Back to Careers Blog

McKinsey Q&A: Gerold

Our ability to collect and process data has grown exponentially over the past few years, and McKinsey Analytics has established itself as one of our most exciting and dynamic new ventures.

Our ability to collect and process data has grown exponentially over the past few years, and McKinsey Analytics has established itself as one of our most exciting and dynamic new ventures. Already touching more than 10% of our work, the group is accelerating our drive to the forefront of data-driven performance.


We recently acquired London-based QuantumBlack, which brings with it a powerful new machine-learning platform called Nerve, and Brussels-based Risk Dynamics, a team specialized in risk model validation. While such tools and methodologies have the potential to generate unprecedented impact for our clients, they’re ineffective without the people who make the complex inferences and decisions around the data they produce. To learn more about who these colleagues are, we sat down with data scientist Gerold, a senior expert in Healthcare Systems and Services based in Düsseldorf.

Interviewer: You’re part of McKinsey Analytics. What does that mean you do on a day-to-day basis?

Gerold: Simply put, I look at huge health care data sets – such as patient information, insurance funds, and treatment costs – to create insights for our clients in Germany and around the world. The most engaging aspect is modeling. I dig through huge amounts of data (a terabyte or more) to find fascinating patterns and trends. Sometimes what I learn provides insights that lead to solutions for problems that weren’t yet on my clients’ radar. For instance, I created a risk-performance model based on patients’ medical histories to help manage healthcare costs against reimbursement rates for state-run insurance funds. I originally built it to show the clients what they could do with the data; it was so successful, it became a standard tool for many insurance companies in Germany and has been used for projects in Sweden, Japan and Saudi Arabia.

Interviewer: Analytics is a fast-growing area – inside McKinsey and in the world in general. How are you innovative at the firm?

Gerold: Innovating in analytics can be tricky. Generating unprecedented insights often requires processing unprecedented amounts of data. You can’t do that kind of work on just any machine. I’m most innovative around finding the resources (e.g., computing power, software tools) we need to do the non-standard things our teams and clients require. It’s challenging and very rewarding to solve those problems – to do things others thought impossible.

At the same time, some of the most rewarding work I do is not necessarily cutting-edge. Showing a client who’s never really used data how powerful a tool can be always makes me feel energized and purposeful.

Interviewer: What are you involved in professionally outside of the office?

Gerold: Perhaps the most fun I’ve had as a data scientist has been in competitions. Most recently, I competed in the Heritage Health Prize, which is an international statistics competition among about 1,400 teams who are all trying to solve the same complex question over a six month period. Of all the teams, ours came in #23. On another occasion, I entered a competition to accurately predict the revenue of fast food restaurants in Turkey. I used the same model I developed for health care, and placed 3rd out of 2,250 participants.

Interviewer: What advice would you share with others considering advanced analytics?

Gerold: What I’ve learned over my more than 15 years in analytics is that success isn’t only about having the most powerful model; it’s also about knowing how to get the right information into that model. In most cases, the richness of data is more important than the complexity of the model.

When interpreting results, you have to be very careful to differentiate between correlation and causality. For example, say you want to look at health care costs by physician. You may see that a certain physician charges more per patient than another. It could be that the expensive physician is providing unnecessary treatment OR it could be that the expensive physician is actually seeing sicker patients who require more advanced procedures. You have to think about what the data means to reach the correct conclusions. That’s the role we play for our clients and it takes partnership. I bring expertise in analytics (and experience in applying analytics in the healthcare space), and I work very closely with my clients and McKinsey colleagues who have a much greater understanding of the industry. Together, we make sure the answers we find make sense.

Interviewer: What keeps you busy outside of work?

Gerold: I’m an avid hiker. This photo is of me on a 900km trek along the Camino de Santiago in Spain. It took 35 days to complete and the views were astounding.

Never miss another post

Receive new stories once a week directly in your inbox