How to Ace McKinsey’s Digital Assessment

Imagine that you are asked to design an ecosystem populated by different kinds of plants and animals. You need to find a mix of species that will co-exist without dying out. How will you decide what to include in your menagerie?

This is McKinsey’s Digital Assessment: the ground-breaking video game that the firm uses to assess candidates’ cognitive abilities.

To win the game, you must understand the many factors that affect the outcome. You must develop a solution, test it, and adapt it based on the feedback you receive.

The Promise: an Unbiased Way to Identify Candidates with Strong Cognitive Abilities

Every year, McKinsey has to sift through over a million applicants and invite those who are most likely to succeed in interviews.

Screening resumes and looking at academic backgrounds is a good start. But it’s biased towards candidates who are good at memorizing or come from a privileged background, at the expense of less privileged candidates with great problem-solving skills.

In response to these biases, McKinsey has used a Problem Solving Test for several years. Although it does provide useful data points about problem-solving skills, the PST has its own prejudices. It favors candidates with high processing speed and familiarity with the problems being tested, and it can be mastered through practice.

McKinsey is now introducing the Digital Assessment to assess candidates’ cognitive abilities in an environment free of biases linked to candidates’ socio-economic background, familiarity with business problems, memory, and preparation.

The assessment was developed by a start-up called Imbellus and was first piloted in 2017. It is now being rolled out globally. Eventually, this will allow McKinsey to rely less on a candidate’s academic background when gauging their potential.

Five Key Cognitive Abilities

The Digital Assessment tests five cognitive abilities:

1. Critical thinking: the ability to form a rational judgement from a set of facts
2. Decision making: the ability to select the best course of action among several options
3. Meta-cognition: the ability to use strategies to make learning information and solving problems easier (e.g., testing hypothesis, taking notes)
4. Situational awareness: the ability to determine the relationships between different factors and to project the outcome of a scenario
5. Systems thinking: the ability to understand cause & effect relationships involving several factors and feedback loops (e.g., anticipating several orders of consequence)

The game captures all of the candidate’s actions (e.g., every movement of the mouse, hesitations, time taken to do an action) and uses data science to score the five abilities. It’s as much about how close candidates come to the optimal solution as about how they think the problem through.

How The Test Works

You have 71 minutes to go through the tutorial and solve five tasks within two scenarios. The way you divide your time between tasks is not set, so if you spend more time in one area you will have less time for others. The time you spend reviewing the tutorial is not counted, so take as long as you need to review instructions carefully.

Imbellus says that they aim to reduce testing anxiety by leaving sufficient time. Recent test-takers have indeed reported that they finished with 5-10 minutes to spare at the end. Time can be tight for some of the tasks but candidates are given indicative time guidance for each task.

The test does not require making calculations beyond additions and subtractions.

Recent candidates have reported experiencing two types of scenarios: ecosystem creation and organism protection.

The Ecosystem Creation Scenario

In this scenario, you’re asked to create a stable marine or terrestrial ecosystem. You will choose the location and the species that will inhabit your ecosystem and aim to have as many of them survive. This will test your systems thinking!

In the example of a mountain range environment, you will have location variables such as elevation, wind speed, soil PH, rainfall, and soil moisture. In the example of a reef environment, you may choose the location based on depth, temperature, salinity, current, and clarity of the water. You will then choose species from different categories (e.g., coral, aquatic animals, algae), who present different characteristics (e.g., required environment, place in the food chain, how many calories they need to survive, how many they provide when consumed). At the end of the day, all of your species may not survive – and that’s fine – but you need to come as close as possible to the optimal solution.

This is a particularly complex exercise. Browsing through the list of producers and animals available, you’ll realize that the interdependencies get complicated quite quickly. Start by simplifying the data set. You may have 8 – 10 variables but not all of them are important. Perhaps all animals can live within any of the wind speeds present in the mountain range, so this variable can be ignored. Some requirements are correlated with others, so they can be ignored as well. Then try to identify broad relationships between the environmental variables (e.g. depth, temperature) and the species. It is more important to find a rough, high-level solution than to optimize one specific relationship.

Ultimately you’ll need to design a food pyramid. In such a pyramid, there exists:

1. Apex predators (top of the food pyramid – e.g. black bear)
2. Intermediate predators (preys on first order consumers – e.g. owl)
3. First-order consumers (convert plants into meat – e.g. pond turtle)
4. Primary producers (convert soil and sunlight into nutrients – e.g. plants)

So how do we determine which units to pick and their location? In order to solve this problem, we recommend that you break down the exercise into smaller steps.

1. Assess available units: Spend a few minutes browsing through the list of producers and animals to determine the groupings and key characteristics. For example, you might find that certain apex predators need a lot of calories or certain producers provide significantly more calories than other producers.
2. Select unit category: You will need to determine which group of producers and animals you will select. You can start with the Apex predators (those who are not eaten by any other species) and work your way down the food chain. You can start with the primary producers (who do not eat any other species) and work your way up. Or you can select the least needy apex predators, then match them with the suitable primary producers, and finally fill in the intermediate predators and first-order consumers.
3. Measure and verify unit survivability: Once you have your full selection of species, you’ll be able to weed out the less productive species (high calories needed and low calories provided). Take time to map the food chain, establish the processing order (highest calories provided eats first), and verify that each unit is able to survive (not eaten to extinction). If you did a good job, you should be able to get all 8 species to survive.
4. Pick the physical position on the map: We have determined the parameters at step 2, now the final task is to pick the actual location on the screen.

Congratulations, you’ve finished the first scenario!

The Organism Protection Scenario

This is a tower defense minigame. In this scenario, you’re asked to protect a native plant against invader species in a series of progressively more complex maps. The invader species plot paths towards your native plant and you’re tasked with using a range of tools (e.g. terrain barriers, introduction of predators) to deplete the invasive species and disrupt its route to your native plant. Your goal is to survive a number of “turns” per map (five to start with) – in each turn the invasive species takes a step on its path towards the native plant. With each turn, the invasive species is depleted or has to take a longer path depending on the features you have planted. At the same time, new invader species groups can emerge in other parts of the map, at which point you can choose to edit your current plan to ensure the survival of the native plants. If you are into video games, this is similar to Plants vs Zombies and to Tower Duel.

Candidates have reported that this exercise is less rushed than ecosystem creation, so you may want to take your time.

While this game is very easy to learn, keep in mind a few points:

1. Learning the intangibles: You start with incomplete information, from a linear perspective (e.g. when will new predators emerge) and an absolute perspective (e.g. how does your terrain interact with a predator)
   
   It is important to learn the interaction rules with the world and its elements. You might need to experiment and make some mistakes along the way, but if you record your findings you will be able to learn and adapt.
   
   For example, a predator which depletes the invading species by 10 units per turn can be placed on bare land or on a forest (terrain barrier). The forest will slow down the invading species by 1 turn, allowing the predator to deplete for 2 turns (instead of 1). The hidden rule, however, is that you have to place the forest before the predator, and the reverse is not possible.

2. Recognizing that long-term strategy is important: Ask yourself what the long-term strategy is and how will you achieve it? (e.g. should you place preemptive blocks for new invaders that might emerge? should you block off terrain access on the right side of the plant? Perhaps placing predators all around the plant is better?)

   It is quite easy to survive 15 rounds, but how many MORE rounds can we survive, and that will actually change how to design the placement and choice of elements. A key objective to note is, you’ll want to plan to survive for as many turns as possible – while only being able to influence 15 turns. Good candidates will last 20 – 30 turns.

Another two scenarios have been reported in the past, but have not been spotted recently:

• Disaster management – In this scenario, you’re first asked to identify the disaster that is happening (e.g., tsunami, volcanic eruption) based on a set of symptoms (e.g., temperature, atmospheric pressure, rain). You’re then asked to move animals to the location where they are most likely to survive, based on the characteristics of the species and of the locations (e.g., forest density, predators, rain)
• Disease management – In this scenario, you’re asked to identify which disease is affecting an animal population based on a set of symptoms. You’re then asked to recommend the best course of treatment based on the disease, the population (e.g., species affected, stage of the disease) and the characteristics of the treatment options. Your goal is to optimize the rate of survival. This is similar to video game Plague Inc.

Other Tips to Ace the Test

In addition to the scenario-specific tips to we shared above, there are a few best practices to bear in mind:

• Do not try to replicate the solution of other test takers: the digital assessment generates a new, unique scenario for every test taker. Your solution will not be the same as other candidates’!

• Before the test, familiarize yourself with the five cognitive abilities being assessed; understanding them will help you demonstrate them.

• Test your ideas and record the outcomes.

• Take good notes of your findings and observations, in a structured way (e.g., important characteristics, relationships between factors, cause & effects).

• Be comfortable making imperfect decisions based on partial information. There is no one answer to the problems. Your goal is to come as close a possible to the optimal solution.

• Manage your time carefully: do not spend too long on the tutorial, be quick at understanding the situation and identifying the problem you aim to solve, so you can leave time for analysis and decision making. Keep an eye on the progress bar!

A Good Score Alone Won’t Get You an Interview

Given that Digital Assessment only tests specific cognitive abilities, they don’t rely solely on its results to make screening decisions. Instead, they use the test to eliminate low scorers who otherwise have a strong academic background or to identify top scorers who otherwise do not have a strong academic background.

As a result, the best way to improve your chance of being invited to interview is to put together a great resume. You can get our best advice by signing up for our Free Resume Course.

Once you’ve passed the application screening stage, don’t let all your hard work go to waste. Make sure to prepare for the case and a fit interview. Subscribe to our Consulting Interview Prep Toolkit get access to all the online courses, sample interviews, case material, and practice tools you’ll need to ace your McKinsey interview.