Introducing predictive AI, a new approach to targeting before health events occur

[00:00]
It increases high value activity on brand.com it provides near human MLR compliant conversations and it delivers real-world data and Analytics back to the manufacturer based on conversations.

[00:24]
Hi, this is Marc Iskowitz editor large for mmm and I’m excited to be part of this sponsored podcast with Swoop introducing predictive AI a new approach to targeting before health events occur. Soup is a leading provider of patient and hcp audiences which they build using artificial intelligence and they’ve got a lot of new offerings that they’re bringing to market including having launched a predictive AI targeting solution and conversational AI to enrich the targeting process on both the hcp and the patient sides and my guest today Katie Carr EVP chief revenue officers going to tell us all about that.

[01:04]
Katie how are you and welcome back to the M&M podcast hey Marc great to be here. So good to join you again for those who are not familiar with soup. Can you give people out there a quick overview sure so Swoop is a leading provider in Custom solutions for both pharmaceutical manufacturers and their aor’s so agencies of record and everything that we do today is powered by AI and real world data our Solutions can range from anything from DTC to hcp audiences all the way down to on-site conversational AI and ultimately Mark what we’re trying to do here is improve health outcomes and lower health care costs overall great. I mean Italian to the triple aim is certainly put you on the right right side of the equation for sure.

[02:00]
And uh the whole rizzandia tray of the company building custom audiences and these new product offerings how facilitate that so let’s get into a chowie you mentioned Swoop uses real-world data for targeting. Can you share an example of how this approach works?

[02:20]
Targeting we use real-world data.

[02:24]
To reach and engage patients and also they’re associated hcps and historically that’s always been based on what has happened in the real world health data, so an example here would be a patient’s diagnosed with the condition. We may want to target that patient and or they’re treating team right and we can find these people while they’re just living their daily lives right Mark so they could be on a podcast such as this or maybe they’re watching TV or checking the weather on their mobile device. They could be interacting with friends on social media or just searching the web. This is all really effectively done in a very privacy safe manner okay great. That’s really good to hear what kind of outcomes have marketers seen using Swoop’s targeting approaches.

[03:17]
Yeah sure, so we see you know increased visitation rates. We see lift and conversions overall. We see really strong audience quality in Target multiples, which really correlate to script lift and ultimately many of our pharmaceutical manufacturers are looking to drive script lift and that’s what Swoop does really really well. We do this for both our Healthcare professional segments as well as our DTC segments and all of our campaigns or for the most part all the campaigns that we work on are measured by some type of third party measurement company to really validate this of the program and the segment provider itself.

[04:03]
Yeah, that’s important for marketers obviously to be able to point to you know script lift as measured by a third-party measuring company for their making sure that their campaigns were successful. That is apply to both the hcp and DTC campaigns and can you share some examples of how client results have been?

[04:25]
Yeah full of a recent household addressable case study we worked with a well-known brand that had been consistently running in the household addressable space where you know they tend to see more consistent to flat results quarter of recorder this brand is a rare. I would say sensitive condition which can lead to increased patient hesitancy so to be able to really move the needle quarter over quarter was something that the client was going to be very happy with we ran this campaign for six weeks and we collected over three months of attribution and the results were really quite astonishing from both DTC and hcp perspective so we saw 20% lift and visitation rates so 20% lift in patients going to see their physician. We saw 34% lyft in screenings so patients going to get screened for the condition and a 24% lift.

[05:25]
in brand conversions, we were measuring the impact quarter of a quarter from the previous data provider, so they had run with a competitor the last quarter and then they switched over to Swoop data for this specific case study, so to be really clear you know Swoop in this scenario outperformed the previous data provider 20% in visitation rates 34% in screenings in 24% in brand conversions, so overall really really strong results very

[06:01]
Now let’s segue over switch gears over to the new product launches. You know you mentioned predictive targeting AI at the top of the conversation. What is that? How does that differ from traditional targeting methods so all targeting up until now is really been based on what has happened in the patient’s history right. You were prescribed. You were diagnosed this event happened and you know this is really what everyone’s doing in the targeting space today. What AI allows for is this opportunity to target? What will happen right so essentially what we’re doing here is targeting the future Mark so advertisers now can identify and target individuals who are maybe at risk of developing a condition pre-diagnosed those who might be transitioning to a new treatment or a patient at risk of becoming.

[07:01]
On adherent within a very short window we also look at targeting patients based on their insurance coverage or getting ahead of progression right. When is a patient going to progress this opens the door to much more sophistication and really comprehensive approach to targeting and I’ll tell you right now. You know this is not something as a company that is new for us right so we have a boutique advanced Analytics District company IPM AI now what we’re doing is we’re automating the technology the high end Analytics that they’ve been doing for years and we’re bringing it to media targeting. So this isn’t our first rodeo here.

[07:49]
Yeah, I know ipm.ai and so you’re kind of leveraging their heritage in this area and then leveraging it for Media targeting a very very interesting. Can you explain how this predictive targeting can be used in a therapeutic marketing context Kitty yeah sure so marketers today can use predictive models to manage marketing resources more efficiently enhance engagement and education to optimize marketing budgets and increase Revenue and what predictive AI can lead to is higher conversion rates and enable brands to engage audiences at critical junctures in the diagnosis and treatment journey anything from pre-diagnosis to treatment selection, so they can really start living with the condition. So what you’re saying is that you can apply this predictive AI technology anywhere along the treatment journey is you’re saying earlier. You can apply it.

[08:49]
When your data suggest a patient is at risk of becoming not inherent meaning they’re already a patient on a treatment and you can sort of intervene there or a different points along that patient journey this can come into play and it’s theoretical to some extent but it sounds like it’s practical in that you’re already using it along that journey. Yes, and you know listen with our predictive AI we’re really solving for a lot of challenges that marketers are faced with today right, so there’s actually five different challenges that we focused on one being disease state awareness with undiagnosed segments one being awareness of product with our rapid adaptor, so we’re looking for those patients who are likely going to switch and those providers that are rapid adapters the third area is understanding coverage, so who’s likely to be covered and the providers with patients with access.

[09:50]
The fourth is competitive and complex journey with progression so patients who are likely to progress on their treatment journey and the providers who are seeing those patients and then last but not least the staying on treatment, so not adherence is a huge problem in the health industry, and we’re able to identify patients who are non adherent who are likely going to be non adherent within a very short time period and those Physicians that are treating those patients who are likely going to be non adherent. Yeah, I mean that we’ve been hearing about that promise for quite a while now, so it’s really interesting to hear that really coming into play. Can you share a bit more about how this predictive AI targeting helps marketers? You know and maybe share some examples of these audiences and their effectiveness sure so let’s talk about non adherence because non adherence is one of the first segments that we brought to market a few months back.

[10:50]
This is obviously a really huge problem in the industry within Healthcare we see over 50% of patients with chronic conditions become non adherent during their treatment Journey right. This is a 500 million dollar on average cost that the health care system occurs in a given year so what we’re able to do here is identify within a 30-day window which patients are likely going to be non adherent to a drug to a treatment to a therapy and these numbers Mark are really impressive so we ran numbers for three different types of drugs treatments and therapies right so we looked at a type 2 diabetes drug. I can’t name the name of the drug but we saw 94% of all patients. Who’d become non adherent within the 30-day window right, so now this allows.

[11:50]
Tizers to say hey I want to become more efficient with every dollar that I spend in market today. Not only am I going to target those patients based on what has happened. So who is on script, but I’m also going to heavy up my media dollars against those people who are likely going to fall off script in 30 days. What if you could talk to those patients about the reasons why they’re likely going to be non adherent and have the ability to change their mind right so we also looked at an MS treatment and we saw 92% of all patients who became non adherent within a 30-day window so great accuracy there. We looked at it depression drug. We saw 92% of all patients who became non adherent within that 30-day window and this is something you know if an advertisers interested in exploring they can reach out to someone at Swoop myself.

[12:50]
Loop wrap and we can run these numbers based on their specific drug of interest and we can let you know before pre-launch the accuracy the number of patients that were able to predict within that 30 day window. It’s really something that is what I think is new and innovative in the market today. It’s really fascinating you mentioned. You know what if you could you know identify all patients and given therapeutic area with a specific drug or likely to become non-inherited a 30-day window That Could allow the advertisers to become more efficient kind of like the classic John Wanamaker problem. I know certain percentage of my ad budget is inefficient, but I just don’t know which you know part to focus it on and reach out to them about changing their minds that that’s really interesting so that’s the kind of messaging that you would then deploy here there’s at the next step correct correct. So you could heavy up on your you know your Media dollars against those patients who are likely going to be non adherent and then Mark so another example.

[13:50]
You could look at is undiagnosed right. What if I could tell you six months in advance which patients were likely going to become diagnosed with specific condition and what if as a brand marketer you could speak to those patients before they become diagnosed right. So you know I think oftentimes myself included you think that undiagnosed patients or typically rare disease or maybe they’re let’s say misdiagnosed, but what we found is you know you look at just type 2 diabetes alone. There are three million people in the US who are undiagnosed with type 2 diabetes.

[14:31]
That’s a lot of patience and I think identifying those high value patients and their Healthcare providers and speaking with them can really Accelerate the time to diagnose this and treatment right and so you know in when we’ve done this. What we’ve seen the results have been particularly impactful, so we recently did a campaign with an RA drug and we looked at five million patients. We were able to predict 42,000 patients six months before they were diagnosed with ra, so if you were a marketer and you were able to reach those patients six months before diagnosis in this scenario the results showed as 10 to 16% lift in a very strong overall lift in lifetime value of those patients.

[15:28]
So being able to target higher in the funnel increases the likelihood that they’ll end up on the brand’s treatment sooner and it’s really you know optimizing patient outcomes in an ultimately driving strip left. Yeah, I mean this is the stuff that you hear you know chief marketing officers at Pharma companies talk about leveraging claims data with AI to then identify people that are potentially candidates for a therapy especially in the rare disease area where it’s hard to target and identify potential patients and so the fact that this has becoming more accessible to marketers on a serial basis is really fascinating rather than a kind of a bespoke you know let’s build it from the Ground Up basis. It sounds like you’ve got a process. That’s not such a heavy lift for brand managers out there. Let’s just switch gears a little bit and talk about the other new offering of Swoop earlier this year you launched a conversational AI solution tell us about that.

[16:28]
As head is that integrate into the existing targeting business. Yeah, so Mark listen. This is hands down game changer okay so 2023 farmer advertisers spent 7.2 billion dollars to primarily drive their patients in hcps to a brand.com which tend to be you know a lot of times. They’re difficult to navigate sometimes. They leave patients and providers searching for answers the results are sub optimal engagement and this assets specifically as an 80% bounce rate. So you’re spending 7.2 billion dollars you’re doing an awesome job on your targeting and finding and driving but you’re sending them to an end point where there’s an 80% bounce rate. This is leading to billions of dollars of wasted advertising so what conversational AI does I like to think.

[17:28]
Of it as an mlr compliant virtual customer engagement solution that lives on your brand.com whether that be DTC or hcp or both the brains behind this is so Sophisticated that there’s really nothing else in market that Compares and when you think about what marketers are doing today Mark did you know the first website was developed in 1991? I believe it. You know it’s funny. There’s really been you know very little change to websites themselves and you know we’re driving our patients and our healthcare professionals to a 35 year old piece of technology in fact. We had a top five firmer. We spoke with recently they spent over 20 million dollars creating an hcp.com.

[18:19]
They understood the problem. They just couldn’t quite nail the solution right and in a little test we recently ran. We looked at trying to find information on a.com and it took us 14 clicks to get to the single piece of information that we were looking for you know listen modern brand websites were never designed to operate as primary engagement tools and we have to start thinking about things differently Swoop conversational AI provides users with an intuitive experience that responds to natural language it continually adapts itself learns. It’s contextual relevant and it’s a human-like dialogue. We spend the last six to 12 months working on the brains of this to ensure that it was sophisticated and it was going to exceed any type of expect.

[19:19]
In the market, it’s not our first rodeo with this either. So you know this company’s been around for a while. We have worked with the top 15 pharmaceutical manufacturers in the US and internationally we just made the solution a heck of a lot better in the last six to twelve months, so when we look at kind of what this solves for it increases high value activity on brand.com it provides near human mlr compliant conversations and it delivers real world data and Analytics back to the manufacturer based on conversations, so start thinking about the data exhaust that you can get from understanding not just who your patient market is but what they’re actually asking about. What are the conversations that they’re having and how can that data exhaust fuel your overall targeting initiatives The Who the what the wear the

[20:19]
Will you talk to your patients in hcps? I think you know even just the data alone is going to be valuable to any brand marketer today sounds like yeah, you just kind of laid out the benefits that marketers gain from that approach. What results should a brand expect to an implementing conversational AI to their site. We’ve seen.

[20:40]
H is engaged with the agent over five minutes per time right so in this you know you can compare this to say what you your field teams experience which is probably around 60 to 90 seconds. We see 800 percent boost in on-site engagement such as things like let’s say request a Rep or ask for sample or download content or registrations for marketing programs right. We’re also seeing 97% of users trust the information that they’re receiving 96% of increase or successfully managed. We see. Oh this one’s really interesting so we’re seeing over 50% of conversations are taking place in the evenings right after your sales force is off the clock in satisfaction rates are very strong. We see between 4.7 to 5 on user satisfaction for conversational AI it’s no longer a novelty but it’s a

[21:40]
For success and I think at Swoop we take this to the next level with supplying back that data exhaust to our advertisers and the value of that is really priceless. Yeah, that sounds like it really is these conversational AI based customer service chat bots you know on these hcp sites really turning around some good performance numbers there finally you kind of wrap this up Katie what do you believe that predictive AI signifies for the future of the Healthcare industry first off. I think brand marketers need to think about both their targeting and they’re engagement as holistically and not siloed you can have a suboptimal site, but really great targeting and you’re not going to get the outcome that you want right. You’ve got all the right patients and Physicians going to a landing experience that bounces 80% of the time or vice versa you could have a really great.

[22:40]
Engagement tool but you not reaching the right patients right and you’re not reaching the right position so I think both in totality. They can’t work without each other right for if you want to get the outcome that you want but to really answer your question. I think simply predictive AI is going to help Healthcare providers and patients make Better Decisions and ultimately it’s going to improve health outcomes in Lower Healthcare costs. Overall absolutely sounds like the promises there this has been really a fascinating discussion. Thank you Katie for talking about these Innovations that are now available for brand marketers both the predictive AI and the conversational AI tools which really have the potential to be game changers for health media targeting and as you say increase or improve outcomes lower costs because it’s making potentially marketing more efficient and improved quality because

[23:40]
getting the right drug to the right patient at the right time which is the goal so thank you so much Katie

[23:47]
Yeah, much Marc it’s always great speaking with you likewise to that end those who have any questions can email Katie with their Media targeting queries or contact her through Eminem and you know that was a wonderful conversation. Let’s have another one. Thank you everybody for listening come back soon. This has been markusquitz for the mmm. Podcast take care.