Stephanie Wierwille: Welcome to the No Normal Show, brought to you by BPDA marketing services firm that delivers the future to healthcare’s leading brands. This show is where we leave all things status quo, traditional old school, and boring in the dust, and instead we celebrate the new, the powerful, the innovative, the bold, all focused around the future of healthcare, marketing and communications. I am Stephanie Weell, EVP of Strategy and Innovation, and I’m joined by Drew Marlatt, VP of Data Pro Products and Technology. Hi. Hi. Hi Drew. Welcome.
Drew: Hey Stephanie, so happy to be here. Thanks for having me on the show.
Stephanie Wierwille: I am so excited to have you, Andrew. It has been, uh, kind of a dream of, of the show really, because you are all knowledgeable about all things data tech and ai, and we’ve been talking a lot about AI and innovation on the show recently. Um, and you’ve really spent the last year or so building bpds own suite of AI workflows, AI agents.
You come to us from, um, a wide. Wealth of, [00:01:00] uh, tech experience, including Shopify. So I can’t wait for you to share all of your thoughts, uh, with our listeners,
Drew: Awesome. Yeah, it’s gonna be fun. Like you said, I’m, I’ve been hiding in the back rooms for way too long, right. So,
Stephanie Wierwille: with the robots, right.
Drew: right. Yeah, exactly.
Stephanie Wierwille: Um, all righty. So we have a, we have a fun agenda today. We are gonna cover a few headlines. First of all, we’ve got a nice little pop culture headline about a song that just came out from, um, from, uh, TBOs, from TLC to raise awareness. HIV pre prevention, so I can’t wait to talk about that. We’re also gonna talk a little bit about some new AI guidelines that came out in the medical space.
And then our main topic today is a really fun one, which is how can healthcare organizations innovate with ai and what are the challenges they face and how can they really, uh, leapfrog over some of those challenges? So before we get into it, just a couple quick notes. Uh, in a couple weeks we are partnering with E Healthcare on.
This topic, actually, [00:02:00] AI and Healthcare Marketing Week. We’re gonna have a full week of virtual programming for marketing and digital leaders, intentionally focused around hospitals, health systems, and provider organizations. And, um, so we’ll drop a link in the show notes. Please sign up for that. It’s five full days of sessions.
I think there’s like 10 plus sessions and, um, it’s gonna be a great time. And Drew, you’re gonna be, you’re gonna be on one of, uh, two of those sessions talking about AI agents and workflows.
Drew: That’s right. Yeah. Once as a participant, once as a moderator, I’m super looking forward to it.
Stephanie Wierwille: Me too. I can’t wait to, um, can’t wait to see all the case studies and hear all the discussions.
Drew: Oh, for sure.
Stephanie Wierwille: and then also we’ll drop a link to our latest report, which is called Crossing the Einstein Divide. We talked about this quite a bit on our show, and I think today’s topic is related to it. We’re not going to go into depth on the paper at all, but just kind of a relevant topic around innovating with ai. So we’ll drop that in the show note as well. Um, but before we get into it, drew, since this is your first time on the show, would love to learn more about you and maybe, um, you know, we’re coming [00:03:00] right off of Thanksgiving weekend. So I’m really curious, what are some of your favorite Thanksgiving traditions, dishes?
Uh, what’s the vibe for you?
Drew: Oh, for sure. If Thanksgiving is by far my favorite holiday, I don’t know what it is. I mean, Christmas is great too, but I feel like Thanksgiving, uh, has the personal connection that that Christmas doesn’t necessarily lack, but maybe gets overshadowed by. Other elements of tradition and commercialism, et cetera, right?
So I feel like Thanksgiving being all about gratitude and family and, and most importantly, food really connects to my core interests. You know, um, I’m an advent home cook. I, I obsess about making the Turkey, so I’m so glad to have just been able to smoke a Turkey for my family. Um, I go, I live in the Midwest now, but I go full southern when it comes to Thanksgiving.
So smoked Turkey, cranberry, barbecue sauce, Barbe, excuse me. Smoked Turkey, cranberry, barbecue sauce, and uh, cornbread Stuffing are the must haves for sure.
Stephanie Wierwille: Cranberry barbecue sauce is something I
Drew: great.[00:04:00]
Stephanie Wierwille: try. That sounds amazing.
Drew: Oh, it’s awesome. Very tangy. Yeah.
Stephanie Wierwille: It’s got all the things, the
Drew: Yeah.
Stephanie Wierwille: the fruit. That’s brilliant. I’ve never heard of that. That’s a first. And it’s, you know, to me sauce is, sauce is thing amazing. So.
Drew: Yeah, I feel like cranberry sauce gets overlooked. A lot of times people just leave it out of the ocean spray can and it still is in the can shaped mold, you know?
Stephanie Wierwille: Yeah.
Drew: you? What’s your favorite part about, oops.
Stephanie Wierwille: I love a carb holiday. I love a carb in general. So, um, for me it’s the dressing, it’s the mashed potatoes, all of that. Um, but yeah, I think I’m with you that Thanksgiving is the best holiday and I’m going to try to start a trend next year of Thanksgiving. Decorations like blow up inflatable turkeys in the front yard. I think we have an, we have, you know, hall Halloween decorations have gotten wild Christmas decorations have of course been wild. Um, we need to blow up Turkey in our lives. That’s my. Take.
Drew: Yeah. Yeah. Give Thanksgiving its space. Give it, its due. [00:05:00] Absolutely.
Stephanie Wierwille: Yeah. Um, alright, well good to catch up with you a little bit. Um, so let’s hop into our headlines. Uh, we, we found this headline, actually Ra, who was on the show last week, shared this headline right after she was on the show. And I think it’s just such a cool example of healthcare showing up in pop culture.
So, anybody remember Drew, do you remember TLC and the song Creeped from 1994?
Drew: Oh yes. I think you and I are of a similar age, so this is right in my sweet spot.
Stephanie Wierwille: It, it’s a great nostalgic moment. Um, so creep was one of TLC’s bigger songs. Um, and so TBOs from TLC teamed up with Gilead Sciences, which is a pharma company. However, they are going to market as care for the culture. So kind of like a brand forward, you know, um, not Branford. Where the brand is kind of hidden, you have to dig to see who created care for the culture.
It’s Gilead Sciences, which is the pharma company that is known for the, um, HIV prevention drug prep. Um, and so, uh, TPAs teamed up [00:06:00] with them to create a song, a remake of creep that raises awareness for HIV prevention Prevention. So instead of the original lyrics, which was soy creep, it’s now soy prep. Um, and kind of a, a great awareness song.
Um, I. Personally love this because to your point, drew, I’m a millennial. I love a throwback. I wanna live in the 1990s. And you know, I think it’s always an interesting play to say, how can we make something like health awareness, memorable and lyrics are a good way to do it.
Drew: Sure.
Stephanie Wierwille: What, what was your thought on this?
Drew: I completely agree. I think it’s so nice and we see this all the time, but I, it’s so nice to see health systems, like innovative ways to push preventative care specifically. You know, it’s such an important part of making sure that the entire health system and our whole population. Um, is more healthy, more productive, uh, better quality of life, all the things, right?
So like to see that done here and, and, you know, HIV tr treatment [00:07:00] and prevention is such a huge win for the pharma in industry in the last few decades that, I mean, it’s a great highlight for sure.
Stephanie Wierwille: Yeah. And you know, you mentioned health systems and I think we have several health system examples that we’ve talked about on the show that lean into culture. I mean, one example we talked about a few weeks ago was Baptist Wealth. Baptist Health. Um, health is wealth, right? Baptist Health and uh, zoo time. Um, so that was a great example of maybe a celebrity forward partnership. Um, so there’s, there’s a million ways to do this, uh, and you know, whether it’s pharma, whether it’s health system, whether it’s association, there’s all kinds of ways big and small. Doesn’t have to be a TLC budget necessarily. Um, but anyway, I just think it’s always a, a fun way to go.
Drew: Absolutely. Yeah.
Stephanie Wierwille: Another quick headline. Um, so we found this past week this AI code of conduct that was released for health and medicine. This was released by the National Academy of Medicine and they created an AI code [00:08:00] of conduct and it gives six. Foundational commitments on everything from ethics and health equity and effective and responsible use of ai. So what I think is really interesting about this is it includes a collaboration of authors such as the Permanente Vanderbilt, Mayo Clinic, Memorial Sloan Kettering, Johns Hopkins, all the biggest names and healthcare, and they all collaborated to say, Hey, AI is moving really fast. aren’t enough guidelines and guardrails for responsible AI use. Um, and so this is a very long paper. We’re not gonna spend a lot of time on it. Um, but for anybody who wants to dig into the weeds of it, we’ll link out. And I think that, you know, the point here is just as the tech continues to race ahead, it’s really important for folks to collaborate and think about the larger, more overarching guidelines. Um, that’s critical so that healthcare can leverage technology, um, in the right ways. Um, drew, what did you, what did you think about the code of conduct?
Drew: I think it’s, I think it’s a really great way [00:09:00] to address the core issue right now, which is that. Responsible AI is first of all a great passion of mine. I, I bill myself as a responsible AI evangelist, right? I think this is something we can and should be doing in a very thoughtful way, and it’s a, we’re in a space where I would argue responsible AI is still much more of a definitional problem than a technical one.
Like we have to be very careful and thoughtful about how we say, it’s easy to say we wanna be just, and fair and transparent, but then we have to say, you know, what is justice? What is fairness? Who are being fair toward, what are we being transparent with, um, and all the things, right? So, uh, this is a really good framework for being comprehensive.
I think, not only on the research side, but on the clinician side too. They cover a pretty wide variety of use cases, so it’s, it’s definitely worth the read. Um, for me, like there’s four big categories that. At a minimum any good AI policy should, should touch on, especially in healthcare, data privacy should be number one, [00:10:00] always, both from a regulatory and a responsibility perspective.
It’s just ethically the right thing to do to protect our patient’s data. Um, the veracity. Of AI’s outputs as well. So is it allowed to lie to you? It shouldn’t be. So you should control for when and how you check for that. Um, is it exhibiting toxic behavior or is it putting something out there in the world or into your patient population or your clinical population that you don’t want to see in that, like influence in that way.
Um, and is it complying with, with, uh. Intellectual property, both internal to your organization, protecting your IP, and also not violating others’ ip. And that’s still an open question about where those boundaries are, what’s considered fair use, what isn’t. Luckily it’s less on the. Um, it’s less of an issue on the like provider side, on the company side of things.
The, the folks building a AI enabled apps or building AI enabled workflows. We worry about IP compliance a little bit less than folks who are building the next foundation model at Anthropic [00:11:00] or something. So
Stephanie Wierwille: That was a great overview and um, we could probably spend all day digging into all the things that you just mentioned, but it is critical. But more importantly, it’s really hard. It’s really hard to,
Drew: it is.
Stephanie Wierwille: this, this, this conversation requires. Requires philosophical conversations. It requires debate, it requires technical knowledge and understanding.
It requires cross disciplines, you know, especially in healthcare, in the room. Clinicians, you know, I know there’s a lot of discussion around the importance of peer reviewed, um, science and bringing that forward. So, um, I think you just unpacked it in a really beautiful, simple way. And I’m just adding like this, if it’s not something that, you know, marketers and communicators are involved discussing right now, organizationally, I think that is an. goal for 2026 is join the conversation in your organization about responsible AI use.
Drew: So glad you said that. ’cause it’s, it’s definitely one of those topics that’s not a, not somebody else, it’s not my problem kind of vibe that doesn’t work here like it. To do this well, [00:12:00] companies need to foster the right culture and they need to bring everybody in at some point in the process, or at least, or at least embed the mindset for everybody in their workflows.
Stephanie Wierwille: Yeah. So maybe that’s a good transition to our main topic because what we’re gonna talk about here in the main topic of the day is what are the challenges that healthcare organizations face when they’re thinking about innovating with ai? Um, there are many challenges. Responsible AI uses one of them, um, a huge one, but there are others as well.
There are cultural, there are people challenges, there are tech challenges, their budget challenges. It’s all the things. So we are always coming at this from kind of a marketing lens. However, more and more what we’re realizing is. You know, marketing is, um, the purview is expanding. Marketing is getting more into things like patient communications, patient experience marketing has a right to bring forward the importance of, um, leaning into AI organizationally in, in terms of how are we communicating this internally and externally, but also, you know, what are we doing operationally to make that patient and employee experience more [00:13:00] meaningful?
So, um. Big, you know, ball of wax here. Um, but why don’t I just start at the beginning, which is it’s the end of 2025, I think. We at BPD have been kind of raising this flag really loudly that hey, marketing is rebuilt from the ground up. For all the reasons I just mentioned. single marketer, every single business leader has to rethink what they do. Um, and so when we think about rethinking it, that’s kind of where the paper I mentioned earlier comes into play. We kind of think about it in two parts. One is, how are you moving faster? That’s critical. We call that Edison thinking, you know, how are you, um. Incrementally improving. How are you building a better light bulb? Um, that’s really important. It’s the work that most marketers are thinking about right now. What folks are not talking about as much is what we call the Einstein approach, which is, you know, how are you re-imagining the universe? Um, and so that requires some big, crazy dreaming. And when you. Think about innovation, both in terms of [00:14:00] marketing communications and organizational innovation.
You have to have both approaches. so I’ll just pause here and ask you, drew, you know, I know this is a core part of your role, your job, and you think about this every day, both for BPD and for health systems. Um, are there any, uh, kind of things you would add just sort of to the overall landscape of why this is happening before we dig in?
Drew: I think that was such a great, uh, summary. You know, it’s a, it’s a shifting sans mentality right now. Things are moving so fast. Uh, it’s been a tool that has democratized so much. So much capability that wasn’t there before that was locked behind walls. So people are eager to take advantage of it. And I, you’re totally right.
I don’t think it’s an either or. I think it’s a yes, both. You know, you need to have strategies that go from the bottom up and from the top down.
Stephanie Wierwille: Yeah, and we’ve done a lot of dreaming on this podcast, so anybody who’s a long time listener knows we have. And I kind of gotten sometimes in the clouds around like, what if this could be possible? [00:15:00] But I think what’s interesting, and maybe it’s timely since we just had Black Friday, but um, you know, probably over a year ago we talked about agents shopping on our behalf, agents, you know, being able to crawl the web and not just know like, what do I wanna buy, but actually buy it for me?
And that’s happening. Right Chochi BT has shop their shopping features, their, um, shopping research features, um, the Atlas, the, you know, kind of agent mode. All of that’s happening. So I think as we sit here today, drew, I’m very curious if you were to, I don’t know, I like to predict, um, if you were to predict at the end of next year, are there things that you think that we are kind of dreaming about right now that might be a reality with the tech next year?
Drew: Oh yeah, that’s such a good point. I mean, the age agentic ai, especially for shopping, I love this idea for like even my grocery shopping by the way, like. I would love to just feed at the links for the five or six recipes I wanna make in a week, and it just goes and puts in a order to my local circular market and I just pick it up.
That would be great. Um, but that’s possible today for, for next year. I think that the, the agentic [00:16:00] usage, especially in the browser, it’s gonna upend a lot of. First of all it infrastructure that we’ve relied on in the past. Um, I would argue the last decade there have been a, a lot of focus projects on robotic process automation, RPA.
So like how do we automate things inside of browsers? And it’s a very complicated and brittle process. I think this is gonna completely eradicate that as a technology when you can just tell your favorite AI tool. Okay. Uh, log into this system for me. Go fill out this web form. Go grab this information, go scrape these articles.
Whatever the case may be. It may even eventually start getting rid of things like APIs and SDKs when you can just. Have an AI agent go crawl through the website itself and just use the front end as it was a user. Um, specifically at the clinician space, I’ve heard a few, uh, friends in this area talk about the rise of the need for AI doctors even so I.
Right now, there’s a lot of folks experimenting in the clinician space with [00:17:00] things like automatically taking, uh, recordings of minutes or transcribing meeting notes. When you’re talking to a clinician, I mean, we’ve all been in a doctor’s office, right? And seen them typing away while they’re half listening to us.
Um, and they’re very, very good at that. But if you take that away from them. Let the machine do the transcription, the writing, the prescriptions, the uh, uh, aftercare summaries, the the treatment plans, all of that, and let the clinician just focus on having the conversation. That’s the low hanging fruit that people are trying to tackle right now in a lot of big and small companies.
Um, but as we’ve seen, like what does that happen when, you know, you, you add a new vector to the treatment. Uh, right now, you know, you might have an issue. You might go to a, uh, um, a primary care doctor for a certain class of issues. Maybe you do a virtual appointment, maybe you go to urgent care or the er.
What does that fourth one look like? Where you just ask an AI doctor instead? And I think that this is a weird case where. Um, the users are way further ahead than the [00:18:00] health systems are. Um, I think users have already been doing this in the last few months with chat GPT and others. Um, and WI think we, uh, could talk about how chat GPT is actually trying to roll that back and curtail that behavior, uh, for liability reasons.
But, um, I think the users are there. I think a lot of people want this either out of necessity or out of convenience. What I wonder about is when will the technology be reliable enough that both the providers and the payers are comfortable with the level of risk that, that this introduces? So it’s really more of a, an organizational question than a technical one at that point.
Stephanie Wierwille: Wow. You just, you just unpacked almost a whole roadmap in a way. Um, let me summarize quickly and see if I heard you right Drew. And then we can kind of unpack like how to get to this mindset get here is harder than it seems and we can unpack the challenges and, you know, kind of like the ways of thinking.
So, um. Tell me if this sounds right. Based on what you just said. It sounds like you said ambient listening is already here. That’s happening. It’s not [00:19:00] just being piloted, but it’s been scaled at a lot of health systems. So in the next year we’ll see that rollout. Probably at mass. We are already seeing people, consumers, patients using AI tools for healthcare research, but not yet.
And we expect this will come, sort of that more agentic use of like booking appointments, filling out forms, like that will come. then what you’re saying that’s kind of like the furthest from where we are is how would you, gonna use the word build lightly, right? Because it may not be a direct build, but if you have ambient listening today, how do you leverage all of that data and those kinds of models to then allow. Provider like communication provider like is important. Uh, verbiage there provider like communication with patients when the provider is not there. IE AI doctor that my take is, we might see pilots of that next year but probably wouldn’t actually see in full until 2027. Tell me if you feel like that’s a decent [00:20:00] summary of your
Drew: Oh,
Stephanie Wierwille: sorts.
Drew: on. Yeah. And I totally agree with that five check at the end on timeline. I think it’s something where the health systems themselves will wanna thoroughly de-risk it before they really let people do that. But imagine that AI doctor that already has and already has read your medical records. Um, of course built inside of the ecosystem of whatever your EHR platform is, right?
Or at least interfacing with it. But, um, the responsibly built tool that can see what you’re talking about. Take those notes from whatever conversation you’re having with it, update the medical records accordingly, pass those on to specialists. Right. So I’m thinking about like, you know, you know, let’s say I am playing soccer and I, I, I hurt my knee and I know like, okay, I’m gonna go to the, the primary care.
I’m gonna talk to him about it for 20 minutes. And he’s gonna say, here is your referral to orthopedics. Maybe the AI doctor is the circumvention of that, where they can just give me the referral to orthopedics, right? Without the intermediary step. It’ll be interesting to see [00:21:00] where the, the dice lay on this one.
You know?
Stephanie Wierwille: It is, and yes, and the dice. It’ll be interesting to see where the dice fall in terms of both physicians, you know how physicians feel about it, which right now, early signals show there’s a lot of appetite for it because physicians are overburdened and there’s not enough of them. It’ll be interesting to see where it falls in terms of consumer perception. So we have a long road to go to get to what you just talked about. So why don’t we, now let’s talk about what are all the challenges in place for that kind of thing. It may not play out like we just said, right? Those are examples. Those are dreams, those are thoughts. Um, there’s a lot. There’s a many a million barriers.
So let’s talk about the
Drew: So true.
Stephanie Wierwille: I’ll, I’ll throw a few out and then you can build, um. I’ll just throw out a few, uh, here, which is in healthcare, you know, we don’t have the capital for all the pilots and pilot programs that, you know, an open AI may have or even a retailer may have. Um, and speaking of that, uh, you know, ex having an experimental mindset is key and budget and experiments go hand in hand. [00:22:00] So we know the importance of pilots and pilot programs. Um, I’ll tie back around to that ’cause I know you have a lot of thoughts. There as you’ve been somebody who’s been doing a lot of those, um, types of, uh, experiments, if you will. Um, healthcare is an operations industry. It should be, right? We are. We cannot tear it down to the studs and rebuild it, nor should we ever do that.
That is not something that especially health system leaders can even think about. So that’s why we end up. With more incremental innovation. And it’s like, what can we add? What can we edit versus how might we completely rethink? Which is usually, you know, the, the mindset we’re thinking about here. Um, and then we’re risk averse, which again, we should be none of these, I’m not calling these out to say, these are bad.
I’m calling these out to say, these are truths. So as we consider, you know, how do we get from here to there? These are, these are constraints that we. To, uh, be mindful of, um, let me
Drew: Yeah.
Stephanie Wierwille: Drew, anything you wanna talk about, especially in regards to the experiment, experimentation, mindset, and what that looks like in maybe your past life at a company like, [00:23:00] let’s say Shopify or, you know, an
Drew: Yeah.
Stephanie Wierwille: and what that looks like in a, in a healthcare, um, setting.
Drew: Absolutely. I would love to. Um, you’re totally right. Healthcare is unique in that it’s like. It’s half infrastructure and half deeply personal, right? So like it is foundational to us as a society, and it’s also so critical to the individual. So like the, the risk is just so much higher when you have either a research or a clinical use case versus what we do with marketing, right?
Not to the belittle what we do, but no one’s life is on the line. There are PR disasters that could happen, but like, that’s about the worst of it, right? So we definitely have more room to fail than a lot of other, uh, domains. But, um, with experimentation. So there was a recent report from MIT that did a study.
They found that the big headline, this is probably I think late August, early September, the big headline was. 95% of AI projects and pilot projects specifically fail. Um, and a lot that picked up a lot of noise and the byline that a lot of people ran with was, uh, AI is not working out the way we thought it was, [00:24:00] and I think they’re missing the point.
My, my take on it is that 95% of AI projects are failing because they’re meant to fail. And there’s a lot of reasons for that. Um, this is an area of rapid innovation. There is a lot of small players. There are a lot of companies trying to do everything all at once for a lot of reasons. Part of it is, let’s be real senior leaders at company executives.
They are people too, you know, and they are, uh, prone to FOMO as everyone else is, right? So like they’re afraid of waiting for the dust to settle and them being too late to the game to be a major player, right? So people are willing to take risks on things that are a little bit more cutting edge. Um, a lot of it is an appeal to an investor group, right?
So a lot of times. People do want to see that you’re using ai, so they’re not afraid that their investment is isn’t going to play out. And then the third and most important thing is people really are just trying to like do what’s best for their organization. And right now, you know, it’s worth it to take the moonshot at sometimes, I suppose.
Um, there’s a few really cool. [00:25:00] Findings in that paper though, that I’d love to call out, one of which was that they, they found that there is a big differentiation between where the idea originates. So pilot projects that start from the top are less, are more likely to fail, but they’re also more likely to be big thinking blue sides.
And some like, or excuse me, Einstein, like projects that we’re promoting, uh, the ones that are more likely to succeed are the ones that start near the bottom of the org chart. Um. Sorry about the dog. Yes. Hopefully. Would you like me to stop to take care of that or no? Okay. Cool. The, the projects that tend to succeed are the ones that start lower than the org chart.
The people who do, uh, the actual hands-on work, when they emerge ideas for AI projects, they’re really tactile. They’re very small scale. They’re more s and like projects that are trying to refine sticky workflows, and they are more likely to add value, mostly in, in the form of cost avoidance. Right.
Stephanie Wierwille: Yeah, so you just beautifully shared one of the solutions, I think, which, um, [00:26:00] let me come back to that, which I think, you know, it’s what you just talked about is how do you actually build an experimentation mindset, and it has to be top down and bottom up. Right? If I heard your takeaway. From what you just mentioned, in order to take this, uh, pilot and experimentation approach, you have to, number one, recognize that there will be some failures. You have to merchandise that so everybody is aware and they know what to expect. And then you have to create this kind of cultural rethink, which is a top down, um, encouragement for big in the bold and a bottom up experimentation for moving faster. And, um. Uh, and all of that. I’m just gonna add to the MIT study point. Um, one thought I had as you were talking about the MIT study around how, you know, AI pilot pilots are meant to fail. which maybe the report didn’t highlight that part. I think it was Jeff Bezos who was recently talking about, you know, are we in an AI bubble or not? Um, and he gave a great analogy, which I really enjoyed, which was, if you go way back to the railroad, um, [00:27:00] building times, that was a huge revolution of infrastructure and. One of the ways that it became successful was everybody got into the railroad game because there was so much money to be made, and there was so much building that some of these tracks were literally side by side and duplicative. So there was waste in the revolution. And I think his point was there has to be waste. You can’t revolutionize, a society industry’s business without having some waste along the way. Now we can stop and pause and go have a cocktails conversation about the, you know, ramifications of that, all of that. But I think just to, you know. Think about this from a organizational standpoint, it is for healthcare.
Who doesn’t have the capital will make major pilot programs. It is. What can you do that’s really small? So some examples are, um, to your point of bottom up, right? A marketing team I think is in a great position to have a top-down, bottom up approach. [00:28:00] Um, top down, Hey, let’s have an overarching vision for our department around what we wanna do.
Bottom up, everybody go experiment and come up with, you know, very small experiments that actually are not. Costly at all. Just how can you automate your workflows using the tools you have? Um, what is one, one business problem that maybe you can find that’s relatively small that you can come up with a 90 day pilot that we can go test?
None of those have to be costly. In marketing back to your previous point, we are not talking here about clinical applications of ai. We need to know those. Um, but you know, we don’t actually have to have to, um, be creating those on our side. Um, so, you know, I just think I’m unpacking a little bit more of what you said there.
Drew: Absolutely. Sorry, administratively. Is it okay if I take two minutes to check the door real quick? Thanks. I’ll be right back. Sorry about this. Yeah.
Stephanie Wierwille: Yes. We’ll edit. Go for it.[00:29:00]
Drew: I’m so sorry about that. The cleaning crew came at the exact wrong dime.
Stephanie Wierwille: No problem.
Drew: Yes.
Stephanie Wierwille: we, we deal with that kind of thing all the time. No problem. Um,
Drew: yeah.
Stephanie Wierwille: I’m just gonna, I’m gonna set up kind of the next part of this conversation here and our, we’ll wrap the last part.
Drew: That’s great.
Stephanie Wierwille: if culture needs to be changed from the top and the bottom. Top down vision from leadership, um, C-M-O-C-S-O and beyond bottom up approach from [00:30:00] all the team members across the department.
Let’s talk a little bit about how you do that. Um, because I
Drew: Yeah.
Stephanie Wierwille: is the, to me, in all of our conversations we’ve had with healthcare organizations about their AI approaches and their AI initiatives, especially in marketing communications, I think the thing that I learned over and over again. Might sound obvious, but it’s actually very hard and it’s a lot to unpack is at the end of the day, it’s much more about changing your culture than it is about understanding the tech.
Yes, you need AI literacy. Yes, you need education across every single team and team member. Yes, you need the right tech and tools in place and you gotta have a strong partnership with legal and it. That is this first step, but the much longer battle is changing culture. So let me just ask you, drew, you know, have you seen anything in your experience that you think is really impactful in terms of creating an experimentation mindset, encouraging that bottom up, grassroots experimentation approach, encouraging teams to dream all, all of that.
Drew: Absolutely. I mean, you, you nailed it, right? [00:31:00] Culture comes from the top always. And here the key culture is how do you build thoughtful experiments in a way that are designed to succeed? Right? So I loved your idea of the 90 day pilot. Really the, the key there is, it doesn’t have to be 90 days, it could be 45, it could be 30, could be 180, as long as you time box it and, uh, you, you build something that has, um, a measurability component to it.
So where, where projects tend to fail is when they don’t do a good job at the start of defining the scope. Or to figuring out like, how do we measure success in this? And that can be a really hard thing to do, right? Especially when you’re dealing with a pretty vague, uh, subject matter. So maybe it’s something like patient experience scores or appointment availability or, uh, you know, mean time for an appointment or something like that.
Uh, it could be a lot of things. It really depends on your use case. Uh, for us, for, you know, cost efficiency work, we tend to focus on things like. How long does it take for a certain workflow to go? So, uh, you know, we’ve done some work [00:32:00] on marketing strategy that we’ve been able to take a week long workflow, scan it down to just like less than an hour, you know, um, and that was a very long, complicated process to get there.
But, you know, it took like, like you said, probably about 90 days and we knew what our goal was from the very start. So as long as you’re willing to do that, you can take some risks more effectively. Um, I think the, the old school stat I heard was that the two most likely ways that a project will fail in general, even outside of AI projects is scope creep or research or resource starvation.
So resource starvation is less of an issue here because these things are so well democratized. It’s very easy for you or I to go into chat GPT or something and like. Build a pretty small proof of concept just to show, Hey, this is possible. We just need to find the right data, the right context, wire it together with the right technology.
Um, but for the scope creep thing, that’s where these things really do struggle because you define, we want it to do A, B, C, [00:33:00] and then you go present it to a stakeholder after 45 days and they say, that’s nice, but what about DEF? Right? Why isn’t it doing that yet? So you need to be very clear with your stakeholders what you’re trying to accomplish.
How good is good enough? And um, I used to um, always hear from like professors in grad school. Keep it sophisticatedly simple like the goal is parsimony. You wanna find the simplest solution that accomplishes your goals and make it no more complicated than that. ’cause as you add more complications, you add more risk.
Stephanie Wierwille: Yeah. Um, so I think one of the things you said in there that I think is there’s a million great takeaways from what you just said, but I think one of them that sticks out to me is, you know, you mentioned like some of the examples that BPD has that you’ve been building in partnership with your team here at BPDI.
Think those are really analogous to health system. You know, pilots that might happen. Team. So for us, the benefit to our clients and and customers is they need, they need projects yesterday. They, you know, the world is moving fast. They need work that’s [00:34:00] really fast. Um. And they need it. That’s high quality.
They need it with high expertise. And so kind of our principle is how do we start on third base, if you will? How do we start all the data at our fingertips and then leverage our human expertise and talent in order to get the client what they need? E meet. Your point about hours or days, not weeks, is a major, you know, customer or client benefit.
Now take that same mindset and put it inside a marketer’s mindset at a health system. The, I think starting with the problem you’re trying to solve, I know it’s a duh, but it’s like rarely done. Um, so we just ran a workshop at HCIC to help health system marketers. Uh, create their own pilot projects. So what we
Drew: Yeah.
Stephanie Wierwille: we started with a discussion in the room about what are the business problems you need to solve?
And sometimes they have their own clients. Their own clients are service line leaders, physicians. Sometimes that might be an internal problem you need to solve. And sometimes it might be a patient problem you need to solve, and sometimes it might be a business problem you need to solve. So we had a really good.
[00:35:00] Robust discussion in the room at HCIC about what are those business problems that are small enough but impactful enough to solve. And then we went from that all the way to what’s a 90 day pilot. And it was so cool to see people’s wheels turning and to see, you know, them working with their chat bot to figure this out.
And what’s exciting is. You know, 2025 has been a year for so many health system marketing teams to be experimenting and piloting inside their own functions. Um, and I hope that 26 is about scaling those.
Drew: Totally, and that’s a good call out, right?
Stephanie Wierwille: those convos.
Drew: I love that. That’s such a great call out that like the 90 day pilot isn’t about profitability. The 90 day pilot’s about proving what’s possible. You know, so like you, you have success there and then you, the success in that is learning that it is feasible enough that it will be profitable at scale, and then you can take the next steps to make it scalable.
Productionable in healthcare marketing specifically, we see this all the time where 98 isn’t enough time to see an effect on [00:36:00] patient populations At. All we’re talking about scales more unlikely on like 18 months or more, you know? So you start small, you design an experiment. Maybe it’s a simple AB test or something more complicated and you figure out, hey, is this gonna work at all?
And if you see the signal, you go for it. Invest in it. Yeah.
Stephanie Wierwille: Yeah. Okay. Well let’s wrap on that call to action. I love that. Um, which is how do we, how do we. One, experiment and pilot. Number two, how do we scale? And number three, how do we make sure we’re thinking long term? So thank you. Thank you. Thank you for joining. Drew, this was a blast. You and I could go on forever.
I just love hearing what you have to say and, and pulling from your expertise.
Drew: Likewise, Stephanie, this has been great and uh, I hope it’s the first of many.
Stephanie Wierwille: That’s right. All right, so for everyone listening, uh, let us know of, you know, your pilots. What are you experimenting with? Are you excited for, for 2026? Um, drop us a note if you would like. And until next time, don’t be satisfied with the normal. Don’t be satisfied with your run of the mill pilot. Uh, really push [00:37:00] it, push it to scale, push your culture, and, uh, push the no normal.
And we’ll talk to you soon.
