McKinsey Solve Prep: The 7-Day Practice Plan (2026)

The honest answer is 8–15 hours of focused practice, not the 40 most prep courses imply. Above ~20 hours you're hitting diminishing returns — the assessment rewards pattern recognition you can't grind, and the marginal gain from a 30th replay is roughly zero.

Three realistic schedules depending on how much runway you have:

One free play of each mini-game in a single sitting: Sea Wolf, Red Rock, Sustainable Future Lab. Do not take notes during the games. Do not optimize. Do not pause to Google what a screen means. The point is to feel the actual format under timer pressure — the surprises matter more than the score.

Immediately after, write down two things per game: the moment you felt most lost, and the moment you felt most in control. That's your Day 5 drilling list. Don't replay anything today.

Replay Sea Wolf with full results visible. The 5-step flow — site read, characteristics, filter, categorization, prospect builder — rewards clean ecosystem reasoning more than speed. Watch where you lost points yesterday: usually trait selection or undesired categorization.

Sea Wolf has a 20% penalty per undesired trait, which compounds fast. Two sloppy selections is the difference between mid-tier and top-tier performance on this game alone. The full Sea Wolf walkthrough covers the trait-by-trait scoring logic.

Red Rock is where most candidates underperform, so it gets the longest day. The 35-minute case pulls together five things at once: exhibit reading, draggable value-card prioritization, bi-directional calculator use, case questions, and a final report. Under timer pressure, candidates default to skipping exhibits — which is exactly what the scoring model punishes.

Spend the first hour replaying with full results unlocked, paying attention to which exhibits you misread. Spend the second hour drilling the calculator FAB and forcing yourself to verify every number twice. The most common Red Rock failure mode is committing to a wrong number early and propagating it through the report.

If you want to drill different scenarios, there are six REDROCK seeds available on paid tiers — the variety matters more for this game than for the other two.

SFL is the situational-judgment mini-game that started rolling out in 2025–2026. The 20-minute scenario branches based on your choices, so there's rarely a single "right" answer — McKinsey is reading process patterns across your decisions. Over-drilling SFL produces marginal returns past two hours; this day is intentionally short.

Focus on two things: how you weight competing stakeholders in the priority ranking step, and whether your branch choices stay internally consistent. Wild swings between branches read as random rather than principled.

Pull out the Day 1 notes. Whichever game showed the biggest "most lost" moment is your Day 5 game. Replay it twice with full results unlocked, focusing on the specific failure mode you noted — exhibit misread, trait selection, branch inconsistency, calculator slip. Two focused replays beat five distracted ones.

If two games tied for weakness, pick the one with the biggest scoring weight at McKinsey today: Red Rock first, Sea Wolf second, SFL third.

Run all three mini-games back-to-back, in the same order you expect on test day, with no pause and no notes. This is the day almost no one simulates correctly, and it's the day that matters most. The fatigue compound — Sea Wolf focus, Red Rock math density, SFL judgment — is real, and the only way to calibrate is to live through it once.

Set up your environment exactly like the real assessment: quiet room, laptop (not tablet), stable connection, no phone, no music, water within arm's reach. Treat the result as your baseline rather than your verdict.

No new content. No replays. Re-read your Day 1 + Day 5 notes for ten minutes, then close the laptop. Sleep eight hours. Hydrate. The compounding work is done; cramming today only adds anxiety to a test that already rewards calm process. The most common Day 7 mistake is panic-replaying, hitting a bad seed, and walking in demoralized.

Day 1 runs entirely on the free simulator — one play of each mini-game with no credit card. Days 2–6 are where the free tier hits its limit: you'll want multiple seeds for Red Rock, full scoring breakdowns to see where you actually lost points, and unlimited replays for targeted drilling.

For a head-to-head comparison against the other major prep tools, see our best McKinsey Solve prep tools for 2026 guide. Pricing details live on the pricing page.