McKinsey & Company regularly holds the Innovation Olympics, in which our own teams bring their best business ideas to an internal competition. In the most recent cycle, more than 2,000 McKinsey colleagues competed, with the goal of creating a bottom-up innovation mechanism that surfaces great ideas and puts them to work for our clients. In this episode of The Committed Innovator, McKinsey innovation leader Erik Roth speaks with some current winners about the process. Three members of the winning team—Iris Roelens, John Law-Lund, and Suryansh Soni—join us to describe how and why they decided to use machine learning to draw on McKinsey knowledge and create digital agents that could serve as virtual consultants. This is an edited transcript of their conversation. You can follow the series on your preferred podcast platform.
Erik Roth: Today we’re highlighting one of the winning teams from McKinsey’s innovation competition. Every year McKinsey holds a call for the best business ideas, whether they come from working with clients or external stakeholders, or are just something that a team sitting in a team room has devised when they’re working together. These ideas get put through multiple rounds of evaluation, testing, and pitching, and culminate in a grand finale where one team’s idea is selected as the future potential business idea for the firm. Today we’re talking with members of the winning team: Iris Roelens and John Law-Lund, who are product managers at QuantumBlack Labs; and Suryansh Soni, who is a McKinsey senior data scientist.
Let’s start by having you tell us a little bit about this winning idea. John, do you want to go first?
John Law-Lund: Thanks for having us, Erik. We’re bringing McKinsey intelligence and connecting that to agents that can bring it to our clients. The idea is built on the vast store of intelligence we’ve developed over time, which all our colleagues use in our client work and research.
Erik Roth: How did you come up with the idea for it?
John Law-Lund: We started with the machine learning models that QuantumBlack, the AI hub within McKinsey, is most known for. First, we made those models accessible to other systems. Then, as things change and advance within the world around us, and agentic solutions have become such a huge part of our work, we wanted to look at how we could bring those models to agents.
Erik Roth: How did this team come together, given that you’re from all over the world?
Iris Roelens: That’s interesting, because initially we were actually two teams, both already working on solutions in the agentic space. When we both got selected for the finals, the organizing committee decided to put us together to challenge us, by seeing how much more innovation we could bring if we worked together.
Erik Roth: What was that like, putting two teams that hadn’t worked together before on the same team competing for the finals?
Iris Roelens: That was very exciting but also challenging. We had to rethink how we would bring together both our innovations to create a one-plus-one-equals-three scenario. It was really fun, because we got to dive deeper into the agent-first innovation and what each team had built. It was also super-fun to work with Suryansh as a data scientist on a day-to-day basis, because he is a domain expert who works with different profiles than what our team in QuantumBlack works with.
Erik Roth: Suryansh, what was it like for you to join John’s and Iris’s team?
Suryansh Soni: It was a wonderful experience. I really enjoyed going from competing with each other to working together to achieve our mission.
Erik Roth: Tell us a little bit about what is this agent does. What does it look like when you bring it to a client?
John Law-Lund: Think of this agent first as a collection of multiple agents that are super-powerful, doing a task that would require a lot of McKinsey consultants to do in a given period of time. That’s the efficiency gain and the impact. For example, if you’re a pharma company discovering new drugs, and you’re working on clinical trials, maybe you want to automate the data analysis for clinical trials at scale. The agent for that has a clinical trial explorer that is like a massive agentic automation and multiagent workflow that can do the automation you need.
That’s just one example. When our two teams began working together, we realized how we can meld essentially all the components from their side with our side to make a much more concrete and comprehensive solution, which will impact our clients at scale.
Erik Roth: How do you know if the agents are going to be as good as humans?
Suryansh Soni: We’re not saying agents will be as good. Rather, we’re looking to offer a tool that will help empower the people who are in the field, and enable them to achieve outcomes in a shorter and much more efficient way, which will generate a huge impact.
Erik Roth: John, you’re part of QuantumBlack, and you’ve been in the machine learning and AI space for a long time, like the rest of your colleagues. How are you seeing this shift to agents? What does it do for how we serve clients?
John Law-Lund: It’s a super-fast-moving world that we’re working in. All of our work is focused on agents at the moment, as it is for many of our colleagues. A lot of our clients are just now trying to figure out how agents work, or begin to understand how agents can work for them and be applied in useful ways. That’s where we come in. We’re spending a lot of our time working on how we can turn the demos and proofs of concept we’ve seen out there into impactful solutions we can bring to our organizations.
Erik Roth: What was the process of actually bringing the various pieces together into one solution on behalf of the firm and pitching it?
Iris Roelens: Both solutions already had a very distinctive positioning before we combined our teams. The solution we built on the QuantumBlack side has a bit more of a horizontal positioning, where we built technology that will enable many teams across practices at McKinsey, whether that be life sciences, or marketing and sales, or operations. Whereas on Suryansh’s side, they had already built applications that were very targeted toward life sciences organizations. Putting both together enabled us to create a full package.
It’s a super-fast-moving world that we’re working in. All of our work is focused on agents at the moment, as it is for many of our colleagues.
Erik Roth: How did you find time to work on these ideas in addition to your regular work? How did you carve out sufficient space to come together as a team and develop something so outstanding?
Suryansh Soni: We brought in the expertise we needed to execute, essentially. We had a full solution built out from scratch before we joined the two teams together, so we had to decide how we could combine it with the power of the other team to make it an even more comprehensive solution at scale that can deliver the results and impact we wanted to see.
Iris Roelens: It was indeed an effort on top of our daily jobs and responsibilities, and that sometimes felt like a bit much. But since it was so related to our product already, and what we were working on, it was actually a useful addition. It helped us generally in our product road map and strategy as well.
John Law-Lund: We were looking for opportunities to radiate our idea among our colleagues and find other opportunities to work with clients, so we felt this could be a good opportunity to get more exposure, which it turned out to be.
Erik Roth: So this isn’t just a competition for ideas and concepts, but also a competition for real businesses that can scale, and have evidence that they’re already working with clients. What was the most exciting aspect of becoming part of this team?
Iris Roelens: What I’m really proud of is the diversity of our team. We have different roles—engineers, knowledge experts, and integrative consultants from across geographies. It’s also pretty balanced across genders. It’s great to be able to bring the full diversity of the firm to this competition.
Erik Roth: Do you think it was a better team because of the level of diversity on it?
Iris Roelens: I definitely think so.
John Law-Lund: Definitely. I think the number of people we had involved, the collaboration, and all the different ideas and perspectives you get from that level of diversity, is what brought us through the rounds and got us through to the final.
Suryansh Soni: The vibes were very positive from everybody on the team. There wasn’t any time where we didn’t have confidence that we could win.
Erik Roth: Your original pitches went through at least three rounds of early evaluation with McKinsey partners and senior partners from different regions and functional or industry expertise, before the field of competitors was narrowed down to the top five teams, who go on to present to firm leadership. How was that process?
Iris Roelens: That was very interesting, because it was the first time the full team met in real life. We had some virtual dry runs before then, but it wasn’t until we were in New York for the event that we met each other in person. That was a very fun experience.
John Law-Lund: It was daunting but also exciting. We were basically doing a kind of Shark Tank or Dragons’ Den pitch to this group of very senior folks from the firm.
Erik Roth: How important do you think it is for an organization to have some kind of mechanism like this, which identifies the best ideas and connects them to funding and other resources needed to scale?
Suryansh Soni: I think this sort of opportunity is quite interesting. It enables a team to think outside the box, and execute the idea as well. It’s empowering to have your ideas recognized, and it reinforces more people thinking outside the box.
Iris Roelens: Our team was a great case of showing that this innovation competition really succeeds in bringing a bottom-up innovation process. Our team was not that big, yet we still made it to the finals and actually won. It shows that any team with the needed capabilities can actually bring up their ideas and get the biggest sponsorship in the firm.
John Law-Lund: I think it’s hugely important as well, because a lot of the great ideas you’re going to have in an organization are going to be from those people who are working very closely with clients on a day-to-day basis. This mechanism for surfacing those ideas and then celebrating them with the funding to turn them into something real is incredible.
Erik Roth: What was the most surprising thing about the competition as you went through it?
Iris Roelens: The level of sponsorship and visibility. The jury members were the top of the firm, and they all took time out of their calendar to be there, guide us, and pay attention to our innovations.
John Law-Lund: The quality of ideas from all the teams impressed me. We saw lots of the teams as we went through the rounds, especially in the finals. All of them were just incredible ideas with hugely talented teams working on them.
Erik Roth: What’s next for all of you now that you’ve won the competition? You have some funding and support, and I know your practices are really excited about the potential for changing the way the firm serves its clients using the technology platforms you’ve started.
Suryansh Soni: Maybe we can think about a new idea that takes our current expertise that we have built over the past six months on agents, LLM, and the combination of two, to make an even more innovative solution, or a unique capability by combining our forces, that can generate an impact.
Iris Roelens: And also, of course, scale our innovation and use cases across many clients and practices. I think seeing it actually used in organizations is really an important accomplishment, and a strong driver for us in our innovation. We’re also looking more into the concept of skills developed by Anthropic, exploring to see how it could provide value for our clients as well.
Erik Roth: It’s a very interesting approach to building capabilities and virtualizing knowledge and workflows. During the competition, did the judges challenge you on something that really helped push your thinking and make your idea better?
Iris Roelens: The judges challenged us to make our pitch really practical and tangible, and not stay too theoretical, but to make it real, bring in client examples, and translate what our innovation will mean for our clients.
Suryansh Soni: One question that came up was essentially, “How are we making sure the output quality from these agents is concrete?” That helped us frame the answer even better in response to that type of question.
Erik Roth: When did you think you might win, or think you might have some momentum that was carrying you to the next round of the competition?
John Law-Lund: I don’t think we ever felt that, even as we were walking up to collect our medal. We were still kind of in shock.
Erik Roth: What advice would you give others for next year?
Iris Roelens: For our colleagues, I would say definitely participate in it. And do take the time to think through the idea and the innovation, because it will help you refine the work you’re already doing. For example, during the competition, we pivoted our value proposition and core messages a few times, because it was over a duration of eight months or so, and the agentic world moves so fast that we had to adapt our pitch. I would say that it’s good to stay flexible and keep questioning your value proposition and pitch as the innovation evolves.
John Law-Lund: There was a push to think about the pitch and distill your idea down. Even though technically your solution may be great, you have to be able to articulate it well to your audience quickly. I think we had three minutes in some of the early rounds to pitch our solution. To be able to articulate it quickly and effectively, and in a compelling way that would get you through to the next rounds, that was a great skill push for us.
Suryansh Soni: I think this will enable people at the firm. If everybody has the spirit of being entrepreneurial in some fashion, it generates impact. When we think outside the box, and become an outlier altogether by thinking of new ideas, that’s where our true spirit comes out. And that’s what defines McKinsey, in my opinion.
Erik Roth: What advice do you have for other organizations that might be looking to hold similar types of competitions?
John Law-Lund: I think the level of support and commitment from senior leadership is critical to the program’s success. It can easily become an internal competition where the ideas don’t really go anywhere, and support from the top prevents that. These experiences of meeting the managing partner, pitching to senior leadership, and then getting ongoing sponsorship accountability is incredible. I think that’s what makes it such a worthwhile enterprise.
Erik Roth: If you could offer one piece of advice to future teams and/or entrepreneurs and innovators in any large organization, hoping to bring something of value to scale, what would it be?
Iris Roelens: My takeaway, but which probably also applies to others, is to be bolder. This solution we developed was very bold. We went out of our comfort zones and believed in ourselves and our idea, which proved to be right. Be even bolder and go for it.
John Law-Lund: Think more broadly about how to radiate your idea out. What’s the largest opportunity you can find to get the most exposure? Before I entered this competition, my expectations of how to interact with colleagues in the firm and with clients was at a certain level. But this experience has taken it up a notch. Seeing the scale of the firm and the impact it has across multiple organizations opened my eyes to the fact that we can do this differently. For people in other organizations, I think they need to explore the way they can do the same thing—what opportunities and people do they have access to, and how can you get your ideas across to them?
Suryansh Soni: From my perspective, I feel there is a pattern that people tend to follow, and it can sometimes become a deterministic approach and your operating model. But if you think like an outlier, you become a little more stochastic in your thought process, and you innovate. I say be stochastic in a deterministic environment. Think outside the box, regardless of where you are.
Erik Roth: Thank you all for joining us today. I hope you inspire many others, and that you continue to be innovative entrepreneurs throughout the firm and wherever your careers may take you.
Related Articles
