McKinsey Redrock Study: The Complete 2026 Guide

What is the McKinsey Redrock Study?

The Redrock Study is the first module in McKinsey Solve, the game-based assessment McKinsey sends to most candidates after resume screening and before live interviews. You play a researcher on Redrock island. You are handed a research objective, a page of study material full of text, tables, and charts, and a clock. Your job is to pull the right numbers out of the noise, run the math, and report your findings before time runs out.

Unlike the puzzle-style games that came before it, Redrock is closer to a compressed data analyst exercise than a video game. Roughly 60 to 70 percent of the questions are quantitative: percentages, ratios, weighted averages, and growth rates, all with simple arithmetic but very little time. The remaining questions test reading precision, chart selection, and whether you understood the research objective.

Depending on your region and role, your Solve invitation will be one of two versions. The 65-minute version pairs Redrock with Sea Wolf. The 85-minute version adds the Sustainable Futures Lab as a third game. Either way, Redrock comes first, and an untimed tutorial precedes it. You cannot choose your version: check the total time stated in your invitation email.

The Redrock format: one Study, then six Cases

Redrock has two parts sharing a single 35-minute clock. Part 1 is the Study, a connected research task in three sequential phases. Part 2 is the Cases section, six short and independent mini cases. Once you confirm your Report at the end of Part 1, you cannot return to it.

Part 1, Phase 1: Investigation

You see the research objective and about a page of material: a few paragraphs of text, two or three tables and charts, and hints toward the formulas you will need. Specific values in the material are collectible. Clicking or dragging them adds them to your Research Journal as labeled data cards. The trap is that far more data is collectible than you need. The objective tells you exactly which comparisons matter, so collect against the objective, not against the page.

Part 1, Phase 2: Analysis

Two to four quantitative questions, usually three, built on the data you just collected. Most are fill-in-the-blank rather than multiple choice, which means no answer options to sanity-check against. The on-screen calculator becomes available for the first time, and journal data can be dragged directly into it. If you collected the wrong data in Investigation, you can still navigate back, but the clock keeps running.

Part 1, Phase 3: Report

Two components. First, a written summary with blanks to fill: numbers and conclusions drawn from your analysis. Second, a visual task where you select the chart type that best presents a given finding and enter the data into it. Picking between a bar, line, or pie chart sounds trivial, but under time pressure many candidates lose easy points here. Once you confirm the Report, Part 1 locks permanently.

Part 2: The six mini cases

Six independent cases, each a tighter information package: a short text, one or two exhibits, one question. The skills mirror the Study, but the cycle is faster: identify the relevant data, set up the calculation, execute accurately, move on. The calculator is almost always available; the journal sometimes appears. Because the cases are independent, a mistake in one never contaminates the next, and a blank answer is the only guaranteed zero.

The Research Journal and the calculator

Two tools define the Redrock interface, and knowing how they behave before test day is half the preparation.

The Research Journal is your workspace. Every piece of data you collect during Investigation appears as a card with a title and description. You can reorder cards and rename them, and renaming is worth the seconds: "Y1 wolf population" beats "Table 2 value" when you are hunting for an input two phases later.

The calculator is a simplified on-screen model: the four basic operations and little else, no powers, no memory functions you would find on a real handheld. It has everything Redrock requires and nothing more. Two behaviors to internalize: it only appears from the Analysis phase onward, and journal values can be dragged straight into it, which is faster and safer than retyping under pressure.

Try a Redrock-style question

This is a simplified, original example built in the style of an Analysis-phase question. Collect the data you need into the journal, then compute the answer. In the real test you would have done the collecting during Investigation, with the clock running.

In the real Analysis phase this would be one of three questions sharing the 35-minute clock with everything else, and the data would already be sitting in your journal, correctly or not.

How the Redrock Study is scored

McKinsey evaluates Solve on two dimensions, and most candidates only prepare for one of them.

The product score is what you would expect: did you get the answers right. Accuracy in the Analysis questions, the Report blanks, the chart selection, and the six cases.

The process score is recorded from how you work: what you collect into the journal and what you leave behind, how you navigate between phases, how you allocate time, and whether your approach looks deliberate or scattered. Hoovering every collectible value into your journal does not just waste time, it leaves a trace.

Time strategy: pacing 35 minutes

The math in Redrock is not hard. The clock is. A workable allocation gives roughly two thirds of your time to the Study and one third to the Cases, and respects that the Cases at the end are the easiest points on the test to leave uncollected.

If you fall behind

Falling behind in Investigation is the most common failure pattern. The fix is mechanical: stop collecting, advance with what you have, and trust that you can navigate back during Analysis if a specific input is missing. In the Cases, never leave a blank. A 30-second educated estimate has positive expected value; an empty field is a guaranteed zero on both scoring dimensions.

The six most expensive Redrock mistakes

1. Collecting everything in Investigation

   The material is salted with plausible distractor figures. Collecting against the page instead of against the objective burns minutes and muddies your journal for the phases that follow.

2. Ignoring the research objective

   The objective names the exact comparison being studied, such as Year 1 versus Year 5. Data from other years or other units is there to be left behind.

3. Unit and timeframe traps

   Mixing monthly with annual figures, or percentages with absolute counts, is the most common source of wrong Analysis answers. The arithmetic was fine; the inputs were not.

4. Sinking into one Analysis question

   With three questions and roughly seven minutes, any single question consuming four of them is an emergency. Flag it, answer your best, return only if time allows.

5. Rushing the Report

   The chart-selection task is easy marks lost to haste. Match the chart to the message: composition suggests pie, trend suggests line, comparison suggests bars.

6. Leaving a case blank

   The six cases are independent and roughly equally weighted. A blank is the only outcome that guarantees zero. Estimate, answer, move on.

How to prepare for the Redrock Study

Redrock rewards two separate skills: fast, accurate consulting math, and fluency with an unfamiliar interface under time pressure. Train both.

1. Rebuild your mental math

Percentages, growth rates, ratios, and weighted averages cover nearly everything Redrock asks. Daily ten-minute drills beat occasional long sessions. Our free math drills are built around exactly this question mix.

2. Practice reading charts against an objective

Take any chart-heavy article or report and give yourself a one-line objective before reading. Practice extracting only the two or three values the objective requires and articulating why everything else is noise.

3. Run the format under real time pressure

At least twice: once untimed to learn the mechanics of the journal, the calculator, and the phase flow, and once against the clock to learn what 35 minutes actually feels like. Use the untimed official tutorial on test day as a warm-up, not as a first encounter.

4. Rest before test day

Redrock is a sustained-attention test as much as a math test. One attempt per application cycle means treating the sitting itself with the respect of a final interview: rested, undistracted, stable internet, full screen.

Redrock Study: frequently asked questions

The rest of the Solve assessment