At this month’s AI Frontiers conference in Silicon Valley, our recruiting table was sandwiched between Facebook and Google. The artificial-intelligence (AI) aficionados who stopped for a chat were surprised to learn that McKinsey Analytics comprises more than 800 data scientists, data engineers, and analytics practitioners. When it comes to applying AI technologies such as deep learning to real-world problems, we’re at the forefront.
McKinsey Analytics has a hand in about 10 percent of our client projects globally—a percentage that is rising steadily. In addition to growing organically, we’ve also made a small number of analytics-focused acquisitions, such as London-based QuantumBlack and Toronto-based PriceMetrix. As a McKinsey client, this means you’re increasingly likely to meet consultants steeped in the art of random-forest modeling and of recursive neural networks—people such as Hugh Skottowe, a particle physicist who contributed to the discovery of the Higgs boson; Martha Imprialou, a computer scientist who joined QuantumBlack following postdoctoral research in genetics; or Luke Gerdes, formerly a professor at the United States Military Academy, who has an interest in using analytics to help counter terrorist networks and insurgents.
Why the upsurge of interest in data analytics? Put simply, the availability of massive computing power combined with equally massive data sets (everything from digital photos to factory-performance data) now makes it possible to apply machine-learning techniques to a wide range of problems in business and government. Hugh works in our Ingenuity group, a McKinsey Solution, building analytics capabilities within insurance companies and using machine learning to predict underwriting risks. Martha’s work at QuantumBlack spans everything from analyzing the effectiveness of pharmaceutical sales to optimizing car-manufacturing processes. Luke works mainly with public-sector clients, for example using social-media data to understand sentiment.
In many of our projects, client data is supplemented with data sourced from elsewhere to generate insights, says Hugh. The rub is that it can take weeks of painstaking work before the data is in a fit state for the analysis to begin. “There’s often a long up-front process of data collection, data cleaning, and data linking before the team can start to generate insights and impact for the client,” says Bill Wiseman, a senior partner based in Taipei. “Analytics is an incredibly exciting, powerful approach to solving problems, but it does require a degree of patience.”
This is true not only for client executives hungry for answers but also for McKinsey colleagues steeped in our tradition of hypothesis-driven problem solving, in which strong hypotheses are put on the table early in a project before being tested and refined. “It was hard for me when I first started [at McKinsey] three years ago, since most people didn’t fully understand my role,” recalls Tiffany Kwok, an analytics expert based in our Tokyo office. “But once I got involved in my first project and was able to create a model that ended up generating significant revenue for a client, people began to pay attention.”
It’s against this background that McKinsey Analytics recruiters and practitioners set out their stall at last week’s AI Frontiers conference in California. Back in London at QuantumBlack, Martha sees the results of the recruiting outreach: “Every week, we get a handful of new hires. It just keeps growing.”
Learn more about career opportunities with McKinsey Analytics.