What "prepare" actually means for McKinsey Solve
Solve is not a knowledge test. There's no syllabus, no formulas to memorise, no vocabulary list. Preparation in the traditional sense — reading textbooks, drilling practice problems — produces almost zero score gain on the actual assessment.
What does produce score gain is three things, in order: pattern fluency (you've seen enough versions of each mini-game that the format no longer surprises you), habit calibration (the right tactical moves — exhibit-reading, calculator use, trait checks — fire automatically under timer pressure), and logistics control (your test-day environment isn't actively working against you).
The methodology below is built around producing those three outcomes — nothing else. If a prep activity doesn't measurably contribute to one of them, it's noise.
Brand new to the format? Start with the digital assessment overview before working through this page.
The 4-phase preparation framework
Every effective Solve prep loop has the same four phases. The phases stay the same whether your runway is two weeks or two months — only the time allocation per phase changes.
Phase 1
Diagnose
Take cold free plays of each mini-game and record where the format surprises you.
Time: ~90 min · Week 1
Ready when: You can name your weakest mini-game and the two screens you got most lost on.
Phase 2
Target
Translate the diagnostic into a prioritized practice plan — weakest game gets most reps.
Time: ~30 min planning · Week 2
Ready when: You have a written allocation: hours per game, ordered by gap size.
Phase 3
Drill
Build pattern fluency on the weakest game and tip-isolation reps on the others.
Time: 6–10 hrs · Week 2–3
Ready when: You finish each game without surprises and your scores have plateaued.
Phase 4
Calibrate
Full back-to-back dress rehearsal + logistics setup + sleep. Don't undo your work.
Time: ~2 hrs over 3 days
Ready when: Dress rehearsal finished without panic. Test-day logistics already locked.
Phase 1 · Week 1 · ~90 min
Diagnose — find your real weak spots
One cold free play of each mini-game in a single sitting. Do not optimise. Do not pause to Google. Do not take notes during the games. The point of Phase 1 is data collection, not score maximisation — a "good" Phase 1 score actually tells you less than a bad one, because it hides where your real gaps are.
Immediately after each game, write down two things: the moment you felt most lost, and the moment you felt most in control. Those notes become the input for Phase 2. Do not replay anything. Do not check answers. Do not move to Phase 2 yet.
Free plays require a free account. Sign up here — no card, takes 30 seconds.
Phase 2 · Week 2 · ~30 min planning
Target — allocate by gap size, not by gut
Look at your Phase 1 notes. Rank the three mini-games from weakest to strongest. The weakest game gets the most reps; the strongest gets maintenance only. This is the single highest-leverage planning move you can make — most candidates allocate equally across all three games, which is roughly the worst possible split.
Use this allocation table as your starting point:
| If your weakest is… | …and your strongest is… | Recommended time split |
|---|---|---|
| Sea Wolf | SFL | Sea Wolf 50% · Red Rock 35% · SFL 15% |
| Red Rock | Sea Wolf | Red Rock 55% · SFL 30% · Sea Wolf 15% |
| SFL | Red Rock | SFL 45% · Sea Wolf 35% · Red Rock 20% |
Once allocation is set, pull the relevant tactical fixes for each game from the 17 tips page — these become your drill checklist for Phase 3.
Phase 3 · Week 2–3 · 6–10 hrs
Drill — pattern reps + tip isolation
Phase 3 is where the actual practice happens. Use two drill modes — never one without the other:
- Pattern reps. 3–5 plays of the same mini-game back-to-back on your weakest game. The goal isn't variety — it's building the visceral "I've seen this" recognition that kills format surprise on test day.
- Tip isolation. One play of any mini-game with a single tip from your Phase 2 checklist held front-of-mind. Don't try to fix everything at once. One play, one tip, drilled until it's automatic.
Scale the cadence to your runway:
| Runway | Cadence | Total | Note |
|---|---|---|---|
| 2 weeks | 1 session every 2 days, 45–90 min each | ~10 hrs | Compress Phases 1–2 into the first 3 days, drill for 8, calibrate in the last 3. |
| 4 weeks | 2 short sessions per week + 1 longer weekend block | ~12 hrs | The sweet spot. Lets you space drills enough for retention without losing sharpness. |
| 8+ weeks | 1 session per week, plus a 4-week intensive at the end | ~14 hrs | Don't over-front-load. Diminishing returns past week 4 — save real intensity for the final stretch. |
Game-specific mechanics for your drill sessions: Sea Wolf · Red Rock · Sustainable Future Lab. If you're deciding which paid tool to use, the 2026 tools comparison covers options.
Phase 4 · Final 3 days · ~2 hrs total
Calibrate — protect what you've built
Phase 4 has one job: don't undo the work you did in Phase 3. The compounding has already happened. From here forward, more reps cost more than they pay.
The three things to do:
- One full back-to-back dress rehearsal in a continuous ~90-minute block. Treat it like the real test — phone away, no pausing, no Googling. Score the gap between this run and your Phase 1 baseline.
- Lock your test-day logistics 48 hours out: laptop check, stable connection confirmed, calendar block with a 15-min buffer either side, water, phone in another room, bathroom break before.
- Sleep on the day before. Day-of cramming has a strongly negative correlation with score. A calm 8 hours beats a panic-replay every time.
The 7-day prep plan Days 6–7 are the calibrate phase as a literal calendar.
Common McKinsey Solve preparation mistakes
Each of these is a recurring pattern in candidates who put in the hours and still underperform. The methodology above is designed to prevent every one of them — but it only works if you don't backslide.
Over-prepping past 20 hours
Marginal score gain past ~20 hours of focused practice is roughly zero. Past that point you're burning sharpness, not building skill.
Neglecting Red Rock because it 'feels like a normal case'
Red Rock is where most score loss happens. It looks familiar, so candidates under-drill exhibit volume and calculator discipline. Treat it as your default priority unless your diagnostic clearly says otherwise.
Drilling outdated Imbellus content from YouTube
The 2019 Imbellus ecosystem case differs from 2026 Sea Wolf — different traits, no terrain bonus, generic site labels. Watching old playthroughs trains habits you then have to unlearn.
Cramming the night before
A bad seed at 11pm anchors you on failure. Phase 4 calibrate is intentionally light: review notes, sleep, hydrate. The work is already done.
Practicing on mobile or tablet
Solve is built for keyboard + trackpad + full screen. Tablet practice burns time on rendering quirks that don't exist on the real test environment.
FAQ
How long does it really take to prepare for McKinsey Solve?
Most candidates who pass spend 8–15 hours of focused practice on Solve specifically. Below 6 hours is risky on a format you've never seen; above 20 hours hits diminishing returns. The methodology on this page scales to that range — 2 weeks at moderate effort, or 4 weeks at low weekly load, both land in the sweet spot.
Can I prepare for McKinsey Solve without paid tools?
Yes, to a point. One cold free play of each mini-game (Phase 1 diagnose) plus reading every guide on this site costs zero and gets most candidates 70% of the way. The remaining 30% is repetition variety — paid practice exists to give you fresh seeds, not to teach you anything you can't learn free.
Is McKinsey Solve preparation different for consultant vs business analyst candidates?
The assessment itself is identical. The preparation methodology is identical. The only practical difference is calibration expectations — Associate/Consultant candidates are typically benchmarked slightly higher on the process-signal side than BAs, so spend a little more time on the calibration phase if you're at that level.
Do I need to learn coding or statistics to prepare for McKinsey Solve?
No. The 2026 Solve format requires only arithmetic, basic ratio reasoning, and reading exhibits. There is no programming task, no statistical inference, and no data-science content. Time spent on Python or stats for Solve specifically is time wasted.
How do I know I'm 'ready' for the real assessment?
Three signals: (1) You finished one full back-to-back dress rehearsal without panicking on any single screen; (2) You can name your weakest mini-game and the specific habit fix you've drilled for it; (3) Your logistics are set — laptop, environment, calendar block, sleep. If all three are true, additional reps don't help.
Start with the diagnostic. Everything else follows.
Phase 1 is one cold free play of each mini-game. It takes 90 minutes and tells you more about where to spend your prep time than any guide can. Sign up to start.