical publication, elite athletics) can offset it. [Directional] After this stage, GPA stops mattering much — case and fit performance decide the rest of the funnel. [Directional]

#Stage 2 — the online assessment

Each firm runs its own, and as of 2026 none of the three still relies on a plain multiple-choice aptitude test alone:

• McKinsey Solve (formerly branded Imbellus, built by a startup of that name before McKinsey absorbed it after Roblox acquired Imbellus's IP in 2020) is a game-based, adaptive assessment. It presents scenario-based problems — commonly reported modules in 2026 include "Redrock," "Sea Wolf," and "Sustainable Futures Lab" — that track not just whether you reach a correct outcome but which data you chose to examine and in what sequence, explicitly grading process over answer. [Established — corroborated firm-history detail] for the Imbellus/Roblox/McKinsey lineage; [Directional] for the specific current module names and invitation-length patterns, which come from aggregated candidate reports rather than a McKinsey publication.
• Bain's SOVA assessment — delivered via the TestGorilla platform as of 2026 — runs roughly 75 minutes and covers numerical, verbal, and logical reasoning alongside a personality/situational-judgment component; it is not strictly timed per question, but response speed is scored. [Directional]
• BCG's "Casey" is a chatbot-delivered online case, commonly reported as 25–35 minutes across 6–10 mixed questions, usually paired with a roughly one-minute recorded video response. BCG has not published an official description of this tool; everything known about its exact format comes from aggregated candidate reports. [Directional]

#Stage 3 and 4 — first round and final round

First-round interviews typically pair one or two live cases with a shorter fit component; final rounds add more cases, often with more senior interviewers (partners at the final stage), and are where firm culture and long-term potential get weighed more heavily than raw technical execution. [Directional] Fit and the PEI covers the behavioral half of both rounds in detail.

#The numbers that get repeated — and why this course won't repeat them as fact

Case-prep marketing routinely cites precise figures: an overall MBB acceptance rate under 1–1.5%, a 10–30% first-round-to-final-round advance rate, a 20–30% final-round-to-offer rate, and a claim that roughly 90% of US MBB hires come from a defined list of "target schools." Checking the most detailed version of this table found in this research traced its methodology to a mix of "firm disclosures" (unnamed), "recruiter interviews" (unnamed), and an "alumni survey" of stated size n=847 conducted by the publishing platform itself — and the same page explicitly concedes "firms don't publish stage-by-stage data" and that its own figures are "estimates." [Speculative] — treat every stage-by-stage percentage in this space as directional flavor, not a number to plan against; no firm has published its own funnel conversion rates, and this course could not find one that has.

One figure has firmer footing: McKinsey told Business Insider in 2024 that it planned to hire roughly 1% of applicants, consistent with 2023 — a specific number attributed directly to the firm rather than to an aggregator's survey. [Established — firm statement reported by a named outlet] The applicant-volume figure it's built on is inconsistent across sources, ranging from roughly 200,000 to over 1,000,000 applications a year depending on the source and year cited, which this course could not reconcile to one number. [Speculative]

#Target versus non-target, and what actually moves the needle from a non-target school

The consistent qualitative pattern — not the specific percentages — is well corroborated: candidates from outside a firm's defined target-school list get interviews almost exclusively through a warm connection at the firm (a referral from a current or former consultant, sustained coffee-chat contact, or an alumni connection), not through cold online applications. [Directional] A genuine MBB referral is widely reported to meaningfully raise screening odds, though no source in this research quantified that lift with a stated methodology; a "2–4x" figure appears in some secondary content but without traceable sourcing, so it's flagged rather than repeated as fact. [Speculative] The practical implication, independent of any specific multiplier: for a non-target candidate, the networking track in The realistic timeline is not optional extra credit — it is very likely the actual gate, and case-interview polish cannot substitute for it.

#International candidates and visa mechanics

All three firms sponsor H-1B visas for at least some hires, typically bridging a candidate through F-1 Optional Practical Training (OPT) or STEM OPT — which grants 12 to 36 months of authorized US work — while the H-1B petition is filed. [Directional] Sponsorship is not guaranteed even for a strong candidate: the H-1B is allocated by an annually random federal lottery outside any firm's control, and firm policy on what happens after a lottery miss differs — Bain has been reported to file once and not re-file after a loss, where some peer firms will re-apply the following year. [Directional] — this is a specific firm-policy claim from secondary reporting, not confirmed against a Bain primary source in this research pass; verify current policy directly with a specific office before treating it as settled. The practical takeaway for a non-US candidate: confirm a target office's current sponsorship track record and policy directly during the process rather than assuming it, and treat "does this office reliably sponsor" as a qualifying question worth asking in a coffee chat, not a detail to discover after an offer.