Redrock Study
35-minute fixed timer spanning all four phases (the clock never stops).
What it tests. Data synthesis, weighted averages, compound changes, chart conceptualization and fast estimation.
Worked example. Iguanas eat 1,200 lbs/week; a vine cuts foliage 15% monthly while 180 lbs of feed is added; with a 900 lb/month collapse threshold, find the net surplus at end of Month 2 (answer: 2,748 lbs surplus).
Common traps. The information saturation trap - McKinsey floods the screen with irrelevant data columns to make you waste time reading linearly.
How to handle it. Search only for the metrics the objective dictates; use the calculator history and drag unrounded numbers into answer fields.
Sea Wolf (ocean cleanup)
Roughly 30 minutes.
What it tests. Algorithmic execution, mental-math agility and tracking multi-variable constraints under time pressure.
Worked example. To hit an average toxin-absorption between 7.0 and 7.5 across three microbes, multiply the target range by 3 and find a combination summing to 21.0-22.5 (e.g. 8.5 + 5.0 + 8.0 = 21.5, average 7.16).
Common traps. Neglecting a prohibited negative trait (e.g. an "Anaerobic Mutation" is an absolute dealbreaker even if the numbers pass) and bouncing eyes back and forth to re-check thresholds.
How to handle it. Eliminate negative-trait options first, then multiply the target range by 3 to avoid slow long division.
Sustainable Futures Lab (SJT)
20 minutes fixed.
What it tests. Stakeholder management, navigating ambiguity, prioritization, integrity and corporate diplomacy.
Worked example. Allocating a final $500K surplus, the best answer synthesizes competing community, sponsor and data-team objectives into one higher-value outcome rather than splitting evenly or capitulating to the sponsor.
Common traps. The "Hero Option" trap - picking extreme idealistic choices (halting a whole project over a minor issue) because you assume McKinsey wants perfection.
How to handle it. Read the whole prompt, identify the binding constraint, gather objective facts first, and never compromise data integrity for political convenience.