On the latest episode of the Change Agents podcast, Alter Agents Chief Strategy Officer Devora Rogers is joined by Katie Zmijewski, Research Director at Alter Agents. Katie shares the early-career, multimethod research project that first showed her what deep strategic thinking looks like, how she thinks about finding “flow state” in research work, why her psychology background gives her an edge in translating complex models for clients, and what makes a segmentation actually stick versus becoming another one clients feel burned by.

What is the insight in your career — recent or not — that changed everything?

That’s a meaty question. What really comes to mind, and what I find myself going back to year after year, is a project from about ten years ago, working on a major global CPG brand’s face care line. We were doing an intense, multimodal consumer journey study — a mix of advanced analytic methods, shop-alongs with eye tracking, ethnographies, you name it, it was there.

What really sticks with me is that we were able to condense that immense amount of data into a cohesive story for the brand. That was really one of my first exposures to truly deep strategic thinking. Early in my career, a project was often about answering a single question, or being one piece of a larger puzzle. Getting exposure to that kind of broad strategy work is honestly what propelled me toward Alter Agents — it’s much more strategic, deep thinking work that we do here.

How did you actually bring all of that research together — the different methods, the different data sources — into one cohesive story?

It’s truly something where we could have drowned in the data — there was so much to sift through, align and weave together. What really helped was grounding ourselves in some advanced analytics — not just stated data, but derived metrics that had real substance to them. That became our guiding light through the whole project. It let us align our visualizations with deep storytelling and high-level strategic takeaways. It was really exciting.

If you think about flow state, my own flow state at work is really those moments of swimming in the data while still being able to think creatively. That project was one of the first times I experienced that, early in my career.

It sounds like that project showed you what a flow state at work could feel like. Did that change how you looked for those moments going forward?

Yes — though being such a monster of a project, you don’t get that opportunity every month, or even every year. So it’s definitely been something I’ve kept trying to find. And I don’t think it has to be a huge, monster project — it could be a segmentation. That’s such a foundational piece of research that everyone in an organization has to align around. You have to really swim in the data again, understand the personas, understand who they are and why they matter. I can get to that same place in a segmentation, which I love doing as well.

You’re working on a segmentation right now for a major retailer, and I got to see you in the middle of narrowing down a lot of possible solutions to bring forward. How do you deal with that kind of overwhelm in the middle of a big project?

You sort of just have to get through it. You have to keep up that stamina, and trust that it will become clear. I’ll work through a few more questions, organize things, set aside what’s distracting, and really focus on the core objectives. It really is an exercise in endurance and stamina to get through some of these moments. But then once you’re through it, it’s like, yeah — we did it, we’re here, we’ve hiked the mountain, we’ve made it.

How did you actually get into research?

My background is in psychology — specifically the experimental, biological and neuroscience side of it, rather than the clinical or therapeutic side. I always enjoyed figuring out the mechanics of why our minds work certain ways, what determines human behavior, and figuring out those answers. Nothing has one singular answer, but there are ways to figure out why something works the way it does.

You work across both qual and quant, but your foundation was originally in quant, is that right?

Yes. With experimental psychology, you’re definitely working in SPSS, crunching numbers and getting yourself to a statistically significant answer, and that’s stayed with me through my career. I’m no statistician — I’ll never claim that — but I think having that understanding, without being too deep in the math, is really what helps me bridge the gap between a model or an output and what our client actually needs. I can sit with a model output or a segmentation profile sheet and break it down into the basics: what our objectives are for the work, what we’re seeing, where meaningful differentiation shows up, what’s rising to the top. I don’t get too bogged down in a specific score, and that’s the part I like. I’d thought for a while about going into a more scientific doctorate-type degree, but I would have missed that human connection — that understanding of how we’re actually working in the real world, real-world human behavior.

If we were at a party, I’d corner you and ask: what’s your favorite research methodology?

That’s such a tricky question for a researcher — it’s like making us pick a favorite child. I’ve worked across so many different types of clients and needs, which is honestly one of the more interesting parts of my work. But I think what I love most are the research programs where I’m helping brands build their foundational understanding — segmentations, A&Us, brand positioning, customer journey or customer experience work. That’s when you can see the lightbulb go on for clients — yes, I know how that connects to our objectives, I know how to move this forward, I know now why our sales data look like this. Those big, meaty, foundational programs are what I love. I guess that skews more quantitative than qualitative, but qual is necessary too — it’s what brings life and light and interest to the work.

We hear a lot from clients that they’ve been burned by a past segmentation. Why is segmentation so important, and where does it tend to go wrong?

You touched on this earlier — sorting and sifting through a dozen-plus solutions. All of those are mathematically valid solutions that could be taken to a client with robust mathematical rationale behind them. But you have to layer on what actually matters to the client — what they care about, what their brand stands for, what they can action on, and who they already know their customer to be. All of that has to layer in, or it won’t stick. If it doesn’t feel like, “I know that segment,” or, “ooh, that’s me” — seeing yourself, seeing your neighbors in the solution — that’s the most important part. That’s what makes it stick.

And without a segmentation at all?

You’re condensing so many different types of people into one singular persona, and that’s not going to resonate deeply with anyone.

What are you working on right now? Anything exciting you can share?

My team and I are working on some meaty foundational work for a couple of consumer and retail clients right now. We’re doing a segmentation for a large global retailer, and it’s in that really fun phase where we’re starting to swim and be creative with the data — I’m loving that. We’re also doing some jobs-to-be-done work, which is a great methodology.

Tell us about jobs-to-be-done, for anyone who isn’t familiar with it.

It’s a way of putting a layer between what people are actually doing and why. So, okay, I bought this beverage because I was thirsty — but really, I was thirsty because I wanted to recover after my workout this morning. So I’m purchasing this beverage not just to drink something, but to help my body feel better after an intense workout, let’s say. That’s the job I’m hiring this beverage to do. It gives brands the opportunity to fuel innovation and figure out where they have room to grow, especially for messaging, but also for product development. It’s a really fun one to live in and to see where a client takes it.

Are you approaching that qualitatively or quantitatively?

It’s actually a methodology that I really believe has to have both, because you have to hear consumers talk about it in their own words before you can do any sort of validation or quantification.