Is the McKinsey Solve Hard? What the Data Actually Says

The difficulty in one number McKinsey won't give you

McKinsey doesn't publish pass rates. Most estimates put the failure rate around 70% — treat that as the informed consensus, not gospel. What we can measure precisely is how candidates perform in practice, and the picture is consistent: the test is hard in specific, fixable places, and most candidates meet those places for the first time on test day. In our study of practice data, 70.3% of candidates with a stated test date didn't start practicing until the final 48 hours — the median gap between someone's first run and their real test was one day the full study. The Solve is hard; the way most people approach it makes it harder.

What's actually hard about it

Three things, in the order they hit you. The format, first: game mechanics you've never operated, explained briefly, scored immediately. There's no warm-up round on the real thing. The clock, second: every game is tight enough that pacing errors compound — spend four extra minutes exploring in the Red Rock study phase and the report phase pays for it. The invisible layer, third, and this is the one almost nobody prepares for: the Solve reads your process, not just your answers. The single worst-performed component we measure across all three games is consistency — in the Sustainable Future Lab, candidates lose 60.7% of the available consistency points (95% CI: 59.4–62.0). You can't grind consistency the night before; it develops across runs, which is precisely why a one-shot format punishes it.

Which game is hardest?

By improvement curves — how much better candidates get with practice, which is a decent proxy for how much the game punishes unfamiliarity — the Red Rock Study shows the steepest curve: repeat runs score a median 20.3 points higher than first runs. It's the most "learnable," which means walking in cold costs you the most there. Sea Wolf is the most mechanical: once the selection logic clicks, it stays clicked (smallest repeat gain, +6.7). The Sustainable Future Lab is the sneaky one — modest score gains with practice (+10.4), but it's where the consistency losses concentrate, and it's the game candidates most often don't know is fully scored. (We adjust all of these for regression to the mean with a placebo test and publish both numbers — roughly half of each raw gain survives. The improvement is real; we just refuse to oversell it.)

Hard, but not a lottery

Some candidates pass cold. That's true, and it would be dishonest to pretend otherwise. But "some people pass cold" is the same sentence as "most estimates say around 70% don't pass at all" — the question is which population you'd rather bet a one-shot, career-gating test on. The Solve is hard the way anything unfamiliar-under-pressure is hard, and unfamiliarity is the one difficulty component that's entirely optional practice all three games in the real format.

Common questions