The Data Scientist Who Tells You What the Numbers Really Say

Amber Meakin: Welcome to Vouched, where we're flipping the script on the job search because people are more than a resume. If you're new here, the idea is simple. A personal recommendation should count for more than a resume lost in a stack of a thousand. In a market where AI and applicant tracking systems screen good people out before a human ever sees them, this show puts one real person in front of you, vouched for by someone who's actually worked with them. Here's how the format works so you can listen the way that fits you. The first few minutes are the vouch itself, the referral, the honest case for why this person is worth hiring. If you're a hiring manager or recruiter and that's what you came for, you've got it in the first five to seven minutes. After that, we hear from the guest directly about their expertise and exactly what they're looking for. So stick around if you want the fuller picture. I'm Amber Meakin, this is vouched. Let's get into it. Today's episode is a little different. I've brought back Tony Zubek, who's been on the show, because there was someone he wanted to vouch for, and the minute he told me who, I was in. Her name is Kari Brandt. Kari is a data science leader who never stopped being a builder. She runs teams and sets the strategy, and she still designs and builds the models herself, the segmentation, the forecasting, the causal work that tells you whether a program is actually doing anything or just taking credit for customers you already had. She's done that for some of the country's best-known loyalty programs for a global marketing agency across health insurance, pharma, and automotive, and most recently for a major automaker. She has a PhD. She's worked at the scale of tens of millions of customers, and she's an analyst who tells you what the data actually says, not what you're hoping to hear. Tony was her client, so I'm gonna let him make the case. Tony, Kari, welcome to the show.

Tony Zubek: Hi, thanks for having me back. Appreciate it. Kari, so nice to see you again. When I first met Kari, she was on the vendor side when I worked at one of the major retailers on loyalty. And every time she came in the room, we knew we had the best data available. We knew we had the highest level of insights. Her reputation far exceeded any expectations throughout the organization. And there were even moments when I presented her work and they said, Where did this come from? I said, this came from Kari Brandt. And they said, okay. So they knew that all the information she provided was accurate, concise, and very applicable to the brand. It was even more so, Amber, that when we renegotiated the contract with that vendor, I made sure that her name was in the contract, that she was not allowed to be transferred off of our brand because of her amazing skill set. And the abilities that she portrayed when we were going through and really designing, augmenting, developing, maintaining that fantastic loyalty program that turned out to be an award winner, partly because of Kari Brandt.

Amber Meakin: Awesome. What a wonderful intro and wonderful recommendation. How long did you work with Kari?

Tony Zubek: we were together for about five, six years. I remember, that was in the beginning when our the program was in a state of decline. And we really needed to focus on how we can improve upon it. You know, we turned a multi-quarter decline into a growth pattern where we received 41% incremental increase in signups. And again, the retention also improved and the frequency as well of the customers in the program. And a lot of it was due to the good work that you did from a data side.

Kari Brandt: Thank you, Tony. It was a great account to work on. Yeah, I could vouch for you too. You were a great client.

Tony Zubek: I tried, you know, what am I gonna say?

Amber Meakin: Tony, I'm gonna ask you, I'm gonna ask you one more question If you're a hiring manager listening, what are you getting when you hire Kari?

Tony Zubek: Everything and more. You know, it data is a telling a story. And to be able to interpret that story and to speak to it to any level of an organization, whether it be a specialist in the marketing area or the CEO, she has that skill set. She knows how to craft gigabytes of data, turn it into something meaningful, and to be able to extrapolate conclusions that we can then continue to evolve the programs that we did. So if you're a hiring manager out there and you're really thinking about hiring a true exceptional data scientist, Kari Brandt has a seal of approval not only by myself, but I know of her colleagues and of friends and her other vendors that she's worked with in the past.

Amber Meakin: Awesome. Yes, I I worked with Kari briefly. We didn't we didn't work for for too long on the same account. I think we were kind of crossing paths. But I I do recall several wonderful insights that you brought along and the stories that you were able to tell during our quarterly business reviews that we did share together. Okay, Kari, what are you looking for right now? I know director of data science is the kind of seat you're aiming for, but what type of team and problem is the best fit? Are you open to regulated environments and remote or a particular place?

Kari Brandt: Well, I I want to stay where I'm located. So I'm open to something in my area or remote so that I can stay where I am. I'm not looking to move if possible. But I would like to do something like I've been doing where I'm looking at customer data and doing data science with customer data to help improve the business, whether that's with a loyalty program or marketing. or with the product team to help them design products better. Any of that works. And I've worked in as as you mentioned, in a variety of industries. So industry doesn't really matter to me. I've done retail, I've done automotive, I've done pharma, I've done lots of different industries and I'm pretty good at getting up to speed on the new data each time I start a new vertical. So industry doesn't matter, but I want to continue analyzing customer data, helping the business make better business decisions based on that data, giving them good actionable insights into what their customers are doing. That's the sort of job I want and I'm looking for.

Amber Meakin: Awesome. And I think we may not have said it already, but you're based in the Dallas, Texas area. Is that right?

Kari Brandt: Correct. Yes.

Amber Meakin: You've spent your career on one question. What is a program actually worth? What pulled you toward that instead of reporting the numbers everyone else reports?

Kari Brandt: Well, I'm I like to solve problems and puzzles. I always want to try to figure out what's going on and solve any issue, problem, question that comes up. And reports don't answer that question directly. You really have to dig into the data and play around with it for a while to figure out what's going on. And so that's my passion. I love playing with the data, figuring out what's going on, building models, forecasts, segmentations. That help the client, the customer, the business stakeholder better understand who their customers are, what the problem is, and how to solve it. I like to take an ambiguous question, frame it into a way that it can be answered by data, and then find that answer and tell the story around the answer. That's what I love doing, and that's where I Do best.

Amber Meakin: many companies and especially this the CFOs can potentially see loyalty programs as a cost center, not a revenue center. you've said members spend more than non members being as being the most expensive sentence in loyalty analytics. For folks outside of this field, why is that line so dangerous?

Kari Brandt: That line is dangerous because you're comparing two different groups of people who have different characteristics and different behaviors. You can't just compare them directly. Members are self-selected. They've decided to join the program, therefore, of course they use your product more. They wouldn't have joined the program if they weren't already vested in your product. Whereas non-members were clearly not vested in your product, and that's why they didn't join. so of course the members are spending more, visiting more, doing more. That's just the nature of becoming a member. You're not going to be a member unless that's going to be the case. just saying that doesn't mean it's true. You have to actually control for those differences and do different statistical techniques to balance them so that you can see what is there really an incremental lift. In what the members are doing. Are they really spending more than they would have anyway? Because they're already your best customers. So are they spending more than they would have if they hadn't joined the program? Is the program increasing their value? And you can't measure that directly against a non-member because they're very different.

Amber Meakin: Yeah, this is why it's so critical to have, you know, consumer insights on the team and digging into the data to help understand, what's the difference in the value of a member versus a non member and how do we convert, non members to members, because there is a difference in in the value that they bring to the organization. So you are leading teams, but you also still build the models yourself. A lot of leaders tend to let that go. Why do you hold on to the hands-on work?

Kari Brandt: I feel like I will better understand what my team is doing if I understand the data myself and the problems myself. If I better better understand what's going on, then I can better mentor, advise, and help other people. If you get too far away from the data and the actual modeling, etc., you may not understand the problems they're encountering. So I like to keep my hands in so that I can better understand what's going on and help people, other people. I also enjoy it. Okay. It's also some of the funnest part of the job is doing the actual data work and modeling. So I don't like giving it up. I like to do it myself as well as mentor other people in doing it and helping them learn. from what I've learned over the years. So I think it's important for a leader to keep their hands in the data if at all possible.

Amber Meakin: Yeah, that makes a lot of sense. And that can even translate into almost any sort of leadership role where, understanding what your team is experiencing and challenged with. Let people do their job, of course, but definitely know what what's going on so you can understand the insights and provide recommendations and that mentorship. I agree, that's very important. You've built engagement measurement from scratch, back when there wasn't even a shared definition of what engagement meant. How do you build something when the standard doesn't exist yet?

Kari Brandt: Well, it's the it's complicated and it's not like there wasn't any potential measurement. most people measure recency and frequency of almost any behavior. so those are two KPIs that just about everybody measures and you can and we applied them to engagement too, but they only show you a slice of engagement, you know, one piece of information. There's lots of there's lots more. To engagement than just those two. So we developed other measures that looked at other aspects of engagement, like consistency, not just how frequently, but like of all the days you could use the product, how many did you use it? Okay. And because we were looking at use of a product in that point, engagement with a specific product, mobile app, etc. So how of the days you could use it, how many did you? That's not, you know, frequency is just how many times you used it, but consistency would tell you what percent of the time were you using it. And then we also came up with intensity, which is on the days you use it, how intensely do you use it? Right? You know, are you using it constantly when you use it, or do you just use it one time each day? So we came up with a lot of different measures that when you looked at them, they were not correlated with each other. So they were in fact measuring different aspects of engagement. And it was just a a lot of it's trial and error, trying different things, seeing what they measure, what they tell you, and if that seems to be meaningful or not, and then codifying it and of course working with your business stakeholders. Because if even if you come up with excellent measurements, if you haven't included the Business stakeholders they may not agree and they may want different measures. And so you need to work with your business stakeholders or clients or whatever to come up with measurements that are both statistically valid and business-wise valid that the business will appreciate and use. So it's it's a long process, not super long, but it is a process. Okay, it's not something you just sit down and come up with on your own and then implement. You have to work through all the data issues as well as involving the people who are going to be using those measurements.

Amber Meakin: It reminds me, when the product team would be designing a new feature. And I would have to ask, have we talked to the data science team about this and what fields they may need to then measure those types of things like engagement. I can think of a couple of features that were just so critical with we did a badge's gamification implementation and we had to make sure we were tying things in and planning that out. At least in my experience over several organizations, I almost feel like it's a afterthought. And my goal has always even been to try and make sure we bring people in early and often to have those conversations because it is so critical and the data is gonna tell us is this working or not? And so being able to tell that story and know what the things that need to be measured are, get your consulting expertise on How does the engineer need to do this to get what you need to to do the measurements on the data?

Kari Brandt: Yes, that that is something I experienced too is they didn't always consult us on what we would need to measure and report on the new feature or whatever was going in. I was at my one company long enough that the data people started including me because they realized if they put something in and it didn't work for me, they'd have to redo it. So They started including me at the beginning to make sure what we needed to what we needed on the back end was being captured because that is a critical element that sometimes gets lost in development of a a feature.

Amber Meakin: Yeah. And I think it's almost in a way for the product team to know how their product performed, you know, that's an exciting piece. If it's doing well, it's great to be able to share that in the future to say, hey, this is the results and value that it drove for the customer. it's such a critical piece. Tony brought this up during our episode. and I'm so interested to hear more. And I think the audience is probably gonna be pretty curious about this too. You trained as an anthropologist and served in the Peace Corps. Before you were ever a data scientist, how did you get from there to here? And how does anthropology still show up and how you work?

Kari Brandt: How I got there is a little windy. Because yes, I did. I got my PhD in anthropology from the University of Michigan. I did more quantitative analysis as a in in anthropology. I was studying teeth and bones and reconstructing diet. So it was a more quantitative, and my PhD was a quantitative analysis. so I had the statistics and the computer skills when I decided I didn't like being a professor. Okay. When you're at a university and you have a specialization like that, you're the only one doing what you're doing at that university. Your other colleagues are spread out across the world in other institutions. So it was a little lonely. So I just decided I wanted something that was a little more collaborative and the people around me were doing the same things I was doing so we could work together. So I decided to leave academia and get into the business world. And the first company that hired me hired me to do data science analytics on their customer data. So that's how I got here. Because from then on, I just took jobs in that same vein and moved my career forward analyzing customer data and figuring out insights. I think the anthropology degree helped me learn how to take a problem, any problem, and figure out how to answer it with data. Because sometimes that's not readily obvious. And you need to be able to figure out what is the best way to answer that question with data. It also really helped me to to figure out how to ask questions so that I understand what the real problem is. Sometimes somebody will come to you and say, I I need this, but they don't really. What they need is something else. They just don't know it because they they they need you need to ask probing questions. So I think that's where my anthropology background has really helped me, is in that is in taking problems. Figuring out what the real problem is and then how to answer it with data. so that's where it still fits in.

Amber Meakin: Okay. Tony said you have a bone named after you. Is that is that right? Is that from your time with with anthropology?

Kari Brandt: Yes, it is from a time when I was with anthropology. I went on several fossil huntings two summers I went out to hunt fossils in the Badlands of Wyoming and I found a very the first specimen of a very, very early primate and so it was named after me because I found the first fossil of that primate. So yeah. It is pretty

Amber Meakin: That's pretty cool.

Kari Brandt: cool.

Amber Meakin: All right, Kari, so one final question. I'm gonna ask if there's someone you'd vouch for the way Tony vouched for you, a former colleague you'd rate or anyone that you've, worked on a project with or that's helped support your team in some way.

Kari Brandt: Yes, I would like to vouch for Greg Hughes. He's in infrastructure, and at one place I was where he was too. the data science team had to maintain some of our own infrastructure to a certain extent. we had our own, you know, software and tools that we were using that weren't fully supported. So we had to do as much as we could ourselves, and whenever we ran into a problem. Greg was always there. He was always helpful. He didn't make us feel stupid because we couldn't deal with the problem ourselves. He was always incredibly helpful and kept our infrastructure working regardless. He always was always very helpful, making sure the infrastructure was sound, which is very important from a data science perspective. Because, you know, getting to the data, being able to analyze the data. Without a good infrastructure we can't do it. And Greg was always there, always solved the problem. A very, very helpful.

Amber Meakin: Yes, thank you, Kari I also had the pleasure of working with Greg and agree he's a he's a great colleague and and partner to work with and any company would be lucky to have him. So thank you for vouching for him as well. And I look forward to speaking to him next week for another episode of Vouched. Thank you again, Kari and Tony, for your time. Much appreciated.

Kari Brandt: Thank you for having us, Amber.

Amber Meakin: Absolutely. Before we go, the line I want to stick with you is the one Kari gave us. Members spend more than non-members is the most expensive sentence in loyalty analytics because members were your customers before they ever joined. That's the kind of thinking you're getting when you hire her. If Kari sounds like someone your team needs, her contact information is in the show notes. Reach out. That's the whole reason this show exists. if listening made you think of someone you'd put your own name behind, tell me about them at vouchedpodcast.com. I'm Amber Meakin, this is vouched, and we'll see you on the next episode.