Skip to content
CaseStar
CaseStarTry Red Rock free

CaseStar field guide · Solve scoring

McKinsey Solve score and decile: what is actually known

Most pages about Solve scoring state guesses as facts. McKinsey has published almost nothing about how Solve is scored. This page separates what McKinsey says, what many candidates independently report, and what is speculation, so you can decide what to prepare for.

What does McKinsey say about the Solve score?

McKinsey Solve scoring: published versus reportedMcKinsey has not published its scoring method or a pass mark and has not published what the result contains. Candidates report that answers and approach both matter, that results are deciles relative to peers, that the result email has no score, and that a raw score and decile can be requested.What McKinsey publishesWhat candidates reportScoring methodNot publishedAnswers and approach both matterPass markNot publishedNo fixed cut-off; deciles vs peersYour resultNot publishedEmail has no score; decile on request
Only the left column comes from McKinsey. Treat the right column as candidate-reported.

McKinsey’s public material says Solve is a problem-solving assessment, that preparation is not required, that each task begins with an untimed tutorial, that scenarios are generated so no two candidates receive identical ones, and that results are considered alongside the rest of your application. It may also ask you to complete a short follow-up verification task. It does not publish a scoring model, a pass mark or a percentile table.

What score and decile do candidates report?

Candidate-reported facts about Solve scoring
ClaimHow well it is corroborated
The result email contains no scoreWidely reported
Recruiters share a raw score and decile when askedWidely reported (at least eight candidates in 2026)
Raw scores cluster between about 499 and 558; the scale seems to top out near 560Widely reported
Deciles are equated by role, region and test versionWidely reported
Results arrive in one to three weeks, occasionally up to sixWidely reported
A result is valid for about twelve months; a retake comes only with a new application, typically 12–24 months laterReported; varies by office
Only 20–30% of candidates advanceReported, not verified

Reported pairs include 558 in the 10th decile, 530 in the 9th, 510 in the 6th and 499 in the 4th. That spread matters: a few points of raw score can move you several deciles, so every answer counts.

If you sat Solve and were rejected, it is worth asking the recruiting inbox for your score and decile. Knowing whether you were near the line or far from it changes whether reapplying in the next cycle is worth planning for.

What is the difference between the product score and the process score?

The idea of a product score (correct answers) and a process score (how you worked) is repeated by nearly every prep provider. It is plausible: Solve logs your interaction with the task, and McKinsey describes Solve as measuring how you solve problems, not just whether you finish. But the specific signals and their weights have never been published. Lists of “process metrics” you will read elsewhere, such as time per question, what you save to the journal, how often you revisit exhibits, are reasonable guesses by vendors, not facts.

For Red Rock this means: accuracy is what you can control and measure, and a clean process (collecting deliberately, not flailing between screens, finishing all six cases) is how you get accurate answers within 35 minutes. For Sea Wolf, the per-site effectiveness shown after each treatment is the clearest visible outcome, and candidates who report top deciles describe finishing one site at 80%.

How should the scoring change your preparation?

  • Finish everything. Unanswered Red Rock cases cannot score. Pace the study so you reach all six.
  • Be accurate before you are fast. A wrong answer delivered quickly is still wrong, and in Red Rock it can cascade into the report.
  • Practise deliberate collection. Whether or not it is scored, it makes Analysis faster.
  • Do not chase a practice-tool percentile. Use practice scores to find your errors, not to predict your decile.

CaseStar’s McKinsey Red Rock simulator reviews every answer and how you used your time. That review is CaseStar’s practice model, clearly labelled as such. See the Red Rock Study guide for the format and the 2026 change log for everything that moved this year.

Common questions

What is a good McKinsey Solve score?

There is no published pass mark. Candidate-reported raw scores in 2026 cluster between about 499 and 558, and those who reported 9th or 10th deciles reported about 520 or higher. Deciles are relative to comparable candidates, so the same raw score can map to different deciles.

Does McKinsey tell you your Solve score?

The result email does not include a score. Many candidates report that emailing the recruiting inbox named in their invitation gets them a raw score and a decile. McKinsey has not published a policy on this.

What is the process score in McKinsey Solve?

A widely repeated description says Solve scores both your answers (product) and how you reached them (process). McKinsey has never published which behaviours count or how they are weighted. Any detailed “process score” formula, including practice-tool scores, is a model, not McKinsey’s algorithm.

Can I retake McKinsey Solve?

McKinsey does not publish a retake policy. Candidates describe one attempt per application cycle, with a result valid for about twelve months and a retake only when you reapply, typically 12 to 24 months later depending on office and role. Confirm with your recruiter.

Does a CaseStar score predict my Solve result?

No. CaseStar grades your practice against scenarios we generate and our own model of good process. It helps you find and fix errors; it is not McKinsey’s scoring and does not predict whether you will pass.

Sources and limits

CaseStar (casestar.io) is an independent practice platform for consulting recruiting: timed McKinsey Solve simulators, including Red Rock, and voice case interviews. It is not affiliated with McKinsey & Company.

McKinsey publishes only a broad description of Solve. Details labelled as candidate-reported come from debriefs and can differ between candidates, regions and test versions. Worked examples on this page are original CaseStar exercises, not assessment questions. Your invitation and recruiter take precedence.