The AI Advertising Podcast: S2
Episode 11
How AI is Reinventing Healthcare Marketing
About This Episode
Healthcare marketing is “hard mode”: strict privacy rules, heavy regulation, and measurement limitations that make the usual adtech playbook fall apart.
This episode explores how AI is changing the way healthcare and pharmaceutical brands reach both patients and healthcare professionals—while balancing relevance with privacy and staying compliant.
Jerrad Rickard | Senior Director & Product Leader, IQVIA Digital
Jasmaan Panesar | Senior Manager, Verticals & Performance, StackAdapt
Disclaimer: StackAdapt does not access or store patient-level data and operates exclusively using privacy-aware, de-identified audience signals activated through vetted data partners.
Transcript
Diego Pineda (00:00:00)
Healthcare marketing is one of the biggest—and most complex—ad categories.
EMARKETER estimates pharma spent $19.4 billion on online marketing in 2024, about 88% of healthcare industry digital ad spending.
And at the same time, AI adoption is moving from experiments to operations—McKinsey found 65% of organizations report regularly using generative AI.
So how do you use AI in healthcare marketing without being intrusive, non-compliant, or ineffective? And what happens when AI becomes the infrastructure that connects planning, activation, and measurement inside one compliant ecosystem?
Today, I’m joined by Jerrad Rickard, Senior Director and Product Leader at IQVIA Digital, and Jasmaan Panesar, Senior Manager, Verticals & Performance at StackAdapt—bringing the AI-powered orchestration and activation lens.
Let’s get started.
Podcast intro (00:01:06)
Welcome to the AI Advertising Podcast, brought to you by StackAdapt. I’m your host, Diego Pineda. Get ready to dive into AI, Ads, and Aha moments.
Diego Pineda (00:01:22)
Most industries want personalization. Healthcare has to earn relevance—without crossing a line. Here’s Jerrad Rickard on why healthcare marketing is one of a kind.
Jerrad Rickard (00:01:34)
I think a lot of the things that you and I would desire out of a normal marketing journey like that the person trying to sell to us or advertise to us really understands our needs takes on a whole different lens when you come at it from a healthcare care perspective.
And so it becomes important that the patient in particular kind of leads that journey by looking for that information first. And and unlike that use case where you’re looking at it from like a retailer packaged goods scenario where there are a ton of potential right options for solving the needs are very specific you know what I mean and very nuanced at times so you know if something I would look at that like if I’m looking for a dress shirt there could be a thousand options for a dress shirt for me. But if I have a specific condition, there might be only one proper therapy. So the information that you give is very different in that kind of scenario. You have to have a deeper understanding and at the same time not expose any of that data.
Because of that, there’s also been a lot of legal constraints and applications, you know, from a governmental perspective or, you know we operate primarily not IQVIA as a broad entity operates in 110 different countries, but IQvia digital’s focus is in the US and in the US, each state has their own governance on privacy. And so navigating that ecosystem becomes extremely complicated because you basically have 50 different unique sets of laws, everything from basically there’s nothing to very you know constrictive laws that have to do with you know exactly what data you’re allowed to use and how you can expose it and how you process opt-ins and consent and what is justified as consent is very different state by state.
And then the final thing is that, you know, like a couple other regulated spaces, I think about automotive and I think about finance, the pharmaceutical industry is probably the most regulated, but you have to be very cautious as there are guidelines for how you state things, when you’re allowed to say them, if you say certain things and you have to say certain other things in order to, you know, meet the standards, you have to have things like the important safety information included in the the marketing materials, or at least links to them. And so, things that again, in the retailer packaged goods space are done all the time, like dynamic content become a lot more complicated in this ecosystem.
Diego Pineda (00:04:03)
So if patient-led journeys and privacy constraints are the baseline… what does “personalization” even mean in this category? Let’s hear from Jasmaan Panesar.
Jasmaan Panesar (00:04:13)
The first thing that we need to acknowledge is that true one-to-one personalization, although it might be very doable in other industries like e-commerce, for example. In healthcare care advertising, it’s very difficult when we’re trying to deal with patients. And in many cases, like we actually shouldn’t be trying to do so just to not be invasive to like someone what tell you what someone’s like health conditions may be. um But in spite of that, there are many tactics that we could run. I would say like in terms of personalization, like I think one idea of personalization is like making the message contextually relevant. In a situation where like we don’t know someone’s condition or like if they’re undergoing some procedure, if they have a prescription, but we can take it to the context of the page in which they’re visiting and then show ads relevant to that page, which allows us to essentially be brand safe while also still personalizing the message and create it in some way.
I think another way that we can do it is by taking other signals that are privacy safe, but like just more… at a cohort level versus the individual levels. Think about leveraging, let’s say, a DCO product for geography or seasonality or like weather signals that can help us inform messaging if certain regions are likely to be experiencing flu activity rather than like knowing if somebody has a flu.
And then the final piece of the puzzle is obviously just like our third-party partners. So the way we operate is we don’t actually receive healthcare care data. What we receive is like privacy compliant audiences. And then all the healthcare data is managed by data companies. And then they actually take that data, model it out, and append other like demographic level data so that we’re actually able to reach people who are likely to have a specific ailment without actually getting that data and knowing who actually has that ailment. So I think like all these things kind of play together to have that privacy safe messaging, although like I’d say one-to-one personalization is not possible, we have ways to still make sure that the best message is in place for patient campaigns.
Diego Pineda (00:05:58)
That’s the “relevance playbook.” Now let’s add IQVIA’s data infrastructure side—how do you build audiences without mixing sensitive datasets?
Jerrad Rickard (00:06:09)
On the direct to consumer side, we launched a product in the fall last year using synthetic trend data. So artificial intelligence creates synthetic trend data in a federated model. So one of the complexities, again, we talked about all the legal implications of bringing sensitive data together in a consumer world, when we’re looking to target direct to consumer audiences, there’s a bunch of different methodologies for doing that in a privacy safe way.
One of them would be to to use healthcare care data or health data, you know, you’re what you’ve been prescribed and what kind of treatments you’ve been given, um information from your doctor’s visits, and then injecting a bunch of noise into that audience to create enough information privacy safety so that you could launch that audience. That has never been something that IQVIA would entertain, that just injecting you know additional people into that audience to make it so that you’re not identifying the patients specifically.
So instead we used artificial intelligence to run through all of the data and create synthetic trends. And then we have those synthetic trends and we marry that up with the healthcare data. And then we create an audience off of that, that is not identifying the actual patients.
Some of the patients are there, of course, you know, again, it gets to the same sort of thing, but you end up with an audience where You know a large portion of them are potential patients for this therapy, but some will not be. But that’s all done without ever bringing the actual healthcare care data into the same realm as the audience data. So they’re handled in a federated manner. And only what comes out of the healthcare data is synthetic trend data. And that synthetic trend data is then matched to the audience data. And we actually see those audiences perform largely better than the other methodology, which is, you know, marrying together the real health data with the audience data and then injecting noise into it.
Diego Pineda (00:08:04)
Privacy-safe relevance can still perform—and sometimes outperform. Now, where is AI creating real day-to-day leverage? That’s next.
Diego Pineda (00:08:19)
In healthcare, speed matters—but approvals and complexity slow everything down. AI is showing up in places that reduce friction.
Jerrad Rickard (00:08:28)
I think the number one is smoothing out a lot of the workflows. So It’s a very time consuming process to take a piece of, especially pharmaceutical content from you know kind of that genesis of your you’re writing those materials using the brand um guidelines and and sales materials that they’ve put into the print ecosystem or other parts of the digital ecosystem. Taking that and turning into a piece of digital content is a very long process.
Again, it has to go through lots of rounds of approvals. And then you have to queue it up and get it sent out in the market. And that’s just to do something in a broadcast way. But if you want to, you know, bring some intelligence into that kind of flow and create something of a trigger journey where you’re actually meeting the needs of the health care provider or the patient, there’s a, that could be a six, eight, 12 month process in the past. So, you know, leveraging AI to make smarter decisions about what are the five pieces of content we need to generate from this?
What are the areas of interest overall for the different segments in your target list so that you know exactly what types of content you want to generate means you can do those things much quicker and you can prioritize better.
Diego Pineda (00:09:48)
On the programmatic side, the question becomes: can AI help you target and optimize without relying on crude shortcuts?
Jasmaan Panesar (00:09:56)
Keyword targeting alone can be pretty misleading. Like a page might mention a disease or a negative, unrelated context and related to that disease, but AI helps understand what the actual intent and subject matter is in the page, which improves both brand safety and campaign performance. The other area I would say is bidding and optimization, which I would actually say is pretty much in line with measurement. The idea here is letting the AI ingest campaign performance data and like it’s actually giving you real-world context or real-world context into like what actually the report means. And then you can take that then actually have AI further ingest that data and actually help that it like bid and like provide suggestions into how you should optimize your campaign.
Then at StackAdapt, we do that through Ivy. So just Ivy being able to be like an AI agent that you can use to help understand your campaigns better. And then the key benefit there as well is it kind of makes things closed loop and it kind of reduces the amount of decisions that a human needs to make, um which is always going to be like a beneficial way that technology is growing. And then I would say the final thing and probably the most exciting for data partners and like how we work with data partners is we’re seeing a lot of LLMs getting introduced into audience modeling. So think of a scenario where instead of marketers needing to know exactly like specific ICD codes or CPT codes to target a DTC segment, they can actually describe their ideal audience in natural language. Something like, for example, patients recently diagnosed with asthma who were exploring treatment options. And then the system can actually translate that into the appropriate healthcare codes, which would then query the underlying claims data, and then would actually generate a model audience that can be forecast and activated very quickly in real time.
And this is is very powerful because in the industry today, like when we’re thinking about audience planning, it often takes like four to five business days, there’s back and forth with data partners, there’s a lot of manual intervention involved versus like leveraging AI as like a data partner is now giving DSPs direct access to query claims data without having to actually ingest the claims data, um which isn’t be insanely important for healthcare. And it just increases the speed to activation and includes our ability to optimize audiences and just test doing different audiences out.
Diego Pineda (00:12:02)
So AI is accelerating planning, improving contextual precision, and speeding activation. But then the reality of compliance rules hits, limiting creativity.
Diego Pineda (00:12:18)
Healthcare can’t always “generate and ship.” A lot of creative has to be approved. Even dynamic creative optimization or DCO has some constraints.
Jerrad Rickard (00:12:28)
You can’t just have AI generate content for you at will, it has to have gone to the FDA and been approved. So that kind of dynamic content gets interesting to say the least. And a lot of the newer advancements from a creative perspective, may not apply to the pharmaceutical industry. Now we did pilots last year, with email, where we worked with a dynamic content company and trained the artificial intelligence on the brand guidelines and what needs to be said. And we made some significant strides in getting some, you know, good dynamic content, but we had to put a lot of constraints around it.
Diego Pineda (00:13:06)
How do you run better tests in programmatic when even changing a color can trigger review?
Jasmaan Panesar (00:13:12)
We’ve explored the approach of essentially creating every single permutation but of a creative that we would think so that we can bring it in the MLR process ahead of time so that when all these all these creatives do get approved, then when we’re actually running ahead of our strategy, we have much more to work with and then we can actually explore some DCO options. Another key caveat for DCO, and obviously StackAdapt has a DCO offering, is that not every healthcare campaign is going to be like that heavily regulated. So like I think that’s more specific to pharma, but there are use cases, like for example, like OTC drugs or even kind of just like awareness campaigns where like we might not necessarily need to go through that spending process and then DCO still might be readily available. Then we can do things like for example, even awareness like around flu season, for example.
Diego Pineda (00:14:01)
Let’s talk about the patient journey across channels. How do you make it feel helpful and not invasive?
Jasmaan Panesar (00:14:07)
In healthcare, sequencing really starts with respecting how people naturally learn about a condition, especially for sensitive health topics. You want to begin with education and awareness and before moving toward more specific messaging around that specific drug, right? So channels like CTV or like connected TV and video are great for the first stage because they allow you to introduce a condition or treatment area in a broader educational way.
And it helps maximize reach because you’re not reaching a single person, or person you’re likely reaching an entire household that’s engaging with that content. And one tactic there that we could use is actually CTV retargeting based on ad completion. So if someone actually watches a video to completion, that’s a strong signal that the content resonated and you can thoughtfully continue that conversation on other channels like Display or Native where you can dig a little bit deeper and deeper into like what the drug is and provide even more further education down the line without it being feeling like an invasive process.
And then digital out of home is another interesting example, just because I’d say it’s naturally less invasive. If I were to see a digital billboard that has a pharma ad on it, I obviously don’t think that that’s specific to me or I’m being tracked.
So that’s just because of that, it’s naturally less invasive. And it’s a public awareness channel, so you’re not targeting individuals at that moment. So if someone is exposed to DOH placement and a ladder appears like on their mobile device in that same area, it also still creates the opportunity for us to continue the message later in a more personalized environment like display or native. so I would say focusing on the kind of sequencing that allows you to introduce a topic broadly and then reinforce it in more digital environments that are zoomed in into the individual can actually like help move patients along the funnel of like symptom awareness, exploring options, and then actual condition management versus just starting off on the bottom and like spamming specific i like identifiers online with with ads.
And then I would say another really important thing that I probably haven’t touched on is is frequency management as well. So with healthcare campaigns, it’s probably of utmost importance. um There’s a lot of emphasis on making sure that your messaging is not um overly aggressive and you’re not essentially like reaching the same individual multiple times or like… like over the course of like a specific day. So you want to be very deliberate with how you frequently tap your campaigns just so you make sure that it doesn’t feel like it’s invasive and like they’re seeing it spaced out instead of like all at once.
Diego Pineda (00:16:23)
And then there’s measurement. In healthcare, you’re often not getting clean conversion signals the way you do in other categories.
Jasmaan Panesar (00:16:31)
I would say there’s multiple things that we’re seeing as a platform. So one, I would say, problem area for pharma advertisers has always been that placing a pixel on a page is extremely difficult. So they’re often limited to top line metrics like CTR and CPM. There’s also other metrics that pharma advertisers use like leveraging prescription-less solutions or audience quality. The issue with these solutions historically has been that typically it takes like a lot of time to receive this level of reporting, so it makes optimization very, very difficult, but we’re seeing data partners make really, really meaningful strides in terms of like providing us with that data much more frequently, talking about like weekly reporting and like monthly reporting. And what that gives us the opportunity to do is, even if it’s on a cohort basis, it lets us optimize our budgets and shift them into like the more meaningful channels that are actually giving us those results. So we’re not just getting an end of campaign report, we’re actually able to optimize in real time and help guide and conversions down the pipe.
Diego Pineda (00:17:29)
Alright—let’s turn this into a practical checklist.
First: design for “relevance with privacy” from the start. Use context, broader audience groups, and compliant partner audiences to stay within healthcare rules—and treat healthcare professional audiences as their own track, where more tailored outreach is often possible.
Second: use AI to speed up the stuff that usually takes forever in healthcare. Things like planning, building segments, choosing which messages to run, and learning what’s working sooner. The bigger shift is having AI-powered platforms that bring your data, your media, and your measurement together in a compliant way.
Third: be realistic about testing. If creative approvals are slow, plan ahead: get more variations approved upfront. And if you can’t test creative quickly, test what you can move faster—like channels, formats, sequencing, and frequency.
Fourth: set measurement up for the world you actually live in. Pixels and perfect tracking aren’t always an option in healthcare. So build a measurement plan that works with privacy-safe signals, partners who can help, and faster check-ins—so you’re not waiting a month to find out you backed the wrong horse.
And finally: get the foundations right. Clean data, strong partners, and tools that let you pivot quickly—because that’s how you stay both compliant and competitive.
If you’re a healthcare marketer right now, the opportunity is huge—but the teams that win will be the ones who can move fast without crossing the line.
Podcast outro (00:19:14)
Thanks for listening to this episode of The AI Advertising Podcast. This podcast is produced by StackAdapt. Visit us at stackadpat.com for more information about using AI in your advertising campaigns. If you liked what you heard, remember to subscribe, and we’ll see you next time.
You May Also Like
Webinar
EMEA
Apr 15, 2026
The Personalisation Gap: What Advanced Marketers Do Differently
Webinar
NORAM
Apr 14, 2026
The Personalization Gap: What Advanced Marketers Do Differently
