Consulting digital assessments are the online screens firms run between the application and the interview rounds. They fall into three families: scenario simulations that put a candidate inside a constructed problem, behavioural game batteries such as Pymetrics that build a trait profile from short timed tasks, and aptitude or situational tests such as those supplied by SOVA that combine reasoning sections with judgement and personality questions. Each family rewards something different, and preparing for the wrong one is a common and avoidable waste.
Which screen you will sit depends on the firm, the office, the role and the intake. Your invitation is authoritative; treat every general description, this one included, as a map rather than a timetable.
The three families at a glance
| Family | Examples | What it measures | Does practice help? |
|---|---|---|---|
| Scenario simulation | McKinsey Solve, BCG Casey | Problem-solving process, data interpretation | Process practice helps; content does not |
| Behavioural game battery | Pymetrics | Risk tolerance, memory, attention, reciprocity | Very little, by design |
| Aptitude and situational | SOVA, numerical and logical reasoning tests | Reasoning speed and accuracy, judgement | Substantially |
The two scenario simulations have guides of their own — McKinsey Solve and BCG Casey. This guide covers the other two families.
Behavioural game batteries: Pymetrics
Pymetrics is a battery of short games, usually taking twenty-five to thirty-five minutes in total, that measures behavioural and cognitive traits from how a candidate plays rather than from self-report. Typical tasks include inflating a balloon for reward with a risk of bursting it, splitting or accepting sums of money with another party, pressing or withholding a key as symbols appear, recalling lengthening digit sequences, and identifying emotion from facial expressions.
The traits these produce — risk appetite, impulse control, working memory, attention, fairness, effort under uncertainty — are then compared against a benchmark profile built from assessments of people already succeeding in the role at that firm.
There are no right answers, and this is the important part. The screen is a fit comparison, not a score. Two consequences follow:
- Trying to perform a profile is the main risk. Candidates who decide that consultants must be risk-tolerant and inflate every balloon to bursting produce an inconsistent profile, because several games cross-check the same trait from different angles. Inconsistency reads as noise, and noise does not match a benchmark.
- Conditions matter more than strategy. Sleep, a quiet room, a working trackpad or mouse, and a single uninterrupted sitting affect reaction-time tasks measurably. That is the part of your result you control.
Reasonable preparation is limited: read each game’s instructions carefully before it starts, since several games change their rules partway through and the instruction screen is where that is disclosed; play the practice round properly rather than clicking through it; and do it when you are alert rather than at the end of a working day.
If you are asked to sit the same battery for a second firm within a short period, some providers reuse your prior result. That is normal and not something to work around.
Aptitude and situational tests: SOVA and similar
SOVA is one of several psychometric providers whose assessments consulting firms use as an early screen. Its format typically combines a small number of components in one sitting:
Numerical reasoning. Data presented in tables or charts, with questions requiring extraction and calculation under a per-question time limit. The arithmetic is not difficult; the constraint is speed and accurate reading.
Logical or inductive reasoning. Abstract sequences where the task is to identify the rule and apply it. No domain knowledge involved.
Situational judgement. Short workplace scenarios with several possible responses, to be rated or ranked. These are scored against the firm’s own view of good practice.
Personality items. Self-report statements, usually agree-to-disagree, building a behavioural profile.
This is the one family where preparation reliably moves the outcome.
Preparing for the reasoning sections
Practise under the real time constraint. Most people can answer numerical reasoning questions correctly with unlimited time. The skill being trained is doing so in roughly a minute, which is a different skill and only improves with timed practice.
Read the question before the exhibit. Knowing what you are looking for makes reading a dense table far faster than absorbing it first and then searching.
Learn the standard traps. Percentage change against the wrong base, indexed values treated as absolutes, a total that includes a category the question excludes, and units that differ between the table and the question. These recur across providers because they are the reliable discriminators.
Do not leave blanks unless the test penalises guessing. Most do not. Confirm in the instructions, and if there is no penalty, answer everything.
The arithmetic underlying all of this is the same arithmetic the case rounds need, so the case maths techniques guide is directly useful preparation.
Answering situational judgement items
Situational judgement questions look like personality questions and are not. They are scored, and the scoring reflects how the firm actually expects consultants to behave. A few patterns hold across firms:
- Raise problems early rather than absorbing them. A response that quietly works late to fix a slipping deliverable usually scores below one that flags the slippage to the engagement manager and proposes options.
- Involve the people affected. Responses that consult the client or the team before deciding generally score above unilateral action, even competent unilateral action.
- Do not confuse decisiveness with judgement. The most assertive option is rarely the highest-scoring one; the option that gathers the missing information first usually is.
- Do not escalate over someone’s head as a first move. Going directly to a partner about a peer’s work scores poorly nearly everywhere.
The best-scoring answer is usually the one a thoughtful consultant would defend in a debrief, not the one that resolves the scenario fastest.
The personality section
Answer it consistently and honestly. These sections contain repeated items phrased differently to detect inconsistency, and an inconsistent profile is a worse outcome than an unremarkable one. There is also little to gain from constructing an ideal answer, because the traits a firm benchmarks against are not the ones candidates typically guess.
Where these screens fit
All of these are gates. None of them is the decision, and none replaces the case and fit rounds that follow. The most common preparation error is proportional rather than tactical: spending weeks on an assessment worth a pass-or-fail and arriving at the interview rounds underprepared for the part that actually decides the outcome.
A reasonable allocation is to give the screen enough attention to remove avoidable failure — read the invitation, know the format, practise timed reasoning if the screen includes it, and set up the room properly — then return to case and fit preparation.
For that preparation, the case interview hub sets out the sequence and the fit interview hub covers the behavioural round. The firm profiles describe how each of McKinsey, BCG and Bain structures the rounds behind the screen.
