Imagine yourself in a beautiful, serene forest populated by many kinds of wildlife. As you take in the flora and fauna, you learn about an urgent matter demanding your attention: the animals are quickly succumbing to an unknown illness. It’s up to you to figure out what to do—and then act quickly to protect what you can.
In November, 520 recruiting candidates found themselves in just this scenario, on their computer screens. They were at our London testing site immersed in a digital, scenario-based assessment designed to understand and measure how they approach and solve problems—in other words, the type of thinking our own people do every day.
As the needs of McKinsey’s clients evolve, our firm is evolving as well, through both acquisitions and organic growth, and we are broadening the types of talent we recruit, including data scientists, implementation practitioners, IT experts, product and digital designers, and software developers.
McKinsey’s standard recruiting process involves evaluating résumés, administering a multiple-choice test, and then interviewing select candidates. “But the multiple-choice test has limitations,” explains Keith McNulty, McKinsey’s global director of people analytics and measurement. “Recruiting only knows if candidates got the right answer, not how they approached the question. Plus, there’s a large amount of strategy, preparation, and luck involved in multiple-choice tests, and if you use them in the selection process, it reinforces the status quo—at a time when you are looking to widen the scope of candidates you’re hiring.”
Our firm has deep experience using in-person, scenario-based assessments (such as case interviews) in recruiting. Yet we miss more nuanced perspectives on the skills of some candidates and miss out on other candidates completely because we can’t meet everyone in person. McKinsey began investigating this problem in collaboration with a start-up called Imbellus, which has a team of data scientists, engineers, and psychometricians who aspire to replace traditional standardized tests and are committed, above all, to science. We also enlisted the help of UCLA’s National Center for Research on Evaluation, Standards, and Student Testing (CRESST) to help identify and measure the skills we need.
The result of this collaboration is a new scenario-based assessment designed to help our firm attract new and different talent profiles from all parts of the world. It not only evaluates candidates in a lower-stress, more engaging environment than a traditional test does but levels the playing field by minimizing the influence of a candidate’s background.
It also provides a lot more insight into a person’s skills. When people analytics are applied to the test results of hundreds of recruits—a fast-growing data set—it can not only help determine whether a candidate is creative but also provide an understanding of a deeper subset of skills. How does this person absorb information? Is he or she an idea generator? Does he take an unusual approach to situations? Is she methodical in her analysis—or tend to follow intuition? Can he easily manage multiple factors? “When people join our firm, we can be more intelligent about placing them in a position—and on a course—that makes sense,” says Keith. “This approach offers, potentially, hundreds of different readings on a candidate and provides literally an anthology of skills.”
What’s next? The team is expanding the pilot to other geographies to collect additional data points before the assessment becomes a formal part of McKinsey’s recruiting practices. Another goal, in the next few years, is to make the assessment available online so we can understand candidates’ skill profiles without bringing them into the office. “This will especially help recruiting in emerging markets,” says Keith. “As a firm, we’re reinventing ourselves and becoming more diverse to benefit our clients. So if the way we serve clients is constantly evolving, our recruitment processes have to as well.”
Meanwhile, back in the forest, more and more animals are exhibiting symptoms, and time is running out. Our candidate chooses a course of action, and the diagnosis is revealed—and so is a way forward for the candidate at McKinsey.