eMarketer Tech Talk: How behavioral data can transform marketing strategy
Speakers:
- Lisa Jillson, VP of Marketing and Design, Arity
- Rob Nendorf, Director of Data Science, Arity
- Fred Dimesa, Head of Aggregated Data and Advertising, Arity
Transcript
Nancy Ferra Santos:
Hi everyone. Thank you for joining us for another eMarketer Tech Talk where we host guests of all areas of adtech, MarTech, and media. I’m Nancy Ferra Santos, senior Vice President, eMarketer, insider Intelligence. Today, we have an excellent presentation. I’m very excited about it. We’re going to hear how mobility can be used to convert valuable customers and maximize your profitability. This is your opportunity to learn. We want you to have an educational journey, so please submit questions in the questions tab. We’ll get to as many as we can during the q and a, so please keep them coming. Now I want to introduce my special guest from Arity. I have Lisa Jillson, VP of Marketing and Design. I have Rob Nendorf, who’s director of Data Science, and I have Fred Dimesa, head of aggregated data and advertising product lines. Lisa, Rob and Fred, thank you so much for joining us. Lisa, I’m going to hand it over to you to take it away.
Lisa Jillson, Arity:
Thanks, Nancy. Welcome, everybody. My name is Lisa Jillson, my head up marketing and design at a company called Arity. Just a little bit of background for those of you that might not know us, ity is a mobility data and analytics company, and we are focused around creating behavioral insights and patterns that help make sense of the how, where, and when people drive. We’re super excited to be here with eMarketer today, and we’ve got just a terrific audience of top-notch marketers on this call, and we’re going to spend the next 25-ish minutes trying to dive into that idea of behavioral data and thinking about behavior data in a different way can really help transform your marketing strategy. I’m sure I’m not alone. Many of the marketers on this call really understand that finding ways to not only understand and connect with the best prospects is tough, and it’s getting even tougher, and I think it’s getting even more important in this very competitive landscape that we now operate in. I’m going to bring in a little bit of data from foresters. Almost 66% of us, almost 66% of marketers say there is more pressure than ever to prove the ROI or the business performance of the activities that they’re responsible.
It’s a common focus area of marketers today that we have to hone in on building better, smarter strategies that help us optimize not just the quantity, but the quality of the prospects that we’re targeting so you don’t waste money converting customers that aren’t going to deliver on that ROI. So before me and my colleagues get into the meat of the conversation, I’d like to ask all of the participants a couple of questions. And this is really informal. It’s not actually a poll, so you can just put your answers in the chat, but there’s going to be two questions. Here’s the first one, do you use behavioral data as a part of your current marketing campaigns and strategies? If yes, go ahead and put yes in your chat and if no, put no, I’ll give you a second there to go ahead and populate it. All right, here’s question number two. Have you used behavior data other than shopping and intent data, those hand raisers data, if you will this time? If the answer is yes, put hell yes, and if you haven’t yet put not yet. All right. We’re going to ask our friends at eMarketer to do a bit of a tally in the chat, and I’m going to come back to that at the end, but let’s go ahead and get started.
[Marketers] have to hone in on building better, smarter strategies that help us optimize not just the quantity, but the quality of the prospects that we’re targeting so you don’t waste money converting customers that aren’t going to deliver on that ROI.
So today we’re going to dive into three different questions. How can you use this unique data to help optimize the KPIs that matter most for your company and in that, help outmaneuver your competition? Number two, how can data about how people behave in the real world, how people drive actually become a critical advantage to understanding and helping you reach your very best customers? And then number three, what strategies and tools do marketers need to go beyond just driving what I’ll call the easy sale that those folks that are already hand raisers, those folks that are ready to buy today and go to not only capturing that but capturing that in a way that delivers significant profitable growth to your organization? Now, back to that Forrester data. Again, over 70% of marketers say that consumer data is already the lifeblood of their marketing strategies and a very common type I would venture to say maybe the most common type of behavioral data that’s used right now is shopping or intent data.
And that’s really powerful. It’s important to know who’s raising their hand, who’s shopping today, and it kind of creates the backbone of what we often see as performance marketing. But the reality is is that so many of us are using it that it’s actually not as differentiating and can actually be pretty expensive from a marketing strategy perspective. It obviously has links to quick sales, but it actually might fly in the face of driving towards profitability. So figuring out what type of data to use to not only attract the most profitable customer but also to gain a competitive edge is what we’re going to talk about today. And that isn’t always easy. You can add unique data sets, but how do you know what data sets are going to help provide a competitive advantage? And how can we leverage those data sets in a way that help you not only to target customers but actually understand what that customer is going to become when they become your customer? How much profit are they going to bring to your organization? Now, the data that Arity focuses on is a data we call mobility data. And that mobility data, we define it as kind of understanding where – even the why – people go from point A to point B, and it gives you a much richer set of insights to really understand what consumer patterns are. But me talking about it’s not the same as looking at some use cases. So let’s go ahead and do that.
The data that Arity focuses on is a data we call mobility data…and it gives you a much richer set of insights to really understand what consumer patterns are.
So here are some use cases on mobility data and it can be valuable to a whole host of different kinds of marketing use cases. Maybe one of the most obvious ones is auto insurance. Auto insurance carriers are always looking for ways that they can reach the safest drivers because those safest drivers are most often the ones that are going to be the most profitable, don’t get in accidents. You don’t cost as much money to those insurance carriers. So they use this kind of data to target some of the very best drivers on the road. And those best drivers lead to more profitability. And in fact, if they focus on those best drivers, it can mean a very significant competitive advantage because the switching between insurance agencies is actually not very frequent. So once you capture those best drivers, you actually have a competitive advantage for a long period of time.
But there’s a lot of other industries, industries that can also use mobility data to create unique and profitable targets efficiently and even at scale. Let me pull out two more specific examples out of this list. One is car manufacturers and dealers. They may want to look at driving behavior data so that they can actually create a profile or a pattern of driving that they can use as a targeting parameter. They could on drivers that are very interested in safety or maybe have a driving profile that is focused around comfort or even fuel efficiency. QSR restaurants or retail establishments, let’s say like Dunkin or, I don’t know, maybe Speedway, can help reach customers or consumers that drive long distances and may actually even frequently pass their locations, but don’t always make a stop. But we could create driving patterns and audiences around those driving patterns that could be really fruitful targets for those different categories.
Let’s dive into another example more specifically. So we can leverage this driving data in a lot of different ways. And here’s a specific example where we’re taking that driving data and we’re actually layering it on top of each other. For example, you could make a target of low risk drivers, those drivers that are driving safely that actually make 90 plus minute commutes a day or during weekdays and they have high annual miles. Now, that specific group of individuals, once we’ve looked at the data in those three categories, could be ideal for an auto manufacturer that might be marketing, let’s say a hybrid or even electric vehicle that has very strong safety and comfort features. So it gives you the opportunity to find a way to build a target that is really unique in your category. So I’ve given you some ways to think about how driving data might layer into some of your marketing strategies, and it can be really valuable in a number of different industries, but how does all of this driving data actually get compiled and where does it actually come from and how is it pulled together in meaningful insights? And for that part of the story, I’m going to turn it over to my colleague Rob Nendorf, our director of data science to go into some more details.
Rob Nendorf, Arity:
Great, thanks so much, Lisa. So mobility data, put simply, is data about how, where and when people move over time. It’s gathered via devices that people can plug into their vehicles directly from connected cars and through certain apps and mobile devices. In every case, the data is captured in obvious ways that provide immediate benefits to the end users and allows them to opt into sharing their location and movement data. The field of mobility data has been growing pretty quickly in part because of technological advances in how the data can be collected in both 2019 and 2020. The market for mobility data grew at about 30% year over year, so pretty significant growth. But user adoption has also been a big part of what’s fueled that growth, especially as consumers have realized that sharing their driving data can have some nice real-world benefits. For example, mobile apps that collect driving data can provide benefits like tracking fuel efficiency, detecting a crash, recommending safer driving routes based on weather and traffic patterns and a whole lot more.
The market for mobility data grew at about 30% year over year.
And you’re seeing a number of mobile user experiences built around those features. Mobility data is not to be confused with location data. Location data is very valuable and provides information about where someone is at a point of time, like a snapshot. This information is useful, but it doesn’t tell the complete story about when and where someone moved throughout the day or specifically how they drove while getting around. For instance, many people drop their kids off at school like I do each weekday morning. Some start from my neighborhood as well. So from a location and timing standpoint, we look very similar in this part of our days, but maybe my neighbor leaves a bit later and is speeding on the way there. Driving much more aggressively with the gas pedal and the brakes. Maybe they’re tailgating other drivers and using their smartphone to text or set music for the kids while they’re driving and understanding these things paints a very different picture of each of us that location alone cannot capture.
Now the most valuable kind of mobility data is data that’s enriched with information from additional sources like insurance claims, demographics, weather, speed limits, and information about the roads that you’re traveling on and much more. This contextual data helps us understand not only how people move, but how people move relative to the world around them. For instance, driving 40 miles per hour on a highway with a speed limit of 50 miles per hour is very different than driving 40 miles per hour in a residential area with a 25 mile per hour speed limit. How people respond to bad weather tells us a lot about the risk on the road, but it tells us about who they are. This kind of contextual data paired with mobility data is what marketers can leverage from Arity to attract customers who really make the most sense for their business. Understanding how people drive is really interesting for companies who care about things like risky driving, things like sudden acceleration or braking, speeding and distracted driving, for example, auto insurance companies want to attract maybe the least risky drivers who are likely to become the best customers and they can use driving risk audiences to reach those most profitable drivers.
Auto insurance companies want to attract maybe the least risky drivers who are likely to become the best customers and they can use driving risk audiences to reach those most profitable drivers.
On the other hand, auto aftermarket companies who sell repair services and replacement tires and glass might actually target the higher risk drivers on the road or long distance drivers who are more likely to wear and tear on their vehicles. So there are different ways that different industries and different marketers are going to use mobility data to drive the most value. And there are plenty of other ways mobility data can be used by industries of all kinds, not just for marketing, for gig detection, driver scoring, real-time traffic insights and more. So hopefully that gives you an overview of what this data is, how it’s captured and how it can be used. But now Fred’s going to walk us through how marketers can bring mobility data directly into their campaigns.
Fred Dimesa, Arity:
Thanks, Rob. The biggest thing that my team is challenged with is we have to take all of this data at scale. We have about 145 million U.S. consumers where we have mobility data that we collect on a regular basis and make sense of it, right? Give you something that you can use to actually change the effectiveness of your campaigns. And so we spent quite a bit of time doing that. I’ll give you some examples on the next page about how we’re actually bringing this together. So as Lisa and Rob mentioned, we kind of got our start looking at risk data specifically around loss and LTV for insurance carriers. And then what we found is there’s a lot of other application for this. So people were interested in risk in the automotive aftermarket, but they were also interested in simple or things like mileage and how much daily mileage do you have versus average length of trip, where are you driving, are you driving on freeways, are you driving on arterial roads, things like that.
And so we start, what we do is we run all this data through our various insight models and then what we aggregate the consumer into segments where there are some common behaviors. So high risk drivers, low risk drivers. We also do things like gig drivers, right? So the gig driver, you want to know is this person a gig driver, yes or no? But you also want to know are they doing rideshare or are they doing delivery right? Different types of drivers have different types of needs. We’ll have different trip lengths, and you may be a rental car company and you may be wanting to actually market to these folks, but you’re also going to want to know how much wear and tear they’re going to put on the car before you actually market to them. So you can actually be much more profitable with that.
We also do things that are much more custom, much more bespoke for a customer. So for instance, if you were an out-of-home company running a campaign for a movie company, you would come to us and we could actually tell you not only are these the people who were exposed, but actually the duration to which they were exposed to specific signs. And what we do is we’ll work with that partner to actually bring in those POIs [points of interest], look for what it is they’re trying to accomplish, and then we build a custom audience for them using this mobility data. So we think we’re really at the beginning of what you can possibly do with this kind of data. And we look to our customers to really kind of help us define new audiences. Some of these audiences you’ll find out there in the wild, and I’ll actually show you next, how do we access this data?
So the primary way you’re going to access this data is when people have opted to share a mobile advertising ID with us, we create these audiences and we publish standard audiences into tools like LiveRamp or directly into a DSP, and they will be syndicated there, and you’ll be able to just go into your DSP, and you’ll search for them and you’ll see various Arity audiences, other audiences we will send because they are more bespoke. We will send directly to your seat on that DSP. In the case where consumers haven’t shared a mobile ad ID, we actually operate our own private marketplace. So this is where we collect the data from where we’re displaying back to consumers, the driving behavior so they understand what they’re doing or not doing on the roads. We can then deliver advertisements within that private marketplace where we don’t need an ad ID.
And then lastly, very specific to insurance, because this is where we come from, we do work with some lead aggregators to be able to push the risk data over there and the mileage data over there. So our insurance customers can target people who’ve raised their hand and said, I’m shopping for insurance. Here’s the risk profile. So the insurance companies know how much should they spend for that lead or for that opportunity. It doesn’t mean you shouldn’t target those people, it just means you should spend differing amounts based on the long-term profitability. And that rule applies to kind of everyone who uses our audiences. So lemme give you an example. We will flip over to a quick case study so everyone can understand how we operate. So we were approached by a national car maintenance company. They were looking for new ways to bring in customers who were likely to need their services.
And so what we did was we did a bit of an experiment. We said, let’s look at high risk drivers, let’s look at low risk drivers, let’s look at high mileage drivers. Let’s look at what we call super commuters. So based on averages, we can see how long people commute, length of time, as well as mileage, and we can target those people as part of these campaigns. And what they did was they ran experiments against these different audiences to see which ones worked, which ones didn’t. And the net was they saw a 13% decrease in their customer acquisition costs and a 2.3 times higher multiple of services and bookings using these audiences. And it turns out the high mileage drivers were actually the one that they wanted to target more, and that’s actually where they then shifted most of their campaign spend. So with that, I’ll throw it back to Lisa for a couple of additional thoughts.
Lisa Jillson, Arity:
Thanks, Fred. Okay, so we’ve kind of taken the audience here through what mobility data is and covered some use cases, maybe give you some thought starters to think about, but I want to revisit those questions I asked at the beginning of the tech talk. So the overwhelming majority of you are using data, which is great because obviously we saw that as well in the Forrester survey that using data is the lifeblood of what a marketer needs. And behavioral data in particular is really a strong indicator of driving towards ROI. But using behavior data outside of intent data was more of a mixed bag. We had a number of folks that said, hell yes, way to go, way to go and try some new things. But there was a number of you that said, yep, I’m not using that yet, which I guess makes sense because that was maybe one of the reasons why you joined us today, the tech talk, which is terrific.
Hopefully we gave you some new things to think about on how you could incorporate behavioral data into your marketing strategies in a way that not only can drive new audiences, but can help you create unique and even differentiating targeting strategies that can help you create a competitive advantage by going after audiences that maybe your competitors are not yet targeting and help you actually deliver more profit and more profitable customers to your companies. That’s only going to accelerate your return on investment. Now, I think I’m hoping that we can rely on our friends at eMarketer here. I’d like to bring Nancy back up onto the group because I think there’s some questions in the chat. At least that’s what I’ve heard. So Nancy, maybe you can help us with coming back and joining us and giving us some of those questions. And we’ll answer as many as we can get to the ones we don’t get to. I’m sure we can answer offline.
Nancy Ferra Santos:
Absolutely. Lisa, Rob, Fred, thank you so much. This was fantastic. And as you said, Lisa, this driving data is such a differentiator in the marketing mix. I really encourage everyone to check this out. When we originally talked about this topic historically, I started my career 30 some years ago in radio and television. And on the traditional side you would think of driving data as drive times and for outdoor media and so forth. But this is amazing to think of. And Rob, you brought up just the example of bringing the kids to school and I was thinking, thank you for that beautiful memories of carpools in the past in my life, but how much this data could be used. Parents, and this is of course, just one example, have such busy schedules. If you’re able to figure out somebody’s schedule where they have that busy day where the kids have a late practice or something like that, you’re QSR and you can go ahead and market to them, Hey, it’s Tuesday, I know it’s your busy day. Come here. And just such incredible ways to target that’s different. This is great. I love it. And yes, we have questions, so I’m going to jump right to those. And speaking of QSRs, we have a question that starts out saying, I’m a QSR marketer, how can I leverage driving data?
Lisa Jillson, Arity:
Fred, you want to take that one?
Fred Dimesa, Arity:
Yeah, why don’t I jump in. So this is a newer area for us, something that we’ve been thinking about for quite a bit. There’s some interesting solutions out there about visitation data, right? Lots of companies keep track of when you break a geofence and when you’re actually at that restaurant. And what we were thinking about was, well, we’re not going to add any value by adding more data to that mix. What we can do though is we can actually look at, sorry, I have a dog next to me. We can actually look at where you’re going, where you’re likely to be going and predict the route that you’re taking. And so we can know, are you going to visit that restaurant between nine and 11:00 a.m. on a Monday, but not on a Tuesday, right? And so you can actually time your advertising to the person to know, hey, they’re likely to be visiting, or conversely, they’re likely to drive right by your store and not stop in and maybe they’re stopping at a competitor store instead. That’s a kind of level of detail that we can see and we see so much of that data, we can make those predictions with a high degree of accuracy. And so that to us is a much better way to target someone, target them before they’re actually in your store, not after they’re already there.
Lisa Jillson, Arity:
Yeah, I’ll actually even build on that. I think Nancy, it’s an interesting question, especially given the pandemic because our patterns have all changed dramatically. Actually. We’ve got some interesting things on the website about it, but it’s been amazing to watch because we have connections to 140 million drivers in the U.S. and because we have connections to so many drivers, we had it almost like a front row seat to how people change their commuting patterns, their carpooling patterns, when they drive, how they drive. And it’s just been fascinating to watch those changes happen. But I thought about it from I’m a lifelong marketer and just trying to figure out how to change some of those drive times don’t make sense anymore. It is just completely upended in the last two years. It’s almost like we’ve got to make our own new norms and the data is certainly helpful to help do that.
Nancy Ferra Santos:
That’s great, thank you. And another question that’s coming in. This is a really big talk in the industry, talking about data, talking about privacy, what do you give up? So the question is why would a user agree to share their driving data?
Lisa Jillson, Arity:
I’ll go ahead and take that one because it’s one that I actually, from an Arity standpoint, I feel really passionate about. We’ve taken a stand that our role is not to just look and collect data and create insights out of data. We also look at our role as very importantly to help make the drivers drive better, safer, more useful for them. So a lot of how we collect this data is tied to end user benefit. I think actually, I think Rob brought it up in his section of our talk today is we create things like fuel insights, fuel consumption insights. And so we work with partner right now on a way that their end users, their mobile app publisher, and their end users are particularly interested in fuel consumption. And so we actually help their end users better understand their fuel consumption based on their driving patterns, which can vary significantly. Again, if you’re hitting on your brakes a lot, your fuel consumption really changes. So we provide a value to those end users in order for them to get the value. They opt in, they get that value out of opting in. We are transparent upfront that providing that value, you’re going to allow Arity to collect that data and you’re also going to have some advertising served up within that app.
We look at ways that we are absolutely giving value back to the end user in exchange for the data that they share with us.
Nancy Ferra Santos:
Great. Thank you. Very helpful. Another question we have is about measurement. How would you measure the success of a campaign using driving data?
Lisa Jillson, Arity:
Fred, you want to start with that one? You want me to start?
Fred Dimesa, Arity:
I can start with that one and maybe you can add some color around this, but in general, it depends on what your objectives are. What are you trying to achieve? Are you trying to, as we talked about in the beginning, just attract more profitable customers, a pool of more profitable customers and influence your bidding strategy. So paying more for the most profitable and less for the least profitable. Really kind of getting away from this one size fits all cost of acquisition. In the other cases, what you can do, I was mentioning in the QSR example, you can actually look at did I drive more people in, right? Did I get more share of wallet from the people who I saw as visitors? And so again, going back to what is your real objective and then how can we actually help you achieve that objective? I don’t know, Lisa, if you want to add anything more to that.
Lisa Jillson, Arity:
No, I think you covered it. I think to be honest, the more use cases that we work with different customers on how they could use this data, the more we’ve expanded our own view of measurement and to think about measurement in different ways. I think one of the areas that we’re most interested in is how can we add profitability? Because this is unique data. And so there has to be a profitability side to this. We have to be able to help our customers and our partners deliver more than what they are delivering now and not just more leads into the funnel, but actually better leads. So that quality aspect is really important. So that tends to be how we work with our customers from a measurement standpoint is how can we measure better quality.
Nancy Ferra Santos:
Thank you. And we have a question about the weather. I like this. Do you see driving data change based on weather? If so, what types of differences?
Lisa Jillson, Arity:
Oh, Rob, that’s definitely one for you.
Rob Nendorf, Arity:
Sure. Yeah. So yes, layering on this contextual data, we do that for a reason. As I went through my section, it does provide value and how we tend to think about weather is yes, it will change the overall driving patterns within a city or within a geo region, but we also like to think about it as how do people react to the weather? So if it’s one thing to drive aggressively on a perfectly sunny day, it’s another thing to drive aggressively in the middle of the night when it’s pouring rain. And so that’s the kind of context that when you pair it with the raw mobility data of how aggressively you’re driving and where and when you’re driving, that’s where you really see differentiation in behaviors that can be quite meaningful.
Nancy Ferra Santos:
Thank you. And speaking of that, driving aggressively at night when it’s pouring raining, we have another question that says, how do you know if a driver is safe or unsafe?
Rob Nendorf, Arity:
Yeah, yeah, it is a great question. So we have essentially the largest data set around that pairs mobility data, such as these driving behaviors that we’ve been talking about with actual insurance claims data. And why that’s so interesting is that includes not just whether an accident occurred, which you can find from multiple sources, but it really has in-depth information on the human outcomes such as hospital bills, lawyer involvement, auto repair costs. And so we have detailed information on all of those outcomes. And so being in the data science team, we really have the privilege to build risk models using sophisticated machine learning and actually validating those models against real world outcomes. And not everyone can do that by any means. And so this helps us design all of our data and analytics capabilities in ways that best capture meaningful behaviors on behalf of our clients and our end users.
Nancy Ferra Santos:
Great. Thank you. And Lisa, Fred, Rob, thank you again for all this great content answering all these questions from our audience. Thank you to our audience for asking all these great questions. We really appreciate it. Before we wrap up, let me take a moment and tell everybody what’s happening across eMarketer’s media channels. You can register for more upcoming Live Tech Talk and analyst webinars at emarketer.com/webinars on the audio side. Don’t forget behind the numbers. That’s our daily podcast and keep an eye out for our newsletters. Also, keep an eye on your email for a link to the recording for today’s presentation as well as the slides. There’s a really good blog post that we’re going to share with you all. Share it around your organization. This was great information. Differentiate your marketing, use driving data. Again, thank you all for your time. Thanks again, Lisa, Rob, Fred, and the entire team at Arity. We will see you all on our next tech talk. Bye-bye.