Mining Your Own Business Podcast

Episode 11 - Building Connected Data Teams at General Mills with Zach Wasielewski

Building Connected Data Teams at General Mills

In this episode, we are joined by Zach Wasielewski, Senior Data Scientist at General Mills.

Evan and Zach discuss Zach’s role in strategic revenue management and how data science teams operate cohesively in a large multinational company like General Mills.

They discuss change management strategies and how General Mills has grown as an established company.

Zach also shares about his background in financial services and discusses his transition to consumer product goods.

We recommend enjoying this episode over a bowl of Cinnamon Toast Crunch. Tune in for a entertaining and informative conversation!

In this episode you will learn:

  • How various business units across the company strategize together through knowledge-sharing
  • How an established company like General Mills has incorporated data science strategies and grown over time
  • What strategic revenue management looks like and what challenges they work to solve
  • How data teams interact with the supply chain and future directions for data science projects in supply chain shortages

Learn more about why we created the Mining Your Own Business podcast.

Zach Wasielewski | Guest

Zach is a Senior Data Scientist on the Strategic Revenue Management team at General Mills working with brands like Cheerios and Cinnamon Toast Crunch. His background is in math and economics and he has worked in the financial world prior to GMI.

He’s been at General Mills for one year where he’s focused on sales forecasting and planning promotions.

If you’re interested in joining the team at General Mills, click here. To learn about Zach’s athletic accomplishments referenced in this episode, check out this article.

Follow Zach on LinkedIn

Evan Wimpey | Host

Evan Wimpey is the Director of Analytics Strategy at Elder Research where he works with organizations to transform deficient data into tangible business value that advances their mission.

He is uniquely suited for this challenge by pairing his professional experience in management and economics at high-functioning organizations like the Marine Corps and Goldman Sachs with his technical prowess in data science. His analytics skillset was strengthened while earning his MS in Analytics from the Institute for Advanced Analytics at NC State University.

Evan almost always has a smile on his face, which is at it’s widest when he is helping organizations use data in innovative ways to solve complex problems. He is also, in a strictly technical sense, a “professional” comedian.

Follow Evan on LinkedIn


Key Moments From This Episode

01:51 Zach’s background and how he got started at General Mills
03:16 What is General Mills?
04:00 Zach’s role at General Mills
06:49 How do data scientists interact with people in the supply chain?
08:55 Data flows at General Mills
10:41 How are data teams organized at General Mills?
15:48 How change is implemented at General Mills
18:36 From finance to consumer goods – how did Zach get up to speed?
23:04 Join the General Mills team
24:34 What data science project would Zach like to tackle in the future?

Show Transcript

Evan: Hello and welcome to the Mining Your Own Business podcast. I’m your host, Evan Wimpy, and today I’m super excited to introduce Zach Wasielewski. Zach is a senior data scientist at General Mills, his background’s in math and economics, and he’s got a Master’s in analytics from the Institute for Advanced Analytics. Prior to coming to General Mills, he’s worked in the financial world. He’s been a data scientist at Fifth Third Bank, and he’s been an analyst with Bloomberg. And I will say in a previous episode, we had Gus Kaeding, who was a professional skier. Zach’s also an accomplished athlete. I know this from his intramural days at the Institute for Advanced Analytics, where we were teammates. But I know he is got some pretty cool athletic accomplishments. So Zach, welcome to the show. Thanks for coming.


Zach: Hey, thanks for having me. I don’t know if accomplished athlete is the appropriate word. Glad to be considered with the professional skiers.


Evan: Zach, it’s almost like a model. I consider myself as like the baseline. So all you gotta do is beat the baseline. And then it’s accomplished.


Zach: Perfect. I’m over-indexed.


Evan: That’s great. Zach, I gave a brief introduction. Can you give an introduction here to yourself and sort of how you got into data science?


Zach: Absolutely. Yeah. So my math and econ background, I’ve always been a big fan of the numbers and fell in love with sort of the prediction, and actually customer support was how I got into it at Bloomberg. And really drove me to want to get back and trying to do some data science to try to predict, “Can we do things better? Can we be doing customer service better at Bloomberg?” And went back, got my Masters, made a lot of friends – and probably why I’m here with you today, Evan. But yeah, so fell in love with it. Went back into finance and tried to solve all the same problems that I was hoping to get to solve and have now migrated over to General Mills.


So been at General Mills for about a year in a space that I absolutely love. I mean, who doesn’t love food? Every day I get to work with data on pizza rolls and cereal and looking at everything else that it’s just like, this is more fun than I would’ve ever expected to join General Mills.


Evan: Yeah, very cool. The product in General Mills seems more interesting than the financial service world. It’s more tangible too, which is the great part. Very tangible.

I think maybe it’s helpful to give, I think most folks are familiar with a lot of General Mills. Brands. Maybe you could just sort of paint us a picture of what General Mills is.


Zach: Absolutely. I mean, General Mills, we have a lot of brands. We cover a lot of different areas. Cheerios, obviously, you know, the cereals, Cheerio, Cinnamon Toast Crunch. We’ve got all the monster cereals. I have a whole display of them sitting behind me right now. But we even have like, I mean, Pillsbury, all of your refrigerated baked goods. Pizza rolls I mentioned earlier. We have Blue Buffalo for dog food. We really run the gambit of a lot of the food that you and your pets would love to eat, which I think is what makes us pretty awesome.


Evan: Very cool. You’re a senior data scientist there at General Mills, and senior data scientists can do a lot of different things. Can you talk about where you sit in at General Mills?


Zach: Yeah. So I sit in the strategic revenue management space. A lot of that goes into like planning, promotions and things. So what I think a lot of people don’t know about consumer product goods, like General Mills – our function is not necessarily to sell to a consumer. We obviously – we put our products in a store. We essentially work. Our customer is a store. It would be your Target, your Walmart’s. So we have to try to make these deals to figure out how much does that store need, what type of promotions would be successful in this store? And we plan all of those on our end.


That’s where the data science comes in of trying to predict what sort of thing is going to be successful. How much do we think sales are going to go up if we did a two-for-five sale versus a two-for-six? So there, there’s a lot of different levers that we have at our ability in the SRM space and the data science is just trying to analyze and figure out the effects of those levers.


Evan: Okay. Very cool. Strategic Revenue Management. I know there’s a lot of other functions at – Is it okay to say GM? When I say GM, I think of other GM.


Zach: We go with GMI. Yeah GM, I think they own the GM probably more than we do, but, GMI works.


Evan: General Mills GMI. Yeah. Okay. I’m gonna try to say G Mills. The mills. At the mills. I would assume you have data science counterparts that live in other spaces. Of course, in operations, you make food, you have a supply chain that moves things around. Do you overlap with them? Do they sit in a different space?


Zach: Yeah. Our data science sort of, we spread very wide throughout all of the different areas. We’ve got people who are in marketing, obviously. I think there’s a huge marketing campaign within data science everywhere nowadays. And we have people who are in deliveries. I handle mostly the volume and sales levers. We’ve got people who are trying to analyze our own internal data, trying to do things like supply chain you name it. We probably have a data scientist or some sort of machine learning engineer sitting in that space.


Evan: All right. Very cool. I think you’re focused on sales to the retailers, to your customers. You mentioned supply chain. I wonder how much interaction you would have with those folks. And I’m just painting thinking of you’ve been there for about a year. Supply chain has had some bumps along the way. And so it’s probably impossible to just take for granted that you’ve got all of the supply that you need to sell to a retailer.


Zach: Yeah. We definitely have to work, especially the data scientists as well as all the people in business. There’s so much knowledge of what one person might be working on. Is something that is still applicable. We are one giant cohesive organization. Even though we split into different sections in business strategy meetings. So people who are working on the supply chain trying to figure out are there outages? Is that, that could be why we’re getting low volume, that could be issues with deliveries. What I learn in my day-to-day job is still something that is important for them and what they learn is equally as important for us. I think supply chain definitely falls into that, like the people in supply chain are definitely a hot meeting commodity at this point. Trying to get some of the wealth of knowledge that they’ve learned.


Evan: Very much so. I don’t mean to paint the organizational math here, but do you, I mean, do you sit in meetings with them? Do you knowledge share? Do you help them with their actual analytics and vice versa?


Zach: Not, not necessarily as much of like helping with the analytics. I think we all have our own trajectories of what we have going on. But definitely a lot of meetings, all of our data science org, we meet fairly regularly just to – here’s the ongoings that are happening in, in my team, my department – just to cross share that information. And a lot of times, like in the case of supply chain, if they’re making tables that are trying to account for supply chain outages, things like that, trying to make sure that we are getting that information and we can, if we need it or we want to utilize whatever they produce it, it flows downstream to us as.


Evan: Okay. That’s, I actually wanna stop on that, on the actual, the data flows. Yeah. A lot of times that’s sort of the ideal. We would love it if the operations and the strategic management, sales, marketing, like, sure, we can all access the same thing. Are there challenges there or does it seemingly work? Like you just made, it, made it sound like it works pretty well.


Zach: Yeah. It obviously has its challenges in, in any big company, there’s going to be, there needs to be so much information shared with everybody. And I think that’s kind of where my opinion of having more information is always better. It’s so hard to know and to get that information of what information is even out there. So there could always be more meetings, there could always be more information sharing, but if there’s too many meetings and too much information sharing, then you get bogged down. It depends. It’s sort of a flip of a coin of if it’s beneficial or if it’s not something that you want more of.


Evan: Sure. Certainly. Yeah. But I think, when the tech and the tools are in place that it makes it easy to share, at least you can try to find that balance instead of “we’re lost”, instead of the strategic management team is buying the same data that the supply chain teams is buying over here.


Zach: Absolutely. No, we’ve got a pretty good process of making sure that we’re cross-sharing all of core data and tables. Everything that we use sort of sits in a centralized location so everyone has access to it. It’s just a matter of knowing what those things are.


Evan: Sure. Perfect. Good. So, I wanna talk about you’ve sort of, I’m thinking about a couple of different ways you can split the way you analyze. You’ve got brands, a lot of food, but, but pretty distinct. My dog would love to eat Cinnamon Toast Crunch, but he doesn’t get to, I get to. You’ve also got a bunch of retailers that are very very, very distinct. So I’m curious, how much of what you do, is it common to everything or how much is focused? You’ve got a target team or just a Cinnamon toast SCR team. Yes. Which sounds exciting.


Zach: Yeah. Honestly, I would love if we decided them by brand and by Cinnamon Toast Crunch. I’d be the first one to apply for that team. But no, we, we definitely, because we spread, again, General Mills is a company that we’re in hundred different countries and we have all sorts of different products like. There’s so many different teams. It could be split by country. I personally just work with the North American retail. And I don’t even work for all of the markets in there. I don’t have to analyze certain markets. So there are ways that we have to dig down a little bit deeper into saying we can’t look at everything on a very macro scale because not everything’s comparable. You know that, I mean, Walmart isn’t going to be the same as Wegmans or a Target, there’s going to be a lot of differences in how they run promotions and how they do things. So we do have to do a lot of divvying up.

I sort of have a good catch-all of, I do handle a lot of the different data analytics for a lot of markets in North America. There’s a few exceptions of more niche and more specific data planning, but they’re all built on that business model of the market. All of the markets that I deal with handle very similar business structures.


Evan: Okay. All right. And then you sort of mapped that same challenge of, you’ve got them for folks doing different things, you wanna be able to share best practices or any heads-up that you. From other markets, From other regions.


Zach: Yeah. Yeah. It’s, it’s really cool having everybody sort of working in tandem on similar problems, similar types of analyses and doing it for different businesses. There are different insights that they get and then we can test, and it really does help almost parallel process all of what we do at General Mills and all of our data sites to say, I’m gonna work on this. You work on this. If you find success, then I’ll try it. If not, I’ll probably scrap it myself and move in a different direction. The data science group, especially in the SRM space is all very interconnected. Maybe not as much so with the people, like I said, of supply chain or, or other areas. But we are all very much one team, even if we work in different things, different products, different stores.


Evan: Wow. That’s great. Yeah, I love the analogy. It’s, it really is like parallel processing. Yep. Ah, that’s great. So on the strategic management team, I’m curious, who’s your end customer? If you build a cool model says, Hey, we need to do we need to sell Cinnamon Crunch. It just, of course, it’s my only go-to example that I could think you gave us a bunch of brands, but that’s the one that’s in my head because I wish I had some.

You know, you want to, you want to discount this or allocate some trade to that. Like who, who are you talking to? Who are you conveying this to?


Zach: Our data science and users are essentially the trade planners, the people who are working with a store. I shop at Harris Teeter currently where I live. If we’re trying to make a deal with Kroger Harris Teeter and say, Hey, we wanna put Sierra on sale, we wanna put Sierra on sale this. The price. We’re trying to negotiate, well how many products, like how many pallets of cinnamon toast Crunch do we need to send you? And how much of a deal is it going to be if we’re gonna put it on sale a little bit cheaper this time than we have historically, how much more volume do we think we’re going to sell? So a lot of what I do is predict volume of what a end consumer people at the store are going to buy. And use that information as trying to figure out if the trade planner is shifting a couple of levers or trying to put things on display. They’re trying to put an add out in Harris Teeter’s advertise weekly advertisement. How much more is that going to get us in sales? And is that worth whatever price we need to pay in order to run those ads? Put it on display. One thing that I’ve learned at General Mills is store real estate is a very hot commodity. Everyone’s fighting over those end-aisle displays. Everyone walks past ’em, you know, the whole thing of the sugary cereals, probably at the bottom where the kids can see it, the healthier stuff might be more at the top. There’s a lot of real estate value and stores are no different.


Evan: All right. Super interesting. So, the trade planners are getting this information. General Mills is an old company. Sounds kind of offensive. It would be offensive, I guess, about a person with a company. It’s like a badge of honor. But they’ve, they’ve been working with retailers, they’ve been doing this planning for long before data and analytics were so prevalent. So I’m curious what their appetite is to consume the analytics that you guys have. A lot of what we talk about on the show with previous guests is sort of the challenge in change management. How do you get people to implement some change based on your insights?


Zach: Yeah, I think it’s always going to be a little bit different. I’m glad that when I jumped into General Mills about a year ago, we’re already in full swing of getting what we built sort of sits in this trade planner tool. So they can use this tool to say, Hey, if I make this adjustment, how is that change that feeds through the model? And that predicts what we think the change in volume is going to be. And that change is then helpful for them so they don’t have to rerun all their calculations, they don’t have to do a lot of work. So while even if like bare minimum, we just match the same level of accuracy of predictions as what they. We’re saving them a lot of time. So there’s a lot of like checks and balances to convince people to be like, Hey, hopefully we’re better in what you’re working on. It depends maybe person by person. But hopefully we’re also saving you a lot of time. And it seems like that methodology has worked really well for us to get people on board have been saying like, I want to use. There’s obviously cases of, of one-off, where people are like, Hey, what’s going on here? We’re trying to make sure that there’s a lot of systems in place to check what they would be looking at. We’re essentially trying to replicate what they’re doing with the model. So we’re learning as we go and, and building on top of what we’ve got in place to make sure that anything that they could be using and testing and looking at historically are what we’re also incorporating into our models.


Evan: All right. Yeah. Certainly gotta build the trust there. Yeah. I love the mention on time-saving. You got a background in economics, we both do this. You find some incentive there that can make your job easier. Even if we’re coming up, we’re not challenging your expertise, we’re just saving you a lot of time. Seems like a great way to get some buy-in.


Zach: Yeah, it’s definitely helped that what we built as a tool is essentially the same thing as what they’re already doing in place. Saving them a lot of time that they could be – they request for a prediction, made some changes to their trade plan, they automatically get that back for them so they can start doing other things. They can probably handle a lot more. They can sort of juggle whatever they need to juggle at the same time, whether it’s meetings and be like, Hey, on the fly in a meeting, I can get a new prediction, new volume output.


Evan: Yeah, that’s great. Yeah. Let folks focus on the higher-value activities. Perfect. You’re not the only data scientist there. You come with your backgrounds, math, finance, you worked in banking. Is there, is there much of a learning curve trying to get up to speed in the consumer goods in the food space?


Zach: Absolutely. And I think, I argue that’s probably the case in most companies. I had the same getting into finance. I’m glad I made the transition, but like you said, General Mills has been around a lot of years. I don’t have offhand how old General Mills actually is, but there’s so many people who have a wealth of business knowledge, have a wealth of data science experience and knowledge in this space in particular that it’s a lot to learn. It’s a lot to adjust to. It’s drinking from the fire hose at least for a couple of months. Until you can start, get your footing on all of the acronyms and things that we use. There’s a whole sheet that we have internally of like, here’s the acronyms in CPG and in General Mills and consumer products goods. Yeah. Exactly. Again, the acronyms, we’re all about ’em, so there’s always going to be a lot to learn. But there’s so many people who have been at the company for so long that it’s not like a traditional “I’ve bounced around a couple of tech jobs and been there for a couple of years”, and that in tech, that’s seemingly normal. In consumer product goods, that is something where like, you usually hit General Mills and you stay. There’s a lot of people that I talk to on a daily basis that are “I’ve been with the company for 20 plus years, 30 plus years, 40 plus years.” I mean, me being fairly fresh out of college, it’s a huge opportunity for me to learn and grow in what I’m doing to match what they know and their expertise. It saves a lot of time that I don’t need to be bouncing around from people to people until I find somebody who’s happened to work to be able to answer my question. There’s a lot of people who have that knowledge and have that expertise to help us grow much faster in what we’re building and make sure that it makes business sense rather than data science sense.


Evan: Perfect. Yeah, that sounds great. You talked about the trade planners sort of consuming your analytics, but you’re consuming their subject matter expertise, right? Their context, their understanding, so yeah, certainly if at the very least you’re trying to replicate the work they already do, trying to get a good understanding and having that wealth of experiences hope.


Zach: Right, can’t replicate if we don’t know what it is.


Evan: Exactly. Exactly. So I’m curious if that’s in the data science world at General Mills, are there folks that work in data and analytics that have a long history at General Mills? Maybe even before there was a data science team or there was a big data science footprint?


Zach: Absolutely. There’s a lot of people who, I mean started data science at General Mills and were called different things before it was called data science. We’re obviously matching the times now and trying to migrate more into what the market standard is for data science and machine learning engineering. So there are a lot of changes that we’ve made, but a lot of people who were here and saw a couple of data science hierarchy shifts of what the titles meant. And so yeah, there’s a lot of people, they’re usually a hot commodity as well with just the supply chain people. They know a lot. They know everything. The business and they’ve had the 10-plus years working with all of the people in business who have been here for 20-plus years. So they were able to absorb at least some of that knowledge and they’re now a pretty high contributor in the data science space.


Evan: Awesome, awesome. You’ve been there a year now, you’re working your way up towards expert though.


Zach: I know, I’m trying, Yeah. There’s a lot of people who, I mean, we’ve got the whole gambit of data science, people with PhDs, heavy statistics and visualizations, people who are in more the machine learning, engineering, the deployment, docker. So we have a lot of specialization within our team. A lot of flexibility to sort of do whatever you think you are best at doing and, and sort of balance with. My teammate and I have a pretty good back and forth of like, Hey, this is definitely in your cup of tea. This will help if we work together, either I could learn it, absorb some of that information, or if we just need to like divide and conquer. We do a lot that way as well.


Evan: Yeah. That’s great. And to be able to have the big enough footprint where you can divide and conquer and learn and share, it’s helpful. It sounds like a good team to be a part of. I don’t wanna put you on the spot, Zach, but is, is your team growing now? Are you looking for folks?


Zach: Yeah. Like I said, we stretch into a bunch of different areas in general Mills. I think we’re always looking for data scientists. I don’t know if we’re ever looking for them in, in the SRM space. Things are changing constantly. But we do have positions open. I think we can probably link the job postings or something. But yeah, if, if you’re ever interested in working in consumer product goods, trying to help sell more Cinnamon Toast Crunch, I’m sure we’ve got jobs open, whether it’s machine learning or the data science space, like I’m in.


Evan: Yeah, certainly. It all sounds interesting, Zach. We’ll definitely post some links here. If you’re interested in buying more Cinnamon Toast Crunch, then yeah you can reach out to me. I’ll help you with that.


Zach: Highly recommend.


Evan: Zach, I have one question for you, one final question here. I know you’ve been there a year. There’s a lot – you’re sitting around, a lot of other experience, but you bring in some fresh ideas. I’m gonna give you the magic button here, and we’ll say everybody is aligned to whatever your ideas are, whatever your focus point and whatever your goal is. Where do you want to put your analytic resources, what kind of problem do you want to do? You want to focus on, given that end users say, Hey, that’s what we’re gonna do here. Where do you point the analytics?


Zach: I think what. What I would suggest, what I’m interested in is, we talked about the supply chain side, trying to identify some of those outages, trying to identify if we’re predicting volume of Cinnamon Toast Crunch sales, but for whatever reason there was a supply chain shortage, a lot of stores didn’t have it. Can we identify that? Can we say, let’s not let this input affect our model impute. So I think there’s a lot of things that we need to move in the direction of. We are already moving in the direction of to try to make sure that we’re identifying data continuity issues. So that’s probably the thing that I’m at least most interested in personally. So, hopefully something that I can get more invested in the future.


Evan: Awesome. And it sounds like a relevant challenge, certainly at General Mills, and certainly in today’s world.


Zach: For sure.


Evan: Zach, it’s been super enlightening. Thanks so much for, for coming on the show. Folks, if you’re interested, you can follow Zach here. Keep in touch with all of these brands. See what openings are on the team. If you enjoyed today’s content, please make sure to like and subscribe and catch the next episode of Mining Your Own Business. Thanks, Zach.


Zach: Thank you.