Product Management looks different at every company, and can even vary across teams at the same company. While the craft of Product Management is held together by a set of core principles, each team has to find their own nuance. It could depend on their industry, the size of their company, and the type of tech they’re working with.
So we decided to sit down with three different organizations, to discover what Product Management means to them.
This time, we’re chatting to Aaron Arnoldsen, Director of Data Science at BCG GAMMA. We talked about the benefits of Product Management training for data scientists and engineers, and the kind of people needed to build a culture of empathy and product thinking.
You might also be interested in: Product Management at Freddie Mac
About Aaron Arnoldsen
Aaron Arnoldsen is a seasoned leader with over a decade of experience working in Product, Software, and Data.
He is currently a Partner and Director of Data Science at BCG GAMMA. In his current role, he is leading AI for large-scale transformations and internal product for Lighthouse by BCG GAMMA. Before his current role, he spent two years as Director of Software Engineering for ML and AI at the Boston Consulting Group. He also previously worked in various roles at Nike, including Product Area Director for Price Optimization, Planning, and Supply Chain Technology. He also spent time working on Senior Revenue Management and Analytics Data at Walt Disney Parks & Resorts.
How did you get started in the tech industry?
I actually started off in the data science space. I started off working in demand forecasting and working in the revenue management analytics department at Disney. I moved over to Product when I moved over to Nike. One of my old colleagues brought me over as a supply chain allocation manager, and one of the first things I noticed is all the challenges that I faced, with not having the right tools, not having the right products built for me.
And it was a real eye opener for me personally, in my own Product journey. How do I make sure that what I build, how I’m thinking about products really are built for the customer or for my clients in a way that is meaningful and that I’m building the ideal solution for them and not just some fancy algo.
Titles are difficult to define across companies and industries. What does ‘Director of Data Science’ mean to you at BCG?
First, it’s exciting that we’re in the middle of this massive evolution of AI and machine learning. Here at BCG, we’re really focusing on how to turn this from a buzzword bingo, to actually scalable deployable capabilities that our clients can actually put in operations and in their personalized marketing campaigns. It’s allowing them to say, how can I think about CO2 admissions differently? How do we solve the tough problems and integrate them into business solutions?
I think one of the keys to that, from a data science perspective, is not so much about what is the perfect algo that we’re trying to build, or what is a perfect solution, but more importantly, are we kick-starting the right business processes? Are we making sure that we’re building the right capabilities that can mature? Are we putting the right team structure around to ensure that what we’re starting can continue to evolve? Because we know over the next couple years, this space is evolving so rapidly that we have to be pretty agile in how we’re thinking about this space today.
You might also be interested in: The Why & How of Enterprise Analytics with Spotify Data Scientist
Would you say that at BCG you have a culture of experimentation when it comes to your approach to Product?
When you start to think about how we integrate solutions that really drive meaningful business impact, there’s not a cookie cutter approach to doing that. Many teams and companies are organized so differently that you have to be able to have that sense of innovation, that test and learn mentality that enables your team to see not only the bigger vision you’re trying to accomplish, but what can we do next to drive ROI right now? What can we learn right now that can drive this product to scale?
So the test and learn mentality has to be at the forefront of how we think about data science and how we think about building products in this space. Because right now I find it very interesting when you start to look across many of the major player companies, one of the biggest questions that CEOs are asking is, ‘yeah, I know AI is great. I know I need it. I don’t know how to get it into my operations. I know I have this data science team, but how do I actually get it into my planner, into my merchants, into finance’s hands so that they can make decisions more quickly with better data. And that is only gonna happen as we test and learn. And we’re maturing that capability in the organization and that happens incrementally. There’s not a silver bullet that’s gonna get you there.
You’re bringing in Product School team training for your data scientists and engineers. What are the main benefits that you’re hoping they’ll get out of Product Management training?
I think that there’s three benefits that I’m hoping that our team takes away from training in product. I’m really not that big on deciding and prescribing PM processes, like ‘you should use Kanban’ and that kind of stuff. Really, you have to look at the problem you’re solving right now. And at the level of maturity your product is. And then you can adjust to the tools, and what I’m excited about, in the way that Product School does training, is instilling a culture of Product Thinking. How do we ensure that our data scientists, our engineers, everyone that’s working on our problems are thinking, ‘okay, where is the customer? What is the mindset of the customer right now?’
So we’re instilling the core principles of a Product mindset, and that’s number one. The second benefit that I’m really excited about with the Product training, is helping teams step away and really get to the core of asking the right questions. I’m guilty of this too as a trained data scientist, but it’s so easy to dive quickly into the procedural side of how to solve a problem that sometimes you don’t take enough time to think about what you are actually solving. And so through training we’re hoping to create a structure for our teams to really understand that part of the puzzle. I think it’s critical.
And then thirdly, I think we are at a time in the AI and machine learning transformation that we’re going to have to focus on scale. And there’s some certain tradeoffs that you have to start thinking about, right? Like from a data science perspective, you might say, ‘what’s the model that gives me the best accuracy?’ Well, maybe as you’re scaling across 250,000 SKUs and 2000 stores, maybe accuracy is a secondary KPI. And the algo that helps you actually have the least amount of run time with accuracy becomes now the critical decision path. And so understanding what the right decision paths are as we move to scale is also gonna be another important and exciting part of Product School training that I’m excited for. For the training to really instill this culture here at BCG.
It’s easy for employers to underestimate how much their teams appreciate the opportunity to learn and receive an education. Are your teams excited about the training?
Well, I’ll say this, we announced it in December and our first training cohort was overbooked within that day. And I’m having people reach out to me almost every day asking to be involved. So I think people are really excited. I think the market right now is really excited and starting to understand what Product can do for them and their careers.
Product and tech people are obsessed with the problem, not the solutions. What are some of the problems you’re the most passionate about solving right now?
I couldn’t agree with you more about being obsessed with the problem. For me, it’s really obsessing over the why statement of what you’re doing. Why is this a problem? What is the impact that my client or my business partners are trying to do differently and what are they actually trying to do?
Early on in my career, I remember simply recreating a process and a data flow for my forecasting group. It wasn’t the most eloquent solution and being a little bit more seasoned now I can tell you, I definitely didn’t make all the best technical choices at the time. But I remember the sense of excitement and gratitude that others on the team had when they saw that I was able to simply solve easy process and data flow problems that impacted their everyday work life. And so I think Product is less about grand visioning and more about exceptional observations and authentic empathy, making sure that we have empathy for who we’re solving this problem for, in a way that we really understand where they’re coming from when we’re creating the solutions.
How do you foster that sense of customer empathy at all levels in your organization?
I think there’s a couple things for me. First, I’ll carve off the leadership portion of that. For me, great innovative leaders, Product Leaders care less about what they can do and more about how they can unblock and empower their teams to be creative. I think the most important skill set that I’ve started to learn in this space, especially when we hit the intersection of Data Science and Product Management, is how to let go more and trust my team to do great work.
While I feel that I have two major roles in this process. Number one, I need to be able to effectively remove roadblocks, whether that’s funding roadblocks, political roadblocks, or whatever challenges that we may have. I need to be able to help remove the roadblocks and enable my team to keep moving. That’s my role.
The second one is that I’ve become less and less afraid of hiring people that are much smarter, and much better than me. I find more and more of my role is less about telling them exactly what to do. Many times we have choices between two to three great solutions. And my role is more about breaking ties about making sure that we’re picking the right solution that fits where the vision needs to go three to four years out. And if I think about what a leader’s role in all of this is, it’s to really unblock the team to be innovative.
Now, answering the second part of that, how are we doing this at BCG? I think the fundamental part that excites me most here at BCG is we care deeply about our clients’ challenges from a day to day operations perspective and how they can be better. Whether it’s, you know, and I say, be better very generally, because like, there’s some that it’s like, Hey, how can I operate my supply chain better?
There’s three areas that we really focus on when we start talking about how we solve a problem here at BCG. One is expertise, right? You can’t get around it. You have to have experts in data science. You have to have experts that know Product. There’s not one skillset, there’s many that we’re looking for. We’re finding people that are passionate about and dedicated to their craft and are willing to become experts.
Secondly, at GAMMA as we have so many different challenges to solve, we have to focus on finding people and that have true empathy, that can work with many clients all over the world, that can really understand what the client’s problems are, and then be able to be in their shoes as we’re solving the problem.
The last thing that we’re focused on is really finding people that have the passion and perseverance to continue on. The whole world right now is talking about AI at scale. Well, there’s the really cool part of AI at scale, but there’s also the mundane parts like ML ops, fraud support, etc. So at BCG we’re thinking about how to find the right people that are willing to tackle the problems, not only when it’s exciting, but also when it’s challenging. And that’s really key. I’ve found the best Product people are the ones that are willing to think through the problem not just when it’s the exciting build, but also when it’s in sustain mode.
How have the events of 2020 and beyond impacted the way you do business at BCG? For example, have your client’s problems changed, or has the way you approach problems changed?
This leads to an area that I’m very, very passionate about. I think we’re at an inflection point right now, where there’s a couple factors that have boiled together at the same time. From a pandemic perspective, it really has altered the way governments, companies, and even individuals use data and make decisions. There’s so many things that were reliant on year over year trends. And in a lot of cases, that’s gone. And now we’re in our fourth wave of the pandemic and each wave is so different. So how do you react in the moment? It’s very different, but at the same time, this has happened right at a moment when data has become cheaper, easier, better to store and we have better computational ways to actually integrate it into our decision making process.
So this massive event has occurred, that forces us out of our comfort zone. But then at the same time our ability to use algorithms has gone through the roof. And so where I’m the most interested and where I’m doing the most amount of reading is in thinking about when we get to the other side of this, the companies, the people, the governments that will be differentiated the most will be those that have taken this opportunity to use this alternative data.
And so we’re really at a cool moment right now, and we’re seeing it with a lot of the questions our clients have today. ‘How do I think about forecasting differently? How do I think about pricing differently?’
But at the core of it, it’s really, ‘we now have more data signals that can influence than ever before and coming in real time, how do we actually make that serviceable and usable in our operations?’ So it’s a really exciting time from my perspective to be in Product, because I don’t think that there’s one product out there, at least in the tech space that isn’t influenced by this movement. And it’s going to be up to us as Product people to really understand what’s happening, and how we can capitalize, how we can create a competitive advantage within our product for our clients or the consumers of our product.
Finally, what advice would you give yourself at the start of your career, if you could go back in time?
I think that there’s three things that I would say, I would say patience, empathy, and comfort with ambiguity.
I think sometimes we are too quick to jump to a solution that is suboptimal because it’s the only solution that we see. Many times great problems are more solved by better service to our clients. And so you need to have true empathy, which I’ve already talked a lot about.
I think the last one’s the hardest. We can try to do all the training, all the hiring, all that type of stuff, but really knowing and embracing ambiguity is what I’ve learned is the most important. Especially when we get into the very beginning of product development, there’s a lot of ambiguity. You’re not gonna know all the answers and it can be overwhelming. But I think if I were to go back ten or fifteen years I’d tell myself, it’s okay to have ambiguity, embrace it, be willing to break that down.
I’d like to end with a quote that I love from the movie, The Martian. It says:
“Everything’s going to go south on you. And you’re going to say, this is it. This is how I end. Now you can either accept that or you can just get to work. You just do the math and solve the problem, and then you onto the next problem and solve that problem and solve the, the next problem after that. And if you solve enough problems, you get to go home.”The Martian, Andy Weir
It’s okay to not know everything. You have to be able to test and learn and be comfortable that you and your team are going to solve the challenges that you face and trust each other.