15 min read

Why Marketers Must Make the Switch to Privacy-Safe Real World Data for Targeting

Featured Image

By: Emilie Branch, Content Manager, Swoop/IPM.ai

Scott Rines, Swoop’s Chief Revenue Officer and Lauren Jacobson, Group Director, Performance Acceleration at Initiative, discuss how to successfully reach a targeted audience using real world data. 

The pharmaceutical marketing industry is at a crossroads. Legacy targeting methods including demographic and clickstream data have been depreciated by recent privacy-driven announcements from Apple, Google, and Facebook. The inability to track patients online has created a barrier between brands and their ideal audiences, making it more difficult to deliver a message that might empower patients to take charge of their health journey. However, despite a whirlwind of seemingly negative changes, an opportunity to target more effectively has emerged. 

As Scott explains, Swoop has an extensive privacy-safe database of health records for more than 300 million de-identified patients going back a decade. ML/AI is used to create custom audience segments for clients based on this real world data (RWD) universe; each segment is completely tailored for each brand and can include a range of criteria. 

This level of specificity is necessary because simply targeting a patient with lung cancer, for instance, falls short. Instead, agencies must go deeper and understand the audience’s exact diagnosis by ICD-10 or ICD-9 code. “The more information, the better, because that gives companies like Swoop all that they need to build the perfect audience,”  Lauren shares. Ultimately, patients are highly informed, “always on” technologically and act as the CEOs of their own care – so if a message doesn’t resonate, they won’t convert. 

Although niche patient groups are being targeted based on sensitive health data, it’s possible to maintain privacy-safety. To avoid potential risk, marketers must work with a company that is accredited by the Network Advertising Initiative (NAI). The NAI is a self-regulated industry trade consortium with strict online data protection guidelines created by industry for industry.

Additionally, Swoop’s segments go through a process called K-anonymity, a scientific way to guarantee patients cannot be re-identified. To achieve this, the audience quality (AQ) score of any segment should not be greater than 50%. This ensures patient protection and is still much denser than anything achievable with “spray and pray” targeting. For instance, while only 2% of the population may have a condition, Swoop can build an audience with a 30-50% likelihood of a diagnosis. Not only does this maximize budgets, but it also enables more relevant messaging to reach the right people.

These audiences can be activated across channels including linear TV, CTV, digital and radio, allowing marketers to plan better and optimize spend based on performance. “We can't get to measurement without targeting the right people,” says Lauren, who adds that marketers must ensure that cookies (or any other clickstream targeting method) “are not a piece of the measurement puzzle.”

Aside from targeting the right patients, reaching HCPs is critical – especially considering current challenges with personal promotion. An added benefit of building custom audiences for specific patients is that it’s also possible to uncover their treating physicians. By giving marketers access to both key audiences, Swoop enables the creation of a full end-to-end campaign, all but guaranteeing script lift. 

Read on for the full conversation moderated by Peter Kane, Swoop’s Director of Growth Marketing.

Peter Kane (PK): Scott, can you provide a brief overview of Swoop and its place in the market? 

Scott Rines (SR): It’s a critical time in healthcare. There are a lot of changes taking place not only on the targeting side but also with how data is collected. At the same time, access to HCPs is very limited. However, digital really can play a part in providing valuable information to patients if we take the right path and become great stewards in the industry. Swoop has a 300 million de-identified patient database that has a 10-year lookback, and with that we create custom audiences for our clients, whether that is the agency or the manufacturer. Every single audience that we create is based on those 300 million patient IDs. We also take in over 65 billion social determinant factors when we create these audiences – all of which are specific to a campaign and never used again.

Lauren Jacobson (LJ): I'm Lauren and I work at Initiative, which is one of the leading ad agencies in the world and part of Interpublic Media Group. Initiative’s clients range from toy to gaming companies, to travel, but the main clients that we have are pharmaceutical.

I’m very familiar with a lot of the targeting that we have been able to do in the past and what we are no longer able to do. Going into 2023, the evolution of targeting is to get closer to the right person instead of taking a spray and pray approach. This has its place in media but when it comes to digital, you want to reduce all that waste and get in front of the right person. 

PK: Scott, what does the current healthcare marketplace look like? Can you elaborate on some of these changes and how they're ultimately going to affect marketers’ ability to operate?

SR: I think a lot of these are in the marketplace that everybody's aware of, like iOS 14. If you have a mobile phone and you've been asked, “do you want to be tracked?” Of course, almost 90% of people say no. Facebook took a big hit on earnings because they can no longer track other sites when you have your Facebook app open and they are now moving away from interest targeting. At the same time as data collection is changing, 15 pending government bills are trying to clamp down on how patients are targeted.

So those are the two things that are sort of fighting each other – the collection of data and how we target, and then also what and how you can target. And like I said, 15 bills are going through state senates to prevent that. That is a real issue because it is preventing us from getting the right information to the patient who may have a disease that there’s a cure for.

These patients might be unaware of that because they’re getting limited time with their physicians. If the ability to target these people goes away, it will be a real challenge in the marketplace.

PK: Lauren, are you seeing similar challenges and are there any others that you'd add and, more specifically, how are you addressing this with clients? 

LJ: The majority of my career has been in pharma. Over the past couple of years, as new bills get passed by Congress and as state regulations change, people are always focusing on owning your first-party data.

If you're a car company or if you're selling cruises or even anything that you could buy from Target or Walmart, it's very easy to own your first-party data. However, pharma is more challenging than that. We can't ask our consumers that haven't even been diagnosed, what's coming tomorrow; to sign up so I can message you as a marketer – we have to be crafty. The challenge is also remaining privacy-safe. We have to think about all the rules and regulations and come together as a pharma healthcare industry to get a message in front of a person who might be diagnosed tomorrow or next week or next year, and their family members or caregivers. 

PK: Scott, let's go back and say there are challenges engaging with HCPs, but that also means there's this unique opportunity to better engage with patients. At the same time, there are competing forces at play; patients want more access to health-related information, but it's countered by this highly fluid privacy environment. How should marketers go about identifying partners and targeting strategies that they can be sure are compliant and mitigate risk while also still being effective? 

SR: Most people think if a company is HIPAA compliant, they're safe to work with. HIPAA compliance is about the collection of data — that's the entry point of the game – everybody should be HIPAA compliant. What it really depends on is, are you an NAI member? If so, you’re compliant with not just how you collect your data, but how you use it to target.

It's important for anybody that's either on the buying side or on the technology side to do everything possible to become compliant with the NAI. For those who might not necessarily be fully familiar, that's the Network Advertising Initiative. It's a collective of technology providers and others who comply with a set of rules and then are audited regularly.

PK: Lauren, from your perspective, how do you determine who you partner with and what level of importance do you place on NAI membership and compliance? Are you seeing clients coming to you with questions on privacy changes or is that something you have to bring up with them? 

LJ: When Facebook is in the news for a change that revolves around targeting, clients are on top of this and come to us. We have to make sure that we're in lockstep with Facebook's policies so we can have the right answer. Facebook is a big portion of dollars. Honestly, it's a broad reaching platform and it's a great way to reach a lot of people at once. 

Anytime you're bringing healthcare information with ID-based information and meshing the two together, you need to think of the consumer first and make sure that everything is privacy-safe. When companies belong to the NAI, I think that it makes the conversation a lot easier to have with clients.

PK: Following up a little bit deeper into the NAI discussion, Scott, their regulations are appropriate for both sensitive and non-sensitive conditions, which I think is important when it comes to being able to make sure you're targeting in a privacy-safe manner.

SR: That's correct. The NAI is trying to strike a balance of protecting the patient, but also being able to find the right patient. You want to find the right patient because we have valuable information to provide them about potentially life-saving cures and drugs, but you also want to watch out with sensitive conditions and when you're targeting, and where they receive that message.

PK: Getting more into the practicality of this targeting methodology, Scott, you mentioned that Swoop has a 10-year lookback period of 300 million de-identified patient journeys and 65 billion anonymized social determinants of health signals. For those that might not be familiar, can you provide an overview of these real world data based custom audiences? What separates them from other targeting strategies now and potentially in the future? 

SR: We take a safe approach to how we build these audiences. We start at a point of ignorance — we don’t know anything about a patient other than the potential demographic audience. Then we match this up to our 300 million patient-level database that's tokenized and we see what the index of these consumer groups against those with a specific condition is. When we push an audience to a DSP, it is a representative audience. We are approved for sensitive and for non-sensitive conditions with every DSP in the marketplace because of how uniquely we build these segments.

PK: What are some examples of what these audiences might look like, whether that's diagnosed, diagnosed with a comorbidity or some examples? How can these audiences be applied across channels and what's the value of that? 

SR: They can be as custom as the marketer wants them to be. Take diabetes, a patient could have failed a first-line therapy or been on a specific drug; it can get as specific as you want. The more specific you get creates changes to the size and scope of the audience.

These audiences can be applied not only from a linear TV standpoint and a planning perspective, but also from CTV, digital and radio. For any communications that you make in the marketplace, you can layer on these audiences and know that at a high level, they have these conditions and are HIPAA compliant. 

PK: Having one audience unified across your activation really helps nail down reach and frequency so you can plan better and optimize the media mix.

SR: Yes, by covering the likelihood a consumer has a condition. In the average population, if 2% of the population has the condition and you can build an audience that has a 30 or 40% chance of having it, you're much more efficient in your spend and you're also giving a relevant message to a patient. The worst thing you can do is give a message to a patient that doesn't have the condition — that's bad messaging. 

PK: Lauren, in your work with clients, how are you using these types of audiences and what outcomes are you seeing? 

LJ: When we get a new brief or we're thinking about a new drug that's coming to market, we work with the client very specifically to understand what are the ICD-10 and ICD-9 codes that a doctor would prescribe to know a patient has a diagnosis – and the list can be long. It can be 10, 20, or even 50 things. If the drug is indicated for a lot of conditions and then we get that list and we share it over with partners like Swoop to say, okay, here are the very specific nuanced audiences that we are looking to find in the real world, put together an audience for us using the methodology that Scott just walked through. Then they can send it over to any DSP we're looking to run with any connected TV partner, etc. It goes on and on. Actually, just last week, we ran a Swoop audience targeting for a digital ad.

We were running that same audience using mobile phones and devices, but also as these people are walking down the street, they have the potential to see our ad on a digital out-of-home board. It’s critical to layer in similar audiences to a lot of different media because you can manage reach and frequency and start thinking about a campaign more holistically versus, this is your digital element.

PK: Lauren, taking another step down our journey and digging into measurement – when you look at how it's currently done and connect it back to some of the changes we've referenced earlier, and the changes that are likely to come down the line, for example, with the deprecation of cookies, what are some future considerations?

LJ: We can't get to measurement without targeting the right people. It's critical for every single marketer to take a look at how you're targeting consumers and make sure that cookies are not a part of that puzzle at all.

I think the best targeting moving forward will be ID-based content. I think that the key pharmaceutical closed loop measurement partners are not the ones that will be the strongest tomorrow and for years to come. There is a company that will rise up that will have a better solution because the tech needs to evolve. The foundation needs to evolve. We need to start thinking of things differently and thinking of more ID-based privacy, safe ID-based opt-in, ID-based solutions that help that closed loop measurement. 

PK: Scott, taking that “think differently” approach, I wanted to talk more about audience quality because you referenced it earlier, saying that 2% of the population might have a condition, but being able to build an audience where 30 to 40% of that audience has the condition is a lot better from a targeting perspective. Can you talk about audience quality and what it’s exactly looking at? How does it connect to script lift and what should marketers keep in mind when it comes to privacy? 

SR: I would say that audience quality (AQ) is the gold standard in the pharma marketplace. It's saying that this person has, or more likely has, the condition and a lot of that is also based on how high you can build it, so you prevent reidentifying the patient. Could you build an audience over 50% AQ? Yes, but not if you're a steward in the industry and doing the right things for patient advocacy. You think about leading indicators because you might not receive script lift for six to eight months down the road. So, what is the right metric to look at and say, my dollars are going in the right places when you don't have script lift yet?

With audience quality, you're spending your dollars more wisely because you're getting your message in front of people that are more likely to have the condition.

LJ: Now all of our clients across a slew of different pharmaceutical companies speak in AQ. This is a known leading indicator before you get all the script information. There's such a correlation between getting in front of the right consumer and people getting your script and converting on ROI.

PK: Is Audience Quality capped at 50% because it prevents the re-identification of a patient – for example in an audience of one?

SR: Yes, we do not re-identify the patient. There are marketing strategies based on zip nine targeting and the challenge with that is you can start reidentifying patients, especially with sensitive conditions. 

PK: Lauren, there's clearly a lot going on in the market. Healthcare marketers have a distinct opportunity to shift how they've been operating, if they haven’t already. How do you go about educating clients and what could pharma marketers start considering when they start analyzing their own programs?

LJ: It's hard for pharma companies to build first-party databases but having your own data will always be preferred. Understanding the technology changes that might be occurring and keeping ahead of that is key. You have to push your agency and be a steward of everything that's going on in the space. Take inventory and then think differently. Think of ways that you can target consumers and think of companies that are leaning into ID-based solutions, whether the foundations are Epsilon, Axiom or Live Ramp.

Opt-in privacy-safe platforms are huge and will be around tomorrow. If companies are reliant on that, that's a good sign that if people do their homework, things will change.

Even if you have a handle on it today, you can't just sit back – you have to keep up with all of the changes. While you’re creating a yearly plan, you also have to think about your alternative: if X, Y, and Z aren't working anymore, what else can you do? 

PK: Is there anything you'd add to why using this type of privacy-safe, real world data is so important because it mitigates the risk associated with a fluid market?

SR: Yes, the changes that are happening today are happening almost daily. iOS 14, can't collect data. Facebook is making major changes, we’re soon entering a cookieless world.

As changes happen, if you're collecting data online, you're going to lose your signal. The way to think about this is, are you working with data providers that have offline data that will not be affected by any of these changes and are they HIPAA compliant? Are they members of the NAI? Most of the folks that have real world data check all those boxes, so it's pushing everything in that direction. 

PK: Is there anything else that you would like to add before we head into the Q&A?

LJ: I do think that this is an exciting time to be a pharma marketer. I think that data is going to be cleaner than ever before and there might be some changes to traditional measurement, but that's not a bad thing. In the past, not everything was as transparent as possible, so maybe numbers were skewed incorrectly. I think the future is promising and the companies that are doing it correctly will rise to the top. It'll be very clear because they are members of the NAI and are very upfront. Pharma marketers should look for that in the industry because those are the companies that you want to align with. 

SR: I think it's a great time to be in this industry. We can solve a lot of problems. As doctors are spending less time with their patients, we can provide a service as long as we do it in a way that puts the patient first, doesn’t re-identify them and adds value. 

PK: The first audience question is, how easy is it to get started with custom audiences? Do they cost money to build, how long do they take? Are there any channel limitations, etc.? 

SR: The turnaround time, depending on what vendor you use, is anywhere from two to eight weeks. Swoop is typically around the four to six-week mark. As long as you understand the definition of who you're trying to reach, that’s how you would get started. You would work with our client service team and they would look at all the diagnostic codes, they would build the audiences and they would then send that to your DSP of choice. We do not charge to build these segments. The only charge is activation when you pay on a CPM through your DSP of choice.

PK: To clarify, these can be activated in the channel of your choosing, whether that's digital, radio or TV?

SR: Yes, we're also approved at Nielsen for TV planning for DDL, so they also can be used as a signal to understand the best shows to purchase on your campaign and then you can match back to the Nielson panel to see how well you did. Everything from radio to connected TV, to addressable TV, to linear to digital –  think about your marketing budget as it is connected to one audience.

There's also the ability to say, here are the patients I'm targeting with the condition. I now know the doctors that are treating these patients and I can create a specific campaign against those doctors. So, when the patient goes to the HCP they’re on the same page. You're bringing the communication with both groups together, which is powerful. 

PK: Lauren, how should pharma marketers start conversations with their agency around audience building and improved targeting?

LJ: We have to understand what it means for a patient to be indicated, or what they are diagnosed with using those codes. It's no longer good enough to target a diagnosed lung cancer patient as that could mean a thousand different things; what kind of diagnosis is it, metastatic, adjuvant, etc.? The more information, the better, because that gives companies like Swoop all that they need to build the perfect audience. 

Normally when agencies have their planning tools and you try to figure out how much budget you need to reach an audience in years past, we were just talking about 18+, 35+, or 18-34, for example. However, now, using health-based audiences in those planning tools can help make a stronger case for a marketer at a pharmaceutical company to say, “If you want me to reach 80% of my target, I actually need X dollars to reach them this many times a month or a quarter. I'm no longer talking about 18-34. I'm talking about a diagnosed patient.”

That same audience that I planned against, I can activate them in seven different channels and monitor it holistically and it’s all privacy-safe. 

PK: Scott, can first-party data be incorporated into custom audiences?

SR: Yes, we can digest first-party data in many different ways, typically through Datavant. We can pull in first-party data from a manufacturer and create a HIPAA-compliant audience that can be pushed out to the DSP. 

PK: How can this technology and data be applied to HCP marketing given the challenges with personal promotion? 

SR: If you think about all the ATP marketing or NPI targeting, they're looking at an NPI list that is used for doctor detailing. Most of the manufacturers are not creating specific NPI lists for advertising. Now we can connect patients and doctors, meaning if we build a custom audience for a specific patient type, we can look at that audience and see what doctors are treating these patients.

The challenge is that a lot of the HCP and DTC marketing departments at the manufacturer don't even talk to each other. This omnichannel strategy can come together but there is a lack of strong communication – they're almost siloed at the manufacturer. Eventually, I think this will be a very common way to go to market. Essentially, you go beyond omnichannel from the channel perspective, to who is being targeted on either side – it’s an echo chamber of the patient and their doctors. This creates a full end-to-end type of campaign. 

LJ: If you think about the diabetes marketplace, there are so many ads that targeting a patient diagnosed with type 2 diabetes isn't going to be enough. You're not going to convert in that way, you need to be more specific and come to the table with a holistic solution.  

PK: Scott and Lauren, thank you again, both for being here and having this conversation.

About the Author

Emilie Branch 

Content Manager

Emilie_Original-01

 

Emilie has nearly a decade of experience in marketing, writing, research and strategy for the pharmaceutical industry. She has contributed to top pharmaceutical publications including American Pharmaceutical Review, European Pharmaceutical Review, Pharmaceutical Manufacturing, Manufacturing Chemist, Pharmaceutical Outsourcing, Specialty Chemicals, and Contract Pharma. Prior to joining Swoop/IPM.ai, Emilie served as Strategic Content Manager and Managing Editor for Pharma’s Almanac where she lent her voice to some of the industry’s top players including GSK, MilliporeSigma and ThermoFisher Scientific. 




 

 

 

 

 

 



How Data Driven Linear Unifies TV Planning, Activation and Measurement

By: Kurt Robinson, Senior Vice President of Business Development, Swoop

Data-driven linear (DDL) represents a tremendous opportunity for health...

Read More

Tune Into Condition-Specific Patients – Why Health Marketers Need to Include Broadcast Radio in the Media Mix

By: Emilie Branch, Content Manager, Swoop/IPM.ai

Pharma investment in AM/FM radio has steadily increased year over year; according to Miller Kaplan,...

Read More