.png)
Reaching patients with rare diseases takes more than broad reach; it takes precision, empathy, and smart data. In this episode, Kate Gattuso Duffy, Global Lead of Media Measurement, Optimization, and Web Analytics at Pfizer, joins healthcare marketing leader John Mangano to discuss how data-driven marketing can make a life-changing impact.
Kate shares how her team uses audience insights and analytics to connect with hard-to-reach populations, reduce waste, and design campaigns that truly serve patients. She and John dive into what it really means to “meet people where they are,” navigate bias in data, and balance rigor with curiosity in every campaign.
It’s a conversation about using data with purpose and how precision makes marketing more meaningful for patients.
Also available on Apple Podcasts, Spotify, Amazon Music, and wherever you listen to your favorite shows.
Transcript:
John Mangano (JM): Welcome to Deep, the healthcare marketing podcast. I'm John Mangano. Deep is the show where we explore how marketing improves patient health, with the people actually doing it. We've all been there — getting to know a new friend, and they ask, "So what do you do?" When I say I'm a pharma marketer, there's always some kind of reaction. Usually I hear about some obscure condition they learned about through an ad, followed by how the ad was "wasted" on them, since they don't actually have the condition or any of the symptoms. But when you think about it, those ads aren't wasted at all — every ad someone sees is educating them about a condition they might otherwise never have heard of. That's especially true for rare conditions. By definition, these diseases only affect a small percentage of the population, but when undiagnosed, they can have a profound impact on the people living with them. Advertising is often an important, sometimes the first, way people learn about them — and in some cases, it's the ad itself that prompts someone to finally talk to a doctor.
Today we're talking about marketing to patients with rare conditions. I'm joined by my longtime colleague and friend, Kate Gattuso Duffy. Kate is the Global Lead of Media Measurement, Optimization, and Web Analytics at Pfizer, and previously worked at Publicis Health Media, where we first got to know each other. Welcome, Kate.
Kate Gattuso Duffy (KGD): Hey, John, thanks for having me.
JM: Have you had that experience yourself?
KGD: I've had people say to me, "All I see are pharma ads on TV — should I even be paying attention to them? Why are they targeted at me?" So, similar experience.
JM: Tell me about your role as Global Lead of Media Measurement, Optimization, and Web Analytics.
KGD: Of course. I joined Pfizer a little over a year ago, and it was a brand-new group — I was the first person in this department, under our media team within the central marketing organization. It's interesting starting a new capability like this inside an organization that's been around for 175-plus years. We've really been focused on advancing and modernizing our marketing across media, creative, and web analytics — understanding how to bring the best tools and processes together to optimize our media and effectiveness, and make sure we're serving patients and providers as effectively as possible with our marketing investment.
JM: Media has changed so much over the last five or ten years — has that made it easier to find sufferers of rare conditions, or does it compound the difficulty?
KGD: Honestly, it's a bit of both. As we evolve, we get more technology, and more places we need to address within an increasingly fragmented ecosystem. There's a real challenge in making sure we move with speed and agility as tech evolves across different silos, while keeping everything connected enough to make agile decisions. On the positive side, though, there's been enormous evolution in how we approach audience targeting with data. Where we used to anchor almost entirely on audience quality as our core optimization lever, we can now do so much more upfront work with audience data — thinking both clinically and through a lifestyle overlay — to be far more advanced in our personalization.
JM: That's really the key difference. Unless you're using advanced digital marketing techniques, even if you're a marketer working in traditional TV or a more conventional approach, I don't think people fully appreciate that, on one hand, marketing is fundamentally a numbers game — a given number of impressions yields a percentage of people who respond the way you want, become aware, and some smaller portion of those take the action you're hoping for, regardless of the industry.
But on the flip side, reaching an individual with a rare condition, or any extremely small, specific group, becomes genuinely hard. You end up spending a lot of money to reach a very scarce profile of person, or in our case, patient. And the harder it is to find them, the less the economics naturally support spending to reach them — which might not matter much if you're selling a luxury watch, but matters enormously if you're offering a solution that could relieve someone's chronic pain, or extend their life. How do you make that work?
KGD: That's a good question, because when you think about reducing waste in the process, it really comes down to using the right data and understanding the patient journey within the clinical data set — understanding, say, the kind of patient claims data we might see three to nine months before someone is actually diagnosed with a given rare condition, when we already know diagnosis itself is a real challenge.
Going through that end-to-end process, the ultimate goal is reaching that specific patient. We know an overwhelming share of the broader population doesn't have the condition, or isn't aware that certain symptoms might even point to it. There's also a whole subset of patients in underserved communities who may not have access to the information we're trying to get to them.
So it really comes down to understanding real advances in patient data, alongside consumer and lifestyle data we can layer in to meet people where they actually are, as people. If we can identify some of those genuinely intelligent signals, we reduce waste, which in turn helps us sustain that investment, and our effectiveness, year over year, while serving a patient population that genuinely needs the support.
JM: What's the biggest difference between marketing for a rare condition versus something unfortunately common, like high cholesterol or diabetes?
KGD: The volume of data is dramatically different between those patient populations, and so is the sample we get to work with. Precision and predictiveness matter enormously more with a rare disease. With a broad-reaching condition like diabetes or a heart condition, there are so many patients, and so many caregivers, that you're much more likely to be effective — whether you're measuring audience quality, or getting a patient back to their cardiologist and adherent to treatment. That's a much easier journey to support across the full ecosystem than with a rare disease, where a patient might need multiple tests, a complicated payer conversation, and so on. So there's a real balancing act in optimization — it's not just better targeting, it's better messaging, better education, better sequencing and orchestration. Without that full mix working together, the rare disease space can be genuinely challenging.
JM: It's a bit of a cruel irony — people talk more openly about high-incidence conditions like high cholesterol, so your odds of being aware of it are much higher, simply because you probably know someone affected. But with a rare condition, it's much harder, precisely because people aren't talking about it — they may never have been personally exposed to it. And that's exactly where the biggest conversation often needs to happen.
KGD: And that's where this work becomes genuinely life-changing for people — it really changes what's possible in terms of care, once you understand it through the data. When people ask why I got into pharma marketing, or what excites me about it, it's this: people tend to assume healthcare marketing has a limited ability to be precise or strategic, but there are actually so many different signals you can use, in a fully privacy-safe way, that ultimately improve the experience for everyone involved.
JM: So true. What do marketers need to do differently when marketing for rare conditions specifically?
KGD: Good question. I'd say: don't try to be everywhere. That's a big trap marketers fall into — wanting to be on TV, wanting to be across every single channel. But in rare disease marketing, you simply can't do that; you need to be precise and strategic with your dollars. Making sure that upfront work is genuinely intentional and purposeful, grounded in data, both about the patient population and who they are as people, is tremendously valuable.
JM: We're both analytics people —
KGD: How many times have I said "data" already?
JM: We're constantly focused on campaign analytics, which is inherently data-driven — and we know all data has bias, that's not inherently a bad thing, it's just something you need to account for in your analysis, or it can get you into trouble. When data is as scarce as it is with rare conditions, what's unique about the bias challenges there? What would you tell your younger self, something that would've helped you find success earlier?
KGD: There's a quote, and I won't remember who to credit, that goes something like: "Data can tell you anything you want it to, so it might as well agree with me." That's a real trap. When you're younger in this field, you tend to think, "Someone's asked me a business question, I need to prove their assumption is right." That's the real difference — learning that it's okay to test and learn, okay for a business question not to land exactly the way stakeholders expected, as long as you're genuinely learning from it. Whether that's "fail fast" or managing performance purely for the learnings, it ultimately helps you improve how you reach that rare disease patient, or provider, with the right content.
JM: I've always found that when data tells me exactly what I expected or wanted to hear, I actually trust it less than I otherwise would — simply because it feels too obvious, and we tend to see what we want to see. There's almost always something two layers deeper in the data that should adjust your takeaway.
KGD: Exactly. Any time I see two numbers on a screen that look suspiciously similar, I immediately think, "Go back and double-check that" — statistically, that's rarely a real coincidence.
JM: People assume what analytics folks do is pure science. Honestly, a lot of it is intuition, and even a bit of superstition — I'll look at something and think, "I can't explain exactly why, but numbers just don't usually behave this way," and there's usually something real behind that instinct. When you're using data to quantify success, what do you do differently specifically for rare conditions?
KGD: There's a lot of added complexity in the data sets available for rare conditions. You need enough sample size to have real statistical confidence in the decisions you're making. Best practice for a common condition might be an incremental new-patient-start metric, comparing a test group against a control, where you feel very confident you've accounted for other variables. With rare conditions, you often have to move to the next-best available metric — you might not have enough sample to look at your true ideal measure, so instead you look for a signal that tells you your media is incrementally driving more patients to the right doctor, or reaching not just the right patient population broadly, but specifically newly diagnosed patients. That's a genuinely important success factor and optimization lever in its own right.
JM: One of the challenges we deal with, especially in rare conditions, is underrepresented populations in the data — for a lot of legitimate reasons. Sometimes it's simply that a population doesn't seek treatment in a way that gets captured in the data. Other times, I've seen doctors code conditions slightly differently, specifically to ensure insurance coverage for that condition. How do you identify that, and what's worked to still extract meaningful insight despite data coverage gaps or anomalies like that?
KGD: It really comes down to how you design the data model for a specific patient population. Large data samples will naturally carry bias when you're trying to optimize toward an underserved or more niche community. So it's about pairing the right data with the right optimization tool for that specific group. You can't optimize a multicultural campaign against general-population data — you need to evaluate it against the right multicultural data set specifically, looking at the right demographic breakdowns, so that broader population trends don't inadvertently override the optimizations that specific patient population actually needs, in order to get the precision that group deserves.
JM: Definitely — in a lot of cases, we don't even realize there's inherent bias built in. In our world, we're a DSP, so we're serving ads to populations that are technically reachable — and not every population is equally reachable, or has a durable ID that lets us reach them. We don't target people based on their health status directly, but we do use durable IDs to identify certain profiles that help us reach the right individual. How have you had to adjust measurement to make sure those harder-to-reach individuals are still counted, rather than effectively excluded from a campaign? Because if we're not reaching certain people, simply because they render differently on an ad platform, or in the measurement system tracking success, we're not just missing out on a business opportunity — we're failing to deliver better health outcomes to an entire slice of the population.
KGD: That's an important thing to think through, being genuinely intentional about how we reach people. You mentioned the evolution of identity solutions and identity graphs, and that's really helped us make progress here. I remember, years ago, working with a publisher focused specifically on multicultural audiences, and we were genuinely under-educated at the time on how to market to, and optimize for, that population using data. We'd go into partner conversations and say, "Your metrics aren't hitting the mark," and they actually helped educate us on the different data sets we needed to think about, and how to be intentional with the KPIs we were setting. Instead of measuring against a broad diagnosed population, we needed to measure against a diagnosed population within the specific factors relevant to a community that tends to be underrepresented in standard data sets.
JM: That's a great point. Not long ago, "multicultural marketing," in health or otherwise, often just meant taking an existing ad, translating it into another language, without even touching the creative, and expecting the same results. There's so much more that creative brings to reaching a different subsegment, and language itself carries so much nuance in how something is actually communicated. And in our business specifically, that extends to legal requirements too — how ISI presents in English versus Spanish, Cantonese, or Mandarin can differ significantly, and can land very differently with the people receiving it.
KGD: Absolutely. It relates to how you work across different cultures globally, too — in my role, I work across both our international and US teams, and there are real differences in how we work with colleagues in New York versus Japan, versus the UK, versus other markets. It's about incorporating those real-world realities about people into your marketing, rather than defaulting to a single, overly specific approach.
JM: Could you give us a couple of examples?
KGD: Sure. You gave a great example earlier — it's not just translating the text of an ISI onto a piece of creative, it's also the imagery itself, the qualitative or quantitative research you need to do on which visuals, creative content, tone, and messaging will actually resonate with a specific audience, and doing the real groundwork to get qualitative feedback directly from those groups.
JM: Great example. When it comes to cross-functional teams, specifically around rare conditions, how do you manage that? You might have a team focused on a specific brand treating a specific rare condition, but you're also relying on legal teams treating all brands roughly the same, plus creative and other functions who may not be fully attuned to that nuance. How do you make sure that differentiation gets carried through every step of the marketing process?
KGD: I'd say bringing data in at the very first step is genuinely important. Everyone can have opinions and hypotheses, but if you can say, "This approach likely won't be as effective, based on X, Y, Z data point we've learned across these five other campaigns over the years, and here's an approach that should grow effectiveness, or perception, or intent," whatever the specific KPI is for that campaign — having those concrete examples up front helps you actually drive influence, rather than launching something into market and learning the hard way, which we've absolutely had to do in the past.
JM: What's the one thing you feel like the industry is still missing when it comes to targeting rare conditions and smaller patient segments?
KGD: I'd say being more genuinely aligned around testing and learning, rather than being overly focused on one prescriptive method. With an audience this niche, this under-discovered, and in need of this much support and education, it's important to try and learn from different approaches, so we can build genuinely better best practices going forward.
JM: That makes sense — new ideas matter. Repeating the same approach over and over might be easier, but it can end up losing the message entirely. With something as unique as a rare condition, there are so many different angles that could genuinely resonate with a patient, especially since we've all been exposed to so many pharma messages already.
Last question — you moved from one of the largest pharma agencies to one of the largest, and oldest, pharma manufacturers. What's the biggest mindset shift going from the agency side to the manufacturer side?
KGD: It's interesting, because my agency experience has actually served me really well here at Pfizer, especially in understanding such a broad cross-section of the industry, and how different teams operate and get things across the finish line. That's helped a lot.
On the flip side, in my role at Pfizer, I get to see all the different touchpoints and how they need to work together — that's been the most interesting part for me. It's not just media, not just creative, not just our websites — it's really this whole end-to-end ecosystem, spanning business analytics, content, legal, and all the complexities involved, that make each person a true expert in their specific product. In my role as a centralized marketer, I get to pull together best practices and set standards, without necessarily over-standardizing in a way that erases the specific, niche needs of how we serve a particular rare disease community.
JM: Well, thank you so much, Kate — this has been wonderful. I hope to see you at some point, driving my old cars along the Jersey Shore. I might spot you off in the distance on your boat.
KGD: That sounds great — give me a little wave.
JM: Alright, thank you so much.
KGD: Thank you.
Links
Relevant Blog Posts: Rethinking Rare Disease Marketing: 5 Questions With Jen Loga
