No items found.
In this episode, we bring you a powerhouse panel of industry leaders exploring how data and technology are transforming healthcare marketing. From the early days of marrying health data with media insights to today’s AI-fueled acceleration, our speakers dive into how speed, personalization, and smarter outcomes are redefining the playbook for pharma campaigns. You’ll hear how agencies, pharma brands, and analytics experts are approaching the next frontier of real-time optimization in healthcare.
Together, they tackle big questions: How can AI empower planners, buyers, and data scientists without losing the human touch? How do we break down the silos of fragmented dashboards and create unified, actionable insights? And what does it really mean to balance innovation with privacy guardrails, regulatory oversight, and patient-first thinking? The conversation is candid, practical, and packed with sharp perspectives on the opportunities and risks ahead.
Looking toward the future, the panelists share their vision of what’s next—whether it’s the rise of citizen data scientists, new streams of health data from wearables and telehealth, or the evolution of personalized patient journeys. Above all, they agree on one thing: the goal isn’t just efficiency or optimization. It’s using smarter data and technology to deliver better outcomes for patients. Tune in for a lively, thought-provoking discussion that will leave you rethinking how healthcare marketing works today—and where it’s headed tomorrow.
Transcript:
John Mangano (JM): Welcome to Deep, the Health Marketing podcast. I'm John Mangano at Deep Intent's AdLab conference. This year I had the opportunity to host a panel with leaders from across pharma, agencies, and analytics. We explored how health marketers are connecting massive data sets, embracing real-time optimization, and navigating the role of AI in driving better outcomes for patients. It was a wide-ranging discussion with diverse perspectives, from boutique agencies to global pharma, all focused on innovation, health intelligence, and our industry's future. I thought this conversation would be interesting to our listeners, so we're bringing it here to the podcast. Enjoy the panel.
Thank you, everyone. This is a particularly exciting panel for me to be a part of. We have not just some of the most representative people as it pertains to the types of clients we work with — we're also going to be talking about how all the data connects, how we're using it, and how new technology like AI, although I'm trying to avoid the term AI, seems to be what we always talk about, is going to change everything. Right now we're in a renaissance in how we're able to leverage that data to go beyond what we've been doing for the last five years. But everyone on this panel was part of the initial renaissance where we pulled together health data and media data to quantify the exact impact campaigns have had. That was five years ago or longer, and five years ago this group predicted that, built it, and now they're going to tell us how they're using the tech we have today, as well as the leading tech that's evolving for the future.
So it's a really exciting time. On the panel: Bill Veltre, who joins us from Deerfield, but has also worked at a few other boutique or medium-sized brands, as well as at PHM doing analytics like this. Kate Gattuso-Duffy, who is at Pfizer, representing among the largest pharma companies out there, as well as having experience with a lot of different agencies in our agency life. Mike Caruso from SSCG, which really crosses across all of it. And Mike Bregman from Havas, where he manages one of the largest analytics teams in the industry. So we have a really diverse and great view of not just what we're able to do and what the future holds, but also how we got here.
So, super excited — first, things are changing. Our ability to do what we do today compared to five years ago is pretty dramatic, and we talk about the level of insight, but it's that velocity that's already changing, so things get faster and faster, and they'll get quicker as we progress. How have things changed from just five years ago, as it pertains to what we're doing and how we're managing campaigns? Mike Caruso?
Mike Caruso (MC): Yeah, I think just the speed alone has increased rapidly. Obviously with AI-powered tools, AI being the backbone of a lot of ad servers and things of that nature, I think it's increased our speed to performance. And when you're talking about any kind of real-time optimization, anything where you're able to pull different levers, that's a key aspect. Especially — I lead the biddable practice at SSCG, it's really all we care about. And when you think about the industry as a whole, you see more and more publishers with biddable capabilities. Previous publishers that were direct-only, committed buys, are now becoming more biddable, and that's because the technology has improved so much and allowed them to have that new capability overall. I think speed to performance, and the industry's ability to build new capabilities that adapt to that, is very important, and that's what I've seen change over the last five years.
Kate Gattuso-Duffy (KGD): Yeah, and you said it as well — five years ago we were talking to stakeholders, to people, to say, "Hey, how do we shift from vanity metrics like click-through rate and impressions to really optimizing on outcomes," and we had to really make that case, that "what" and that "why." I think the same kind of thing is going to start to happen in the next year or so when we're talking about the term real-time optimization. For healthcare, that wasn't necessarily a term that allowed us to be strategic with the optimization and the data sources we'd be able to optimize in real time. So now it's really that shift of looking forward: how do we make sure we move quickly in healthcare with the right data sets, at the right speed and latency, to improve that personalization.
Bill Veltre (BV): We've worked on a lot of campaigns together, so when I think about speed, I think about those five years ago and the speed to education — my favorite topic — ultimately being able to make sure clients are up to speed on the data sets available to them, so they can mine and find the nuggets that are going to create impact. Being an advocate from the agency side now, no longer on the client side, it's an interesting position to see how marketers have changed with the speed of change in marketing technology. The marketers managing many of the individual brands we're servicing — the ability they have to mine that data, the curiosity that's available today, is so different than five years ago, and I think it's all because of the focus, or the late need, on results.
JM: Mike, what do you see at Havas?
Mike Bregman (MB): Yeah, there's a lot of change within the agency, and I'll park agency change — we'll probably pick up on AI and our tools in a bit — but I think it's interesting to start from the client perspective. Our clients want growth now more than ever. I think they're under a lot of pressure to get things done quickly. So the idea of the brief itself, how you think about a campaign plan, how you receive those funds, what the right allocation tool, process, mathematical equation is — that's being done really rapidly, and I think our clients want to see that done in a smarter way. So it's less about always-on experimentation, it's about reinvention of the plan itself, and some of that has to be new creative elements, new orchestration, new design, new thinking. Where I think our clients are asking us to help them is stress-testing some of their own innovation initiatives — that could be personalization, data orchestration, third-party data partners, or some of their attribution models and MMM models. But I think all of that begins with: what is the purpose of the budget itself? How do you make it that much smarter, more agile —
JM: So it's funny, because we're all trying to avoid the AI term, because it does seem to take over the conversation — and I don't know if you've noticed, AI is of course in the news, but it doesn't matter what we're talking about. It could be physical trainers training athletes, and we can improve their swim stroke. It could be doctors, like radiologists, reading X-rays. When we talk about AI, AI is just the enabler. Rarely do we actually talk about the use itself. So I wanted to direct this to you specifically, Mike Bregman — what are the personas you see within your teams using AI, and how is it different from persona to persona?
MB: Yeah, so there are four primary personas we think about for AI, and there are hundreds of use cases — Google just came out with a study with 350-plus use cases. We're not trying to unpack every use case; we're trying to think about the user. How do we help the user be that much smarter with AI? This idea of art and science is really important — there's really human and machine. The machine itself is that AI element we're trying to build into the process to drive more efficiency. The human has to figure out what problem they're solving. There are four main personas. The first is the planner. Planning is a lot of art, and they have to figure out what's the barrier they're trying to solve, what's the consumer journey, who is the patient, who is the physician audience, et cetera.
There's massive research they have to go through — AI helps them sift through a lot of those ideas and package it in a much more agile way. The second is the insights pro. When you think about insights overall, you're getting a lot of campaign data back — you're looking at prior results, thinking about the right way to create the creative components, the right way to use different channels for different missions. The insights pro also has to sift through data, but they have to use it in a smarter way to compound the activities that are already in market. So it's the agility of getting the insights back. The third is the buyer, and I think DSPs are really critical to getting the buyer to be smarter. The AI needs to be conversational, and you need to be able to prompt specific messages — is this publisher working, how do I think about frequency capping, is there a regionality element? There are millions of questions you can get, and figuring out those protocols, those prompts for how to work with the AI — that's a key playbook we're developing. And then the fourth is my favorite: the citizen data scientist. There are a lot of elements of SQL and Python and R code — we can geek out a little on this panel — and I'm assuming there are folks who don't know the basics of how to write those. With AI, there's actually prompt management, where you can write — the AI can write the code for you. You can put that into different clean rooms and do smarter analytics on the back end. So attribution, the closed loop, the things that used to take three months, where you'd have to wait for some of these script-lift studies — now it can be done much more rapidly.
JM: So the thing I see is that all of that's an enabler, but foundationally, it's about having the data connected underneath. In every one of those examples, it really was about the data itself — what data sets are we combining, or need to combine, to be able to deliver these insights? I think you're spot on — the best of those is now: how do we get them all to a place where we can enable more of it?
MC: I can maybe explain a bit on the Omnicom side of things, how we're thinking about it, and how it connects to the data as well. Those personas are obviously very realistic personas at the agency level, for the different types of people who'd need to access AI tools, data, et cetera. What we have is what we call Omni — that's our operating system and orchestration system, and it sits on a foundational layer of a ton of data from third parties. A client's first-party data can come in as well, as can any proprietary data we may have, and obviously there are compliance guardrails for that foundational layer. But we've also built specific applications and AI agents for every one of those personas. So not only do you have that layer feeding all of them, or access to that data, you now have purpose-built applications or AI agents that speak specifically to a data analyst, or a planner, or a bid campaign manager, where they can leverage that in each specific use case.
The good thing is that because everyone has access, including our clients, we can keep a repository of whatever you build for others to see. Data silos are a big thing in our industry, but organizational silos are a problem too — if someone's doing really good work in one area, and another team or department can't see it, you lose that cross-collaboration. This kind of solves that too — you can use a specific AI agent or application, but someone else can see what you did and build on that for whatever their persona specializes in.
KGD: Being on the client side, you really get a front-row view into the fragmentation of data. Something I maybe didn't even realize a year ago is how many dashboards we have to look at — which is a dashboard, right? Ideally we stop talking about dashboards and start talking about how our data connects, and use AI to drive decisions. But until we get those dashboards to connect, and the underlying data to connect, we're still going to have six or seven data sources that might be telling us different things. So how can we bring that data infrastructure together, while all the different data models have their own structure, their own piece? That's where I really see a ton of opportunity in AI — to help us say, okay, across Adobe Analytics, to IQVIA data, to spend data, to DSP data, everything we have at our fingertips, market research, everything else on the business side — can you write some sort of script to help us normalize all that, bring it together, and really help us do our jobs quicker and faster?
BV: I thought you were going to say the other bad word, "omnichannel," but you didn't — thank you. But I think you're right, there are a lot of dashboards to look at, and it's ultimately left up to the advocates of your business, like an agency, to make sure you understand all the data sources available to you. From a Deerfield approach, or even from an agency approach holistically across the industry, we want to make sure we understand each individual stakeholder's need for data. Because when you're talking about an internal analytics team trying to do a mixed model, their data sets and what they're looking at are ultimately very different from what my media team is optimizing off of. There's connectivity between those two, but we want to make sure we're bridging the gap so the mixed models and the analytics happening at the client match what's happening on the front end — really connecting those two dots.
JM: Mike Bregman, you probably manage the largest organization amongst us. How do you manage that level of accessibility across your teams?
MB: We try to democratize as much of the data as we can. It's hard, because you need really good training, really good systems access. The tools themselves have to be standardized in terms of taxonomy and lexicon, because to go into a massive data set, no matter the health data provider, you have to know what you're looking for. AI can help to some extent, but you have to stress-test what the AI is actually solving for. Prompt management is definitely a lot of training, but even once they go in, you have to make sure the AI is using the data sources to answer the right questions and giving you a viable answer. There's a lot of stuff you get from AI that doesn't quite break the ceiling. What we try to train all our analysts to do is go in with a good hypothesis.
There has to be a good learning agenda. The learning agenda has to be proven in some scientific way, and mapped to effort versus the value it delivers. If there's really high value, very low effort, that's a really good test to run. Then they have to figure out which data partners to use, which could be an RFP process, or internal stakeholder interviews to figure out what's worked and what hasn't for their clients. Once the test is in market, how much time does it need to reach statistical significance? It could be a really expensive test with a really high-value vendor, a million-dollar effort on behalf of a client, in which case you need better org design, better stakeholder interviews, and so on. So we train all our analysts to go in with a curiosity mindset, but to be mathematically driven all the way through the process — and if they're building their own sophisticated model, who's best equipped to help them succeed.
BV: I think that works for large brands, larger agencies, but when you're working in small- to mid-size pharma, there's not a lot of time or money to actually do that test-and-learn. You have to be very agile, and the learning agenda becomes more important, where you're setting up based on what you know, because you may not have the budget to test and learn across a ton of different channels or collect a lot of different data sets. So it's a lot about risk — being able to take the risk needed to drive the business forward without having all the answers or all the data sets to analyze. I think it's definitely a counterbalance.
JM: When you look at the future — short term, the next year, year and a half, and then long term, five years, which is a long time in the world of data connectivity and AI — what does one to two years look like with what we're doing today? And what are the things, in five years, that we don't even have eyes on right now, that will be the future state we're working in?
KGD: Sure. When you're thinking one to two years out, ideally, we spend the next year really being purposeful about how we want to use the technology and what business questions we want to answer with it, and then doing all the back-end assessments to make sure we have the right guardrails, the right trade-offs, and the right processes and decisions in place. So to me, the next year is probably build-and-test, then build-and-scale a little. But when we talk about scale, we have to be careful, because every audience, every patient, every provider is going to have a different patient journey, provider journey, different hurdles. So why we're scaling, and maintaining that audience-first approach in how we're building these systems and technologies and connectivity, matters. Year two, I think we're off to the races — AI becomes a tool in our toolkit, whether it's, "Hey, I have a 75-page analytics deck, can you recap this for me in five minutes and pull out the key insights," or, "Hey, I have anomalous data, I think my spend is wrong, can you figure out where things might pop up?" It becomes that element, or that team, that used to take a week, two weeks, three weeks to do a thing, and helps us do it really, really quickly. Five years out, we're near — I always caveat, cutting the "near" out of "near-real-time optimization" for healthcare — because there's a patient journey you have to take into account, understanding the time from exposure to conversion for your patients, and that change in behavior with your providers. I think we'll increase our propensity for real-time optimization, so we can really have everything working in the background to drive those incremental outcomes.
MC: Yeah, I think what Kate's saying makes a lot of sense. Whenever we're working with different pharma clients, they're obviously investing in AI right now, and we see that across the board. I think some companies are more advanced than others, but everyone's kind of starting, right? So in two years I think a lot of companies are going to have a higher sense of depth in that technology. On the agency side, almost every tool we've used over the past two years already has some level of AI woven into it, and I think that's going to advance from an insights perspective even more. Look at the Cortex announcement today — that was Copilot a year and a half ago for DeepIntent. So I think that level of advancement in the tools we use is going to continue. Five years from now, I think we're going to have a lot more data sources than we do now — not just in marketing, but in health and wellness broadly, whether it's wearables or telehealth or different devices being used that have new health data that might be marketable in the future. So one priority for me and my team is staying on top of what data sets we can leverage — are they compliant, but also new, unique, do they let us qualify our audience in newer ways? I think that's going to continue to grow as the media landscape gets more diverse in the tools people use at a public consumer level.
MB: I think there's going to be more and more orchestration happening in the media landscape. I think right now we've gotten quite good at data orchestration — taking third-party partners, first-party if you can get it in pharma, weaving that together, figuring out the right identity layers, putting that into a clean-room element of some sort, and then targeting. I think we're good at that use case. Where I think we're going to get even better is content orchestration, and a lot of that is personalization-driven, personalization-theme-driven — thinking about the right message to deliver at the right time to that patient, wherever they are in their journey. From there, it's thinking about the media journey itself — they're on their patient journey, but what's the media journey we want them to be on? How do you think about potential switching that has to happen between different medications, different brands? How do you intercept them with new types of information as they're going through the journey? We don't quite have the real-time data points to figure that out yet — the data's still monthly, sometimes weekly, but mostly monthly, and you have to wait for it to refresh with the right cadence to make some of those changes in campaign design. So I think it's going from identity layers and data orchestration, to content, to the journey itself, over the next few years.
BV: Autonomy, right — automated learning models that make decisions based on goals. I think that's where we're heading, in a place where it's not going to be created equal for every single channel, but ultimately in the programmatic landscape, there's going to be a lot more trust put into AI models that generate better impact for us. And I think that's going to come with a human touch and oversight, but that training is going to be critically important for platforms like DeepIntent, as an example.
MC: Just to piggyback on what Mike was saying — I think personalized journeys are going to be more available, easier to hit, but as we see more and more channels become biddable or built into DSPs — look at the growth from a format and channel perspective a DSP can do now versus five years ago — EHR, gaming, point of care, digital out-of-home, these things are getting more advanced in DSPs, and with that comes the back-end data. When you can see those clinical-effectiveness metrics based on exposure for those channels, that helps fill in the journey you need to hit, knowing that journeys aren't linear anymore. I think that data will really help propel us toward getting more personalized.
KGD: I think when we say "personalized," privacy comes along with that. Right now there are new careers, new teams being formed to help govern and oversee AI privacy and application in healthcare, and I think it's going to be really interesting — a whole team of people learning a new technology that's evolving really quickly, while marketers are also trying to use it quickly and start to refine the data it feeds, what it teaches it, and how it learns. Seeing those two come together is going to be interesting.
JM: And privacy is something we manage to probably more than most industries. What is it about privacy that concerns you most — not the legislative side, I don't think any of us want to predict what's going to happen there — but ultimately, as we do what we do, how do we deliver on our privacy promise while still achieving our goals?
KGD: I think guardrails are going to be really important. You touched on five years ago, when we were talking about why we need a shift from clicks and outcomes to script lift — well before that, six to eight years prior, we'd done a ton of work making sure we had the right infrastructure and the right guardrails behind the scenes, the right data flow that protects patient privacy. So making sure that guardrail is first and foremost, and being able to manage that conversation and have those trade-off conversations about how we test, how we do new things, how we grow our data sets — but first testing with data sets we feel confident are privacy-compliant, and then moving into this world of ever-changing building blocks.
MC: And obviously rigorous due diligence in the partners you work with, and internally leveraging AI to monitor what's evolving in the privacy space. To Kate's point about things we know are compliant — what can we leverage, with AI now in the mix? Contextual has always been in play — we know it's compliant — but now there are AI-powered contextual tools and partners you can work with that get much more specific. Another way we're trying to be compliant is not only refining the targeting, but refining the inventory we're trying to stay on top of. If we know a patient audience is really interested in certain things, relevance to that audience is what we should strive for — not only using AI-powered contextual, or a client's first-party data, which is obviously compliant and owned by the client, but also using new technology to reinforce and reinvigorate historical compliant data sets.
MB: Of those 300-plus use cases I mentioned, every one of them has a legal packet behind it, and I've never dealt with so many lawyers — in a good way, mostly because they're trying to stress-test to make sure we don't get ourselves in trouble, that our clients don't get in trouble, that ultimately the partners we work with are all within the legal safety net we want to be in. There are a lot of gray areas right now in how to use AI and legal compliance — everyone's trying to figure it out, and I'm certainly not the expert. We want to make sure we protect the privacy of every patient and physician, that when we target, we're doing it in a way where the best models, algorithms, and AI prompts reach the intended target and eliminate waste in the system, but in a way that's federally approved. We're figuring it out in real time.
JM: And that's the thing — while there certainly are AI privacy experts, if any one of them went on sabbatical, they'd probably come back and not be outdated, because the environment is changing so quickly for all of us using AI too. So the last thing I'll ask all of us: looking at the marketer of the future, the next five years, our skill sets are going to have to dramatically change. What would be the skill set you'd look for, to bring in a team that can be future-proof and grow with the next five years of data centralization, connectivity, and the use of tools like AI and other things we don't know about yet?
KGD: I'd say agility and innovation are our key areas, because we'll need to move quickly. We'll need to be able to say, "Okay, our process to create this report took 95 steps before — now it needs to take three, and we're able to send this out a week, three weeks, four weeks earlier." So being able to move with uncertainty, and have that agile ability, while also keeping an eye toward how this could get better, what I could do differently, what new view I can take with the data, and how I can bring value.
BV: I think it's probably a couple things. One is curiosity — that cannot change. We still need to train folks coming up in the industry to remain curious, ask really good questions, and not just trust the data — ask for more data, try to uncover those individual nuggets that will drive better impact for your clients. In addition, I think it's having a patient-centric mindset. We can't lose sight of who we do this all for — AI is great, tech is nice, data is awesome, but at the end of all of that is a patient, and we have to make sure we don't lose sight of that, and keep talking about and training on it. And the third piece is remaining constant on the fact that we're creating impact — there's a forecast behind every single brand we're responsible for contributing to, so making sure folks understand how that forecast was derived, and how media plays into it, is critically important.
MC: Just to piggyback on Bill's point about the patient — that's an important piece that isn't really going to change, in the sense that it should already exist if you're doing it well. There should be a certain level of passion about what the brand treats and how it helps patients, and that should be what your strategy is based on. As technology, AI, whatever you want to call it, evolves, it gives time back to reinforce that point. If your data is automatically pulling, or you're getting insights quicker, it should give you more time to truly learn what your patients are looking for, what their challenges are, what HCPs' challenges are in prescribing, and really understand the brand. The technology advancements should give you time back to really think about your strategy and how it hits home for the brand.
The second part, which is more new, is that with that technology, we need employees across every department to really know how to dialogue with it. Prompting is obviously a very common thing talked about in the industry, but since most of these technologies are conversational in nature now, having teams trained on how to ask the right questions — now that they've had time to learn more about the brand and tell the system what you're trying to achieve — is going to be a key factor for success.
MB: We're going through a couple of retraining programs for our teams. We're taking all our analysts right now — we have a pretty big group — and putting them through what we're calling the DAT, the Data Automation Team training. A lot of our analysts come in and don't know SQL, which is one of the foundational elements of data mining. They're stuck in Excel in some ways — nothing against Microsoft, I think Excel is a fantastic product, but it's a limiting factor. It's a static table where you can't put in a billion rows, so you need net-new skills. AI can get you some of the way there, because you can ask within a prompt what you'll find if you look for X, Y, and Z parameters, but you still need those skills to become a more advanced user — a citizen data scientist, as I mentioned earlier.
So we're going through this program, and one thing we're trying to do is make re-education exciting for our analysts, because some of them have been on the job for a year, and they're part of this assembly line of building dashboards. Dashboards are great — we have thousands of them we push out every week — but it's also a limiting factor. The dashboard can only get you so far. It's a static wireframe template where you feed in similar types of data, and we found we're actually rebuilding dashboards for our clients almost quarterly — there are new asks, new data sources, Adobe's changing things, what happens if I pull in CJA data — they love their three-letter acronyms at Adobe. So we constantly have to rebuild the dashboard, and the analysts are in this constant process of feeding new information into a static template.
So we're trying to redesign how information is democratized at the agency and for our clients. If you think about social media platforms, they all have a scroll feature — you scroll through Pinterest and get lots of different images relevant to your interests and the contextual areas you're searching for. Snapchat, Meta, they're all in the same game. We're trying to think about how dashboards can feed those same kinds of insights. You have your trading module, and within that platform there has to be a static screen, but what if there's a way to move the information around? We're thinking about mining the data through SQL and other advanced querying, but also making the dashboard itself more of a living object. So it's some of the ways we're thinking about re-envisioning how information is provided for our teams.
JM: So there's so much going on in the future for us, and I know I'm excited. What's the thing that excites you? This is our last question. Bill, we'll start with you.
BV: Platforms like DeepIntent. Was that a good answer? No, I'm kidding — that is the best answer. Partnership. Being able to create solutions and be part of things that are going to drive the outcomes of tomorrow, with platforms like DeepIntent, the product launches coming today — all of those things are what makes me love doing this. Helping more patients through technology, and this platform specifically, I'm very thankful for.
KGD: I think Bill said it earlier — making sure we're focused on improving patients' lives and outcomes in the process. There's nothing better, from an analytics person's perspective, than seeing a report and saying, "Okay, we increased our new patient starts by X amount, and that helped us do X more things with the dollars and grow that incrementality." To me, that's really exciting — that speed to market, that speed to improvement, personalization, and relevancy, so we can ultimately improve those outcomes.
MC: What I'm most excited about is more technologically savvy organizations across the board, both on the client side and the agency side. We just spoke about education, so I might repeat myself a bit, but I view AI, and the latest in it, as an idea accelerator. With training — one thing we've been trying to do is lay out our blueprint for AI, showing different departments how AI is used across all these different things — once that training is fully fleshed out, and you have a large mass of people who can accelerate their ideas for business outcomes, I think you're going to see really creative strategies, really informed measurement. It'll only improve our client success once it's not just five AI experts at the company, but the entire company knowing how to use it. That idea acceleration is what I'm most excited about.
MB: I love the fluidity in the marketplace right now. I think we're getting much more agile — there are a lot more opportunities to shift dollars around. And by the way, I know we're trashing on television a little here — there's still a huge role for television in the media mix, and it's also becoming a lot more fluid. There are better options, and it's switching to CTV and other types of video inventory, but I think there's a performance mindset, and data is having a golden age right now, which is great.
JM: That's all for this panel, and I'd definitely love to talk more — that was such a fun panel. It's amazing to speak with industry leaders and hear directly from them about the future of our industry. Now I'm joined by my colleague Jen Loga, Vice President of Media Measurement and Analytics here at DeepIntent. Welcome, Jen.
Jen Loga (JL): Great to be here. Good job on the panel — I feel like that was a real front-row look at how fast things are changing, and how deep the challenges still are, and how far we have to go.
JM: Yeah, things are changing so fast. I don't think we would've spoken about anything a year ago that we spoke about today, and I'm willing to bet the same next year. Let's start with what I think was the big theme for the talk. Everyone wants to talk about AI, but what really matters is the data underneath. I loved what Mike Bregman said — how AI isn't the story, it's just the machine, it's the tool, and it's up to us humans to figure things out.
JL: Yes, I loved that, and I think what made it feel more tangible was how they broke it down by persona — the planner, the insights person, the media buyer, the citizen data scientist. Those roles aren't theoretical — they're real people within our teams right now, and AI is helping them move faster and smarter. I often say using AI feels like cheating on my homework, but I'm so much more productive and can do things much quicker without it — I mean, with it.
JM: Yeah, I like to say there's AI, and then there's RI — real intelligence. AI can democratize the insights, but it's really about the RI in interpreting it, making sure it's right. And underlying all of that is making sure the data is connected. I think Kate Gattuso-Duffy really hit that when she said we don't need more dashboards, we need data that talks to each other.
JL: Which really brings us to measurement. Five years ago we were still selling clients on the idea of moving past click-through rates — still a little of that — but now we're really focused on optimizing for outcomes. Are we doing that at scale yet?
JM: I'd argue it's bigger than it's ever been before. I don't know — what is scale, compared to anything we've had in the past? We are doing it at massive scale, but if scale is having the majority represented, I think there's still some scale to be attained. Still a lot of fragmentation, and a need for education on that fragmentation as it pertains to data measurement and what tools can and can't do. This is bigger than ever.
JL: Bill was talking a bit about risk-taking and agility — he mentioned smaller pharma brands, which is a good reminder that innovation isn't just about scale, it's also about mindset.
JM: And if we're really serious about personalization — and we should be, because personalization isn't just a CPG thing — it's getting the right message to the right person, which can be nuanced. Personalization in the world we live in is specifically finding the right person and making sure we're giving them a message that gets them healthy. And we do need to talk about privacy when we're talking about personalization. That conversation was refreshingly honest — no one pretended to have all the answers, but everyone agreed the guardrails are critical.
JL: I really appreciate how it all came back to the patient, and to the person — that need for privacy, while at the same time really trying to help people get on the right therapy faster. It's not just about the models and the metrics, it's really about helping people get better and feel better. I think that's what really excites me, and I think it excites you and the rest of the panelists too.
JM: Yeah, in the end that's all that matters. It's funny, because we do hear people criticize health marketing sometimes — people don't want to see as many ads — but I don't know if people are really doing the math to figure out how much, and how many people, are being made healthier and having their lives improved. That's so important.
JL: Well said. Thank you for letting me crash your podcast this week.
JM: Come back anytime. And for our listeners, we'll link to the full panel video in the show notes. Until next time — that was Deep.
Announcer (Voiceover): You've been listening to Deep, the health marketing podcast. Deep is a presentation of DeepIntent. Opinions shared by the guests represent their own perspectives, and not the views of their company or organization. If you'd like to learn more about the guests, the show, or the topics, check out this episode's show notes in your podcast app, or visit deeppodcast.com. Got a question or a suggestion for the podcast? Drop us a note — our email address is podcast@deepintent.com. If you like the show, please leave a comment and give us a five-star rating on your favorite listening platform, and be sure to subscribe so you never miss an episode. The Deep Podcast features original music by Diaphonic. The show is produced by Robert Haskett, with Ben Abramowitz, and hosted by John Mangano. Thanks for joining us.
Links:
- Bill Veltre on LinkedIn
- Kate Gattuso Duffy on LinkedIn
- Mike Caruso on LinkedIn
- Mike Bregman on LinkedIn
- Deerfield Group
- Pfizer
- SSCG Media Group
- Havas Media Network
- Jen Loga on LinkedIn
