The Honest Programmatic Conversation Healthcare Marketers Need

Announcer: Welcome to the Ignite Podcast, the only healthcare marketing podcast that digs into the digital strategies and tactics that help you accelerate growth. Each week, Cardinals experts explore innovative ways to build your digital presence and attract more patients. Buckle up for another episode of Ignite.

Evan: Hello, everyone. Very excited to be here. Very excited for all of you to be joining us. My name is Evan. I’m the VP of Paid Media here at Cardinal, and this is–

Lisa: Lisa Fisher. I’m an associate media director specializing in programmatic and paid social.

Evan: We work hand-in-hand specifically on wider strategy, and forecasting for all media channels, including programmatic. Very excited to be talking about really the honest conversation healthcare marketers need to have and understand about programmatic in the current industry.

Today, we want to talk about four key areas that we want to explore. Programmatic and healthcare, specifically, the details related to programmatic that is important for healthcare marketers to understand, the real reality of programmatic, our readiness framework. How should you go about doing this, and then how and when to test. Really want to jump through. It’s an overall general overview of programmatic.

It’s exciting because programmatic, I would say, is the most confounding, least understood channel in digital marketing. I think if you really know what you’re doing, you can really wield it powerfully. Let’s start with the issues and challenges to address. Lisa?

Lisa: Yes. When running programmatic media, you are often relying on a third-party vendor who is managing the programmatic inventory for you. They’re going to be the ones who are helping you determine the campaign structure, determining which channels to run on. We need to be the stewards in guiding them, and telling them how to set up our campaigns so that we’re making sure that our ads are getting to the right audiences.

Programmatic can run across a variety of different platforms and in a variety of different mediums. You really need to know when you want to consider programmatic versus another channel. One of the four pieces of any campaign that Cardinal would run is search marketing. We utilize search mainly for driving leads. When you are getting to a plateau or you feel like you’re not reaching net new audiences, you need more prospecting, you need more engaging video content, that’s when you would consider running programmatic.

If search performance is hitting a ceiling, that’s really the number one moment when you want to say, “We need to open our portfolio and be reaching people through programmatic channels.” Also, you know that you need to build awareness, but how does that get quantified? How do we measure the success of that awareness that we’re running on programmatic? We’re going to discuss that.

Programmatic, as I mentioned, we’re running it through other channel DSPs, so we need to have visibility into where our ads are actually showing up, because it can be a black box similar to Meta or any other environment. We need to know what to look for to measure that success. Healthcare then adds extra complexity. We’ve got HIPAA, we have privacy needs in order to protect proprietary data relating to individuals. We talk about how we tackle that in the programmatic ecosystem.

Also, the vendor landscape is noisy. There are so many DSPs out there that are all vying for your competition. We’re going to show you what to look for and what not to look for when you’re choosing your programmatic platform. Then reaching the right audience is hard, particularly in the healthcare space. You really want to find ways to hone in on specific audiences based on a diagnosis, or a certain procedure that they’re looking to have. That is something that we have really found the right mixture of partners in order to target the right audiences in programmatic.

Evan: Yes, that’s right. I think the main summary to take away from this at the beginning is just that it’s not as simple. You should not really just be running programmatic and saying, “I got a bunch of impressions.” There’s so much more that we’re going to cover that goes into understanding how and why you would do this, because usually when we talk to brands or other marketers and they’re just driving impressions, it’s not, so what is the question. The details matter here.

Lisa: Adding value to the impression.

Evan: That’s right. Before we dive into more holistic concepts about programmatic that are important, let’s talk about healthcare-specific concepts. Tracking compliance is a huge category here. There’s two major things we want to talk about. The first are pixels. We know pixels. That’s a common thing. It’s Google Ads, it’s Meta. It’s any paid media platform is going to offer those for tracking.

They are in programmatic really essential for ongoing impact measurement. Particularly prior to any super advanced media mix modeling and multi-touch attribution measurement, pixels are really the only way to see estimated results. The main reason is because so much of what happens on programmatic is really not click-based. You’re essentially showing a billboard to people. If they’re not clicking on it, you don’t have any tracking. Without that, without the pixel, the platform’s not going to be able to see if someone ends up converting on your website sometime later. Pixels are important here.

They can, however, be a compliance concern. Tracking of users, tacking of pages with specific medical conditions, and sharing that information with the platform can be a problem. We’ve seen companies get in trouble for really wayward versions of that, but it’s something to consider because we want to be very cautious about identifying information about users and their site visits.

Finally, ultimately, we would say that the use of pixels and compliance is up to the business and a company’s legal teams. It’s all their comfort and their posture on how they want to approach the use of pixels. Some companies are okay with it, as long as we’re using proper vendors and approaching it in the right way. Others really want to have an extra layer of caution.

That extra layer of caution can come through the form of a BAA, which is a business associate agreement. It’s basically just a contract that helps ensure a vendor or a partner will safely handle data and protect personal healthcare information, PHI. That’s super important. We want to make sure that the platforms, the vendors we’re using are correctly safeguarding that stuff.

The good news is that if you’re able to sign BAAs with different vendors or platforms, it can unlock that confidence for different marketing opportunities. Pixel deployment right away is one thing that you can feel more confident and comfortable about, as well as more retargeting strategies. You’re not as concerned as you were before, plus, of course, knowing that they’re safely handling your information.

Certain vendors are willing to sign these BAAs. We’ll talk about those in a second, but as a note, the big ones don’t. Google and Meta, they are not signing these things with you. That’s its own separate conversation. As it relates to programmatic, we can look at the next slide and talk about some healthcare DSPs here.

Like we said, compliance, what these DSPs can offer, ones that are specifically created for the healthcare industry. They offer expanded brand safety tools, and the ability to sign BAAs which unlock the pixel use. We have a few different DSPs here, Illumin, AdTheorent, which has been now acquired by Cadent, and then DeepIntent. These are all just examples that you might explore of healthcare-specific and focused DSPs that have these tool sets and offerings for healthcare marketers.

Audiences, another big thing is that they will always have a dedicated selection of healthcare third-party audiences, more so than traditional DSPs. There’s a lot more granularity and options there. Finally, they have different inventory and tools; fully vetted publisher lists, they have unique inventory placements like in-office, in-doctor’s office placements on those pads you sign in on and different things like that, and then measurement tools. An example would be prescription lift, which is a crazy metric, but very cool. These options are there, and we usually recommend if you’re in healthcare, you should go for a healthcare programmatic partner.

Finally, specifically with healthcare, advanced healthcare audiences. We talked about the DSPs. They’re going to offer a lot of these audiences more advanced than typical DSPs, non-healthcare DSPs, but there’s other options to take your targeting even further. We have a few options here as well; Definitive Healthcare, PurpleLab, and Swoop.

These are all audience and healthcare technology vendors that help really provide that extra push to even further granularity. More options to target healthcare providers, deterministic options, so you’re literally targeting the people that have these certain conditions or characteristics. Claims-based audiences based on insurance utilization. For being diagnosed or prescription-based audiences, you can get that information. Of course, you can use all of that, not only deterministically to actually target those people, but as a better seed audience for lookalikes. You’re really getting advanced with who you’re targeting.

Lisa: This is all hashed data. There is no PII associated with any of this data. It’s coming directly from the third party into the CRM or into the technology platform where it’s being ingested. There is no handling of any PII. It is completely complicit from that perspective.

Evan: Compliant.

Lisa: Compliant. [laughs]

Evan: Last note here is, we invite you to learn more. This topic in particular, third party healthcare audiences we are discussing, Rich Briddock will be covering it at 4:00 PM today, so don’t miss it.

Lisa: All right. We’re going to shift into the reality of programmatic today. There are a lot of claims that programmatic vendors will make and promise to you. A lot of these things can be achieved. It’s just tweaking the lens with which we are viewing these promises. Really just level setting on what is actually possible versus what the claim is. Programmatic vendors will promise that they can reach upper funnel audiences with prospecting in a premium inventory at scale. Also, say they can do precision targeting with third-party data utilizing some of the more standard audience partners, or third-party vendors like BlueKai, or just any of the standard off the block third-party vendors.

They will say that full funnel attribution is something that they’re able to illustrate by connecting your impression directly to a conversion, that you can find efficiencies in your costs by optimizing spend across the different audiences based on performance, and then also cross-channel exposure, which is really maximizing reach. As we were saying earlier, expanding beyond the search audience whose intent is already there, finding people who are net new prospect audiences, and then connecting them from OLV to maybe a display impression, and through to paid social, and then through to search. Connecting all of the channels together.

However, in reality, there are difficulties to a lot of the actual implementation of this type of measurement. For instance, it is hard to measure direct ROI attributed to programmatic platforms. Programmatic is always going to show value in more of an incrementality or impression-based regard, rather than a last-click regard. We’re used to on the search side of things, people with intent go to Google, they search, they click, they convert.

On programmatic, it’s a different user flow, and we need to think about the attribution differently. It is starting out with a bigger screen engagement, a video impression on maybe CTV streaming on Netflix or Hulu. You don’t have the click capability within those platforms. It’s thinking about how you’re going to measure success from those bigger screen engagements, and then ultimately, telling a broader story of how the overall attribution to programmatic is being told, because you can’t directly correlate it with an ROI like you would be able to in paid search.

Another consideration is the inventory where you’re serving your impressions. There are a lot of low quality websites where you will see impressions being served on a daily basis just as a user of the internet. We make sure, as marketers, that we’re not serving on these low quality sites. We want to make sure there are literal made for advertising sites that are just walls of ads. You want to make sure that you’re not showing up in any of those placements. You want to make sure visibility is there, that you’re showing up above the fold, that you have a good type of placement on a quality premium site.

Then again, to my previous point about last click conversion rate, that is something that we’ve had to learn and acclimate to that there is not a lot of last click attribution going to programmatic. We have a lot of workarounds for how to justify and explain what the value of programmatic is without that last click. Oftentimes, last click is going to search or another really direct channel. It’s connecting the pipes, so to speak, so that you can say, “The last click went to search, but programmatic had a touch point here, here, here, and here. That is ultimately what assisted a search conversion that otherwise would not have happened.”

Evan: Just to summarize it, really, a lot of people we see, they’re running search, they’re doing great. They’re running out of room and they need to expand, “Hey, programmatic.” You shouldn’t be using it for conversion. There are some aspects you can try to do some direct converting behavior, but if anyone is trying to shift over to programmatic thinking it’s going to be similar to search, they will be really disappointed. It’s a long game. You’re forming relationships. You’re not measuring direct conversions. It’s a different beast, and you have to be prepared with that in mind.

Lisa: We’ll show you, ultimately, how to measure the success of programmatic. I wanted to start off with some of our best practices just from running several programmatic campaigns across decades at this point. Really, this is the landscape that we’ve identified as being the primary place to start when you’re thinking about approaching programmatic. You really want to start with that upper funnel, CTV, streaming TV, OLV, which is online video, not YouTube because YouTube is Google. It’s anywhere else on the web that you would see an online video. Then audio. There’s also streaming audio. Actually, a lot of programmatic vendors now are able to tap into linear TV as well.

There are a lot of different upper funnel, more branding types of channels that you can run on. Really, for those channels, we look at video and audio completion as the metric for success. Just looking at engaged viewers; who is watching the ad, and then actually completing it when it’s in a skippable environment to show that they’ve had the opportunity to maneuver away from the ad, and they haven’t because they’re interested in it.

Then going further down really in the consideration stage is where we take audiences who may have already been exposed to a CTV ad or an OLV ad, and we teach them more about the brand. These are people who already have some level of understanding about the brand, but we want to really educate them about the unique offerings, and what differentiates this brand from your competitors. That’s where we really rely on the middle of the funnel to drive more engaged website sessions, clicking around on the website, clicking to multiple pages, learning and exploring more.

That’s really within contextual display. It’s display placements within websites that align contextually with your brand. If it’s dental that is the brand, we can align contextually within content that is relating to topics those users would be reading about.

At the bottom of the funnel, we can do retargeting display. Again, to Evan’s point, it’s not going to have the same conversion rates as paid search, but we’re able to retarget visitors who viewed the middle of the funnel, or who engaged with an ad at any point and drive them to a lead form.

Evan: What’s really going to be important is how do we verify our efforts? Again, if we’re saying that direct conversions are going to be a very small portion of that, and even when we’re talking about the bottom of the funnel, maybe re-engagement, retargeting at the bottom of the programmatic funnel, you still may be looking at view through conversions, rather than direct conversions, because, again, people are browsing. It’s not as common with something like healthcare decisions. Maybe t-shirt purchases are easier to get on programmatic, but something nuanced like healthcare decisions are often not captured through last click on display.

Lisa: It’s a longer journey to get there.

Evan: That’s right. If you’re not going to holistically be really expecting this to be a last click conversion machine, then how do we understand what it’s doing for us? We wanted to outline two major areas; the tools, the kind of methods you might use, and then the major metrics that you’re really going to want to look at in terms of measuring the success of your programmatic efforts.

The tools, lift studies; brand lift, awareness lift, conversion lift. There’s different variations, but lift studies can help inform you, “I’m running this inventory, and it does seem to be supporting improved website visits or increased engagement once I’m running it.” That can be informative. Path and journey analysis, taking a look at the behavior of people moving through your designed funnel. How is that functioning?

Frequency evaluations. It’s definitely really important with knowing that this is a conversation you’re having with the user. One frequency is not going to be great. You’re going to show them one ad, one video ad, one display ad, one bottom of the funnel. That’s usually not enough to really influence them. What is the ideal frequency? What are you seeing as the ideal frequency?

Holdout tests are going to be a big one. If you’re launching something new, you can obviously compare between a market that you’re not running programmatic in and another one, or if you’ve been running it, consider removing that coverage and to determine what the difference might be in terms of traffic volume and a bunch of other metrics.

Then the most advanced stuff that we’re not really going to talk about here is multi-touch attribution, medium mix modeling, and impression tracking. This stuff is really advanced modeling, and it takes a lot of brainpower and data organization to squeeze anything out of it. That is going to be your last step.

Metrics, I think the most important stuff is going to be view through and incremental conversions. Not last click, but did I show my ad to someone and then later they came to our website and converted? Website traffic volume, you’d like to see an uptick if you’re influencing more people, you’re increasing that traffic. Engage sessions; are you hitting the right people, and are they spending more time on more pages on your site rather than quick bounces?

Branded search volume; are we seeing an uptick in demand? Is your name getting out there, and therefore more people are going to be searching for you? Then finally, conversion rate efficiency; whether or not certain different channels you’re seeing are getting a lift from that exposure and relationship you’re building.

Lisa: Successful programmatic campaigns will utilize precise audience targeting. We know that programmatic can run broad targeting, and that is the basis that I would say that any awareness campaign should start pretty broad. I would also say that we need to layer in third-party targeting as well for some audience precision and making sure that we’re targeting people from the earlier slide that we mentioned who utilize that third-party data and prescription and claims data.

Patience pays off. We know that we need to be running this test for a significant amount of time, and give it enough time and spend to have statistical results. Multi-channel beats single-channel. Just running one channel in a silo isn’t going to have much impact. When you run it simultaneously with CTV, OLV, display, and multiple touch points, you’re really going to see the attribution come through on the organic side. Then measuring beyond the last click, considering those signals like website traffic and the correlation really in the amount that you’re spending in upper funnel and what that is doing to the lower funnel conversion rate.

One client that we wanted to highlight, we’re spending almost $1 million a quarter, and we saw a 43% lower incremental CPA on this client after we shifted our spend more into CTV, OLV, and display. Also, when we started running more spend in awareness, we saw a 33% increase in inquiries on their website. That is just organic people coming through to their website without necessarily having the CTV or the OLV exposure.

Evan: I’ll add that that’s huge because this particular client was their biggest problem. The rest of our paid media support with search and bottom of the funnel efforts elsewhere was hitting goals and doing really well, but their website was suffering. No matter how well we were doing in direct converting stuff, it was not offsetting the losses in organic. This increase is massive because we shifted our attention up funnel and really drove that engagement.

Lisa: Running those upper funnel ads drove more people, 33% more traffic to their website, and then there was a 30% increase in brand search traffic on Google. We were able to spend more in brand and capture that demand, and then convert it.

This next section is talking about when is programmatic worth testing. We touched on this earlier, but it’s really when you’re looking for scale beyond search. When you’ve reached a point where you feel that search is saturated, maybe you’re already maxed out in your brand search volume, you can’t take any more, and you need to get more prospective audiences engaging with your brand.

Also, when you need to target niche audiences like the third-party audiences that are available through accessing claims data, prescription data. Again, you can target people based on in-person versus virtual visitation. That is when you can plug into programmatic to utilize those audiences. Also, brand really does help drive performance. When you feel that maybe you’re stagnant, or you’re at a point where performance is what it is and you need some element of evergreen or branding support, that is another great place for programmatic to fit in.

How much should you invest? This is a basic rule of thumb, but ultimately, you want to make sure that you have enough spend to warrant or to show significant results. We would say at least $10K per month per channel. If you want to run display banners, $10K per month minimum to get you started. Really start out with testing in the beginning with 10% to 15% of your total media investment for the first one to three months.

Then in the second phase, you’re really taking the learnings from the first phase and optimizing and validating them. You’re taking any patterns that you may have noticed, and you can actually do AV tests to statistically say definitively, “This is actually the winner, this type of creative performs better.” Then this would be another 15% to 20% of your budget within months four to six.

Beyond that, month seven plus, that’s really where we’re looking to scale. This is what we do personally with our own clients. At this point, we’ve tested multiple variables, we know the channel mix, the creative mix that’s working. This is where we can start injecting more money into the areas that are working with that ideal mix and frequency that we discussed.

Some red flags when you’re approached by vendors because sometimes they will approach you or when you approach them, there are some red flags. If they’re not transparent about fees, I think that’s a big one. There needs to be transparency about exactly what they’re charging for, and what comes with that cost that you’re paying? Also, if they’re unable to explain their incrementality method, or explain ultimately what value we’re able to give to an upper funnel impression. Also, if they don’t give you access to log-level data, which is the raw data, that’s a red flag.

If they’re not able to have a BAA, that is something that can potentially be worked around, but it’s nice to have, especially in healthcare. Then pressure to commit to a long contract, again, that’s a red flag and you want to be able to have some flexibility within the contract.

Evan: All right. Last section here is how we approach testing programmatic. Breaking it into a handful of steps. Really, starting small, learning fast. The most important thing is taking some of these things that we’ve discussed to heart, and making sure you have a plan before you do it. Not just, “Let’s run some programmatic and see what happens,” and then try to figure out your measurement framework after. Start with understanding how you’re going to do it and measure it.

You want to start small to control your risk. You have to gather a bunch of insights and really get used to the structure you’re building. Single market or service line. If you’ve got 100 clinics, 10 clinics even, or multiple service lines you’re trying to cover, don’t try to cover everything all at once. You can either cover your brand holistically or specific offerings, but isolate it so you can funnel more data and funds into a narrower funnel to get better results.

Lisa: [unintelligible 00:27:30] your learnings in one.

Evan: That’s right. This is a long game, so three-month commitment or more is really where you want to start. You don’t want to try to measure it for a month or two and then throw in the towel. Like we said, a clear hypothesis about what we’re testing and how we’re going to view the success measurement, and then really reduce the testing variables. Changing the website and conversion actions you’re measuring and other things are going to really cause a problem with your analysis.

Step two would be really testing audiences and not just channels. You really have to do both at once. CTV versus online video, you’re going to want to make sure that you’re covering audiences effectively in both. You want to compare claims data versus contextual placements, so actual audiences versus just the context of where you’re applying your targeting. Test the different channels and sequential messaging. Then just continue to iterate based on performance with that narrow view.

Lisa: Then measuring beyond the platform metrics. Don’t just trust the metrics that are being shown in the platform. You have to validate them against your own data. You can look in GA4 at website visitation for your client. You can look at additional incrementality studies and brand lift studies to quantify that actual data that’s coming through, looking at cross-channel impact analysis.

You can also use call tracking. If you are a practice that has calls that are in place, you can actually track calls and have AI detection of what the context of those calls is to verify that you’re actually getting people booking appointments. Then the final step is scale what works, kill what doesn’t. Given that you’ve had enough time to successfully get statistical data, if it’s not working, just cut it and start investing more in what’s working. Making sure that you’re increasing investment and you’re scaling in the areas that are working, and identifying additional opportunities that are similar to what is working.

Evan: We suggest being ruthless. Know what you’re trying to drive, whether it’s video completes or engagement. If you’re running different channels and one is better than the other, be ruthless and get rid of that other one and narrow your focus.

Lisa: Then our last slide, we’re going to end on a decision tree for you. This is really just when should you consider testing programmatic? Are you seeing that search CPCs or costs are rising? Do you need to expand your volume? Then you would want to move on it. Do you have the correct amount of investment and so on? All the way down through the decision tree, this is a guide for you to utilize.

If you have any questions, you can reach out to us on LinkedIn or through our other communications channels. We’re happy to help and get you started in the programmatic space.

Evan: Yes, very happy to help. Again, a super exciting topic, programmatic, very complex. We love it and are happy to help. Please reach out. Thank you, everyone, for joining. It’s been a pleasure. Have a great rest-

Lisa: Thanks everyone.

Announcer: Thanks for listening to this episode of Ignite. Interested in keeping up with the latest trends in healthcare marketing? Subscribe to our podcast and leave a rating and review. For more healthcare marketing tips, visit our blog at cardinaldigitalmarketing.com.