Webinar: How to leverage driving data in auto insurance advertising
Hosted December 8, 2020
Insurance marketers continue to look for new ways to reach the right audience. With traditional advertising, it’s impossible to know whether you’re targeting the most profitable drivers and short-term KPIs like clickthrough, quote, and conversion rate can mask the bigger picture. The result? Wasting valuable spend on drivers who were most likely to click and convert, but who will be unprofitable down the road.
In this panel, Arity experts will share how driving behavior and risk data can inform customer lifetime value when used in targeted advertising. We’ll help you understand how to target the best prospects for your business and how to start tracking metrics that matter.
Speakers:
- Jen Gold, Director of Product Marketing, Arity
- Lisa Jillson, VP of Marketing and Design, Arity
- Fred Dimesa, Head of Marketing and Aggregated Data Products, Arity
Transcript
Jen Gold:
Thank you so much for joining us today to talk about how to leverage driving data in insurance advertising. The opportunity to use telematics or actual driving data to target prospects with ads is a new and exciting concept, and today’s conversation will help you understand how to use this extremely valuable data to target the best prospects for your business and ultimately help your company grow profitably. Today, I’m joined by three of my colleagues here at Arity: Lisa Jillson, Fred Dimesa, and Louisa Harbage-Edell for a panel discussion on this topic. Just some housekeeping before we get started. Your lines are muted and we’ll hold questions until the end, but of course we totally do encourage questions and please use the q and a function to submit questions. We’ll get to them at the end as time permits. We hope we get a lot of ’em. So by the end of this session, you’ll have a clear understanding of how to apply telematics data to your marketing efforts and how to ensure that ways you’re measuring success really align with your company’s broader goals.
In terms of today’s agenda, after our panel introductions, we’ll do a quick level-set around the concept of telematics. I’m sure many of you’re familiar with telematics, so some of what we’ll cover today should serve as a nice refresher. Then we’ll jump into a discussion about how marketers, like many of you can use telematics data to target the best prospects, how to optimize your campaigns to the right metrics and how to measure success. Let’s just start with an introduction to Arity. In case you’re not familiar with our company, Arity is a mobility data and analytics company that combines massive amounts of data cutting edge technology and decades of transportation expertise to help partners make data-driven decisions that solve real problems for real people. We’re able to help insurance carriers connect with the right people at the right time based on their actual driving behaviors and mobility patterns.
We have over 400 billion miles of historical anonymized driving data with more than 23 million active telematics connections and 10 years of experience analyzing, driving data from cars and mobile devices [as of 2020]. So now to introduce our team today, and I’ll start with myself. I’m the moderator today obviously of our panel. I’m Jen Gold, I’m one of the directors of product marketing at Arity. My background is in the ad tech space. Actually having worked most recently at MediaMath, one of the industry’s largest DSPs at Arity, I work with our product and sales teams on offerings that help marketers create meaningful connections with their customers using the power of mobility data. Lisa?
Lisa Jillson:
Thanks, Jen. I am Lisa Jillson, and I currently head up marketing and design at Arity, but my background is really in the B2C advertising space, having worked some for a number of different agencies, both big and small, including the likes of like Leo Burnett and Ogilvy and Mather. In my most recent prior role though, I was on the client side actually at one of our customers, Allstate, and I was leading marketing and communication strategies for Allstate’s telematics products including Drivewise, Milewise, and Allstate Rewards.
Jen Gold:
Great. Louisa?
Louisa Harbage-Edell:
Hey, I’m Louisa Harbage-Edell. I’m currently the other product marketing director at Arity, specifically with responsibility for our insurance solutions as well as sharing economy. I spent the last two decades working for and with insurance companies first as an actuarial analyst at Progressive and then as a consultant helping dozens of companies improve their pricing and products primarily through telematics.
Jen Gold:
Thank you, Louisa. And last but not least, Fred.
Fred Dimesa:
So I’m Fred Dimesa. I joined Allstate – Arity about, and I said that specifically because I joined a different division four years ago with the purpose of trying to solve this problem. How do you use telematics in marketing? And my background is actually in mobile working for IAC and Liberty Media and Marketing Solutions. So I run our whole marketing solutions product set as well as our aggregated data lines.
Jen Gold:
Terrific. Thanks everyone. So to start things off today, we’re going to take a quick poll to see how insurance marketers are actually measuring success today. If you are an insurance marketer, please answer to the best of your ability, and if you’re not an insurance marketer, please answer based on what you think your marketing colleagues would be likely to say. The question is what’s the main KPI that you or your marketing team use to measure success today? And there are five options. Now I realize that most marketers use numerous options to or numerous KPIs to measure success, but try to choose the one that probably aligns the best with the KPI that you use the most. So brand awareness, cost per click, cost per quote, cost per acquisition, lifetime value.
Give it another minute, just choose the one that seems to align most with how you usually measure success. And we’ll close the poll once we feel like we’ve got enough responses. Awesome, thank you. Terrific. Thank you so much for answering and we’ll come back to the responses later in the webinar. So now let’s start with some of the basics and then we’ll get to how it’s all relevant to marketers. So as I mentioned and has been mentioned, a word you’re going to hear a lot today is telematics, which is a pretty well-known concept in the world of auto insurance. But for me, I mean having spent my career until now in the ad tech space and not in insurance, I’d never actually heard of telematics before joining arity, but since I’ve been totally immersed in the insurance world, I now realize how critical telematics is to our business. Can one of you please explain just basically what telematics is?
Louisa Harbage-Edell:
I can take that one. So telematics is really the idea of collecting data about people’s driving behaviors. Things like when, where and how they drive. And by how we mean things like hard braking or cornering, even distracted driving. We collect that data in a number of ways today. So through the computers, they’re built into our cars or via mobile apps on our phones that track location and movement. For auto insurers, the way that typically works is they get a new customer to sign up for a telematics based policy, often called a usage based auto insurance or a UBI policy. They typically often them a participation discount to incentivize them to enroll. They collect data on the policy holder during that first policy term and then usually at that first renewal they offer them an updated rate price to reflect the actual risk of that driving behavior.
Jen Gold:
Got it. Okay, that’s great. Thanks, Louisa. So let’s talk a little bit more about what telematics actually does for insurance companies. Why do so many insurance companies offer their own telematics programs?
Louisa Harbage-Edell:
I mean, coming at it from the product and pricing perspective, telematics is just incredibly powerful data. We use that, as I said, to price policies based on the risk each individual’s actual driving. Historically as actuaries, we used proxy variables like age or gender, which we were very excited about at the time, but they were exactly that. They were just proxies. I mean, this is what we always really wanted was the actual behavior on the road. No data in the world can tell us more about how a person will drive in the future than how they’re driving now.
Lisa Jillson:
Let me go ahead and build on what Louisa just said. Telematics programs. It seems like all of the top players now have them, but from a marketer’s perspective, they can be tricky. Just from a consumer connection standpoint, consumers already have a bit of a barrier on how much they trust their insurer. So when you’re trying to market a telematics program, you kind of get this barrier of you want me to become your policy holder and you want to know where I go and what I’m doing, and that feels a little uncomfortable for me. So as a marketer, that value prop can be a hard thing to communicate, and it’s a fairly complicated program for consumers as well. That said, it doesn’t negate the fact that the data itself can be incredibly powerful for insurer and frankly, the value proposition on the consumer side is you can save a lot of money if you’re a safe driver, you want to get rewarded for it.
Jen Gold:
So it sounds like telematics is incredibly important to insurers, but also it can be really challenging to get at the actual data. Fred, what are your thoughts on the value that telematics brings to the table?
Fred Dimesa:
Yeah, I think I come at it slightly differently because I’m looking at telematics from a…how can marketers use the data? While we know it’s really valuable, our view is it comes too late in the process. The person’s already a customer, and in a lot of states you can’t use telematics to price with. And so it’s largely been limited to the world of pricing and it’s been very important and growing, but the question is can it add more value? And the answer is yes. We’ve found a way to market using this data to all 50 states. You don’t need a telematics program in order to use the data because it is anonymized. It can’t be used to price unless you’re actually running through one of those specific programs. And it really creates a powerful way to target the right set of customers.
Louisa Harbage-Edell:
If you think about it from a product and pricing perspective, that’s been a huge gap there as well. As I mentioned, right now we don’t get that data until renewal. So if you had it upfront, we could offer our best prices upfront. Right now, we don’t even know if the participation discounts we’re offering are justified. So Fred thinks about it from a marketing perspective, but it has real power from a pricing and product perspective as well. That kind of data at the point of sale would be a real game changer across the board.
Jen Gold:
Yeah, that makes a lot of sense. I mean, it sounds like it would be incredibly valuable as an insurance marketer to get that data, as you said upfront, able to make decisions around marketing efforts based on the way people drive. So do insurance marketers today actually try to use driving behavior in their acquisition efforts, or is that really happening?
Lisa Jillson:
I’ll take that one and I’m going to put on my hat from my previous role. Marketers would love to be able to use driving data, but frankly until now it hasn’t been available. They do use a lot of proxies to kind of get at it, but it’s awfully hard because the data’s just not available. You don’t really know who people are or how they’re going to drive. You might have a location, you may know that they were on a car lot, which could trigger a shopping indicator, but not how they got to that car lot. And so that’s really the difference now is that through what we’re doing now with arity, we’re making that data available so that you really get the best understanding of how people are driving now and you have a good understanding of whether or not they’re going to get an accident in the future.
Fred Dimesa:
I think the other thing that we realized was we had to find a way to do this at scale, marketers have a lot that they need to deal with. They have a lot of spending that they need to manage, and you can’t do this as one-off experiments that can never grow. When we started tackling this problem, what we saw in the market was lots of people with little tiny apps who maybe had a couple thousand drivers in it, and that’s just not going to solve a problem for a marketer. And so what we’ve done is we’ve built up a significant scale of telematics data so that you can use it very widely and in all of your digital channels.
Jen Gold:
Yeah, interesting.
Lisa Jillson:
I’d go ahead and add on frankly, insurance marketers frankly are really sophisticated. They’re always looking for different data usage. They have the ability to really test a lot of different things, and I think it’s because insurance in and of itself is a data-driven business, but to Fred’s point, until you get to scale, it doesn’t ever get beyond the test stage. And once you get to scale, then it’s a whole different equation.
Jen Gold:
That makes sense. And I was going to say, I know from my background, having worked at a massive DSP I mean scale, but targeting both are really, really critical. You can’t have one without the other. You can’t just have scale. You have to be able to target and optimize, but you have to be able to do it at scale to have any real kind of impact. So that makes a lot of sense. So let’s take a step back and talk about how all this really works. First, let’s revisit the poll that we took earlier. Okay, really interesting. So I’m just seeing this live with everybody else, so give me a moment to digest. It looks like customer acquisition. So the top two winners here, cost per acquisition and lifetime value. Interesting. Cost per acquisition is a very clear winner. Almost 50% of the respondents are saying that CPA, cost per acquisition, is their main KPI.
Now again, I don’t want to say that anyone that answered any of these is eliminating or ignoring the other KPIs. I’m sure that you’re all using a really healthy mix of these KPIs, but it’s interesting that CPA cost per acquisition or conversion is the biggest one. So we’re going to talk a little bit more about how telematics data can actually help you measure more than just CPA. Telematics data can help you measure cost per acquisition and other KPIs, but really can help you get to the heart of maybe even a more important metric to you and your company – which is lifetime value. So let’s get into talking about that. First, I want to ask Lisa specifically, because you’ve been in marketing, insurance marketing for a long time, does this resonate both with your own history as an insurance marketer and also with insurance marketers that you’re talking to? I mean, when you’re speaking to insurance marketers, what kind of KPIs are you typically hearing they’re using to measure success of their campaigns?
Lisa Jillson:
So efficiency is always going to be important, especially when you’re talking about the lower funnel. But what is most important really depends on what part of the funnel you’re trying to fill. If you’re looking at upper funnel, we know that awareness is a really key driver of getting somebody from the upper funnel all the way down to the bottom. So awareness is going to be key. And so you’re looking for things that are going to resonate from an awareness standpoint, but on lower funnel it’s going to be about efficiency and it’s going to be how am I going to drive those shoppers that are in market now down to quote and bind? And I think the key is the equation, and we saw that a little bit in the poll, is understanding the value of that customer. What is that right value? What is that right target?
How do we get not only the upper funnel, but even the lower funnel to be through that funnel with the right target audience in mind? And as a marketer, especially even as an insurance marketer, we don’t always have all the tools we need. We haven’t had all the tools. We need to understand what that value is, at least not at the upfront. We can make some educated guesses like you might be a homeowner or you may have two vehicles. We can append data to our buys. We can look at a lot of different metrics in the upfront, but it’s awful hard to equate what the value is going to be at the backend. And that’s basically because we don’t have a way to measure risk upfront. And so while a cost per acquisition is really important, you’ve got to have a cost per acquisition that’s balanced by what the value of that consumer is going to be at the end.
Jen Gold:
Yeah, that makes sense. And I think I know from research and from talking to countless marketers and insurance carriers, a real typical company, KPI, and for many companies even outside of insurance is profitability, retention rate, overall customer lifetime value, and of course combined ratio. So those are kind of more company-wide KPIs. Typically our marketing teams commonly using the same KPIs as those overall company KPIs or are they kind of measuring different things and is there a misalignment there?
Lisa Jillson:
There’s no misalignment. Again, I go back to insurance marketers are pretty sophisticated and those they are absolutely looking at profitability, which is what you would expect out of any company to measure success. I think the question really is how granular can you get on projecting what that profitability is going to be? And you can use telematics data to get far more pointed on projecting that profitability and you can now start to do that at scale. So typically what an insurance marketer would do is they would get a profitability metric, an allowable acquisition cost if you will, that comes from their internal analytics team, but it’s kind of based on an average across a wide swath of people, not necessarily specific profitability on specific subsegments. Now we can do that with telematics data.
Fred Dimesa:
Yeah, I think just building on that, that’s really the goal. We want you to be able to target the individual based on the individual’s risk to your business. And the challenge with insurance is it’s a very long tail. It takes forever to figure out profitability. This isn’t like selling jeans when you sign someone up, you have a good guess at the risk they represent because they fit that broad average, but if you actually knew their true risk upfront, you can really do some one-to-one marketing. You can really drive some effective messaging and you can do that in the upper funnel doing your brand awareness with the right audiences as well as figure out who of the people who are just raising their hand and saying, I’m shopping for insurance, are actually worth going after. Because sometimes the people shopping for insurance aren’t actually the people you want to bring onto your book of business. Yes, you have pricing to kind of manage that, but at the end of the day, if that person signs up, they still represent a risk to your business and your pricing may not.
Lisa Jillson:
All right, Fred, I’m going to argue with you though. It is like selling jeans because, if you’re a woman and you’ve ever tried to buy jeans, it is impossible to figure out what jeans are going to fit you. So there you go. True. I think it’s a lot more selling jeans than you think, but you don’t know. So if I knew which jeans fit upfront, I would know which ones to pick. So now with telematics, I kind of know what customers fit, what customers are going to fit to get me to profitability. Intent is a piece of this. You’ve got to look at intent as a critical part of the equation. But what you can’t do is you can’t only look at a tent, especially if intent is driven because of what might also be a shopping trigger. Like let’s say intent is driven because somebody got in an accident and now they’re shopping for more insurance. That isn’t necessarily a good harbinger of whether or not going to be profitable. And if the people that are shopping and we don’t know what this percentage is, but if the people that are shopping have a high percentage of people that are actually not as profitable, then that’s a problem and being able to start to piece that out is a real advantage for any insurer.
Jen Gold:
Got it. I love the jeans debate. I mean I get sense at Lisa’s point in the sense it is one of the hardest things to shop for in the world is good pair of jeans, but also to Fred’s point, I get that a pair of jeans has a certain cost associated with it and you set the price and boom, your profitability once it’s bought with insurance. I mean, I know one of the big challenges is you can sell a premium but you don’t know for sometimes years and years down the line how profitable that policy really that sale of that policy really was. So to be able to know that information upfront, to have that crystal ball and be able to project into the future and understand how risky a person is likely to be and align your cost of acquisition with their profitability would be incredibly helpful.
Helpful. But coming back to the conversation, it sounds like it can be really challenging to connect the metrics that an insurance company overall uses to measure success and with the metrics that marketing teams are tasked with for measuring success. To your point, Lisa, insurance marketers are incredibly sophisticated, but sometimes they’re just gold with metrics that they need to either hit or stay under and that’s the way that they’re both rewarded and for lack of a better word, punished if they go over. So I mean, talk to me about that. Is there a disconnect and how that works?
Lisa Jillson:
Well, I don’t know that I would say there’s a disconnect, but you’re definitely incented to try to come in under your cost per acquisition. And I think what would be interesting to see now that we have the ability as marketers to use telematics data in this equation is how might we better leverage a cost per acquisition or leverage an allowable acquisition cost for our marketing program including this data because it would change, frankly, somebody that has a better risk profile is going to make my company more money as a marketer. And so I want to spend more to get that.
Louisa Harbage-Edell:
I think ultimately what we really want to be doing is paying attention to that relationship, right between what I would like the customer acquisition cost and the customer lifetime value would be probably what I would argue for, which if you’re not as familiar with it, it’s just the profit a customer generates over the entire duration of their time as a policy holder versus just a single policy period. I works for Progressive in the early two thousands, I think I mentioned that upfront, and specifically I worked as a marketing analyst then it was actually one of the things we focused on back then because they were very clear that the bottom line was if we could figure out how to align this to acquire a new customer, as Lisa just said, with that long-term value that that customer generated, that we came out a winner. And you kind of see that in progressive’s financial results and their growth and their profitability, they didn’t mind spending a little more if they knew that those buckets of whoever they were targeting were going to generate better revenue for years.
And they knew to cut back if they were going to be less profitable. But in the early two thousands, we were still making guesses. We were still going after buckets, right? It was like we knew pet owners might retain better because they were highly passionate about the safety of their pets or something, whereas now we’re talking actual telematics data for that targeting, which is totally different. There’s actually a study published this summer by the institutional investment firm, William Blair that said basically this exact same thing, the ratio between acquisition cost and lifetime value really is the most predictive indicator of success for an insurance company. And because of the way Progressive focuses on this, this is from the William Blair article, their lifetime value customer acquisition cost is like three to five times better than their competitors, which again, we all see play out in the way the market responds. So that’s just kind of goes on to reinforce, I think exactly what you were saying, Lisa, that really is sort of that golden ratio if we can nail that.
Jen Gold:
Well, that makes total sense to me, and I mean I’ve seen that William Blair study and Progressive definitely is kind of head and shoulders above the rest specifically in terms of that ratio between lifetime value and customer acquisition. So I mean, you’ve sold me. That makes total sense. So my question then is why aren’t more insurance marketers using metrics that include lifetime value and using that metric of the relationship between the customer acquisition cost and the lifetime value as a benchmark for success? It makes total sense.
Lisa Jillson:
So short answer is they’re absolutely doing that. I don’t know of any marketer, not just insurance marketer but marketer that isn’t using a lot of data to better target their buys. They’re looking at segmentation around different policies, segmentation around a household, even how a potential prospect might influence others through word of mouth. So there are a lot of additional data points that as a marketer you want to leverage to go forward. I think the key is that none of those have been as focused around risk. I think the closest thing has been the idea of credit factors. So for those in insurance about 20-ish years ago, there was a pretty clear correlation understood between what your credit risk was and what your driving risk was and what your ability is to leverage that kind of data. And the upfront is always a benefit. So you can use things like credit scores or household income or home ownership, but insurers haven’t had the telematics data, which actually is far more likely to show risk than any of those until now. So there is an ability now to have a much closer proxy to how somebody is actually going to behave as a policy holder. And it’s going to be really interesting over these next few years as that data is now available to see who jumps on board the way they did 20 years ago when credit was first discovered. There were a number of people that jumped on board very early and it kind of changed the equation of where they sat in the marketplace.
Fred Dimesa:
I think just again, building on that, the biggest thing here is that the data enables you to come in at your target acquisition costs. It just means that you may change your bidding strategies, right? You’re going to bid a lot less for someone who’s just an average driver than you might for someone who is a stellar driver. It doesn’t mean you have to overpay, it just means that that’s where your focus is. It also may change your channel mix. Right. Now you have a channel mix that is dictated by performance and how you measure performance. And so factoring this in, you may actually shift and say, you know what? That channel while looks great from an acquisition and brings me way in under budget. It’s really a bad channel for us. And then you’re going to just shift dollars to the better channels that have better risks in it. And I think that’s the mindset shift that has to happen. This is something that, yes, you pay for the insight, but that insight is incredibly valuable and it doesn’t mean it has to blow up your budget to deliver it.
Lisa Jillson:
And I’ll go back to volume is just going to be critical to that success because while you pay for the insight, you want to be able to leverage that insight on as many prospects as possible because growth is still going to be an issue. Growth will always be an area that marketers want to drive towards, and you can’t better harness where you’re going to focus your growth resources unless you have scale to be able to use that data across the whole prospect set.
Louisa Harbage-Edell:
I’d add, it’s not just that we have this data at arity and we have access to it. It’s that the data itself when we think about telematics and Arity’s data specifically, is so predictive because ities models are tied to the largest database of actual loss data out there. I mean at all in the United States. And so the models that we’ve built on that are highly sophisticated, so we’re able to get real differentiation. So when we look at the tiers that we’ve set up for targeting that marketing, we’ve run tests that show that the top quartile of drivers are on average like 3.3 times more profitable than just the average drivers on the road. So that’s some real differentiation.
Jen Gold:
So that’s…
Lisa Jillson:
Not like 3%, that’s like 300%. Yeah,
Louisa Harbage-Edell:
Exactly.
Jen Gold:
Let me get that straight. So the top quartile of scored drivers based on their driving behaviors are actually on average 3.3 times more profitable than average drivers on the road.
Lisa Jillson:
Yep. Right.
Jen Gold:
That’s tremendous. I mean, that’s a huge, huge jump in profitability. That’s really powerful. Wow. Well thank you guys. So let’s look at this a different way for a minute. Take a step back and ask why is it so critical for insurers to target the best drivers? Anyway, that might seem like a silly question Now after you’ve given me a lot of reasons why insurers should target the best drivers, but right now, I mean anyone who’s turned on a screen in the past couple of years knows well that the model we’re seeing today are the big brands, blanket messaging, everyone with generic ads. I don’t feel like I’m being targeted. I feel like I’m just seeing generic ads everywhere for some of the biggest insurance companies. So what’s wrong with that model? Why don’t we just keep doing that?
Lisa Jillson:
I’ll take that one. First of all, I’m not so sure having been an insurance marketer for a long period of my career that I would say they’re all generic. So I think that there was a lot of really good value props by insurance marketers, which is all a good thing. But there’s also a tremendous amount of money being invested to try to go after the market and sway the market. I think at last measure, it was something like seven or eight billion a year spent on insurance marketing, which puts it number four or maybe number five on the list of overall categories, which to me is a little insane. It almost feels like even when you’re a marketer, you’re in somewhat of an arms war to try to go after convincing prospects. Part of the reason for that is that there’s all kinds of evidence that driving awareness and consideration, again, that top of the funnel is so important at what comes out at the bottom.
If you don’t have top of mind awareness, you’re not one of the first three companies that somebody thinks of when they’re ready to shop for insurance. You don’t have a shot at getting them to quote and potentially become your customer. So there’s a tremendous amount of money spent throughout the funnel, but especially at that top end of the funnel. But now that digital allows us, digital marketing allows us to use data across that whole spectrum of the funnel, I think that we’re going to see that model change. It’s not that I necessarily think that there’s going to be a huge drop off in spend overall, but I actually think insurers are going to get a lot smarter about how they spend that. I’ll use the example of Louisa and Fred. If they look on paper, they’re the same, but I know that Fred is a much better driver than Louisa. It’s not really the case, but let’s just say Fred is a much better driver than Louisa. I’m going to spend more to get Fred, and I may forego getting Louisa, and it’s not even just to get them to shop. I want Fred to have me as top of mind, and I care less about Louisa, so I’m still going to blanket the airwaves, but I’m really going to focus my investment on the ones that when they’re ready to shop are most likely to come all the way through the funnel. To me,
Louisa Harbage-Edell:
I’d also say there’s a real opportunity to level the playing field here because let’s be honest, there’s a lot of insurance companies that can’t compete with billion dollar mass marketing ad campaigns. So they have to spend smarter, which this kind of advertising would allow them to do, right? They could target the most profitable customers without having to go crazy with some of the mass market, and they might not have the top of funnel awareness, but it might allow them to compete in a different way. Knowing who those drivers are and targeting the ad dollars specifically at the audiences they want then would be really critical. They’d have to be laser focused if they’re going to compete. And the only way to do that would be by using targeted ads based on higher prediction values of things like lifetime value with telematics.
Fred Dimesa:
And that’s actually, we’ve seen that in our marketplaces today. We actually see very large companies come in with very large budgets, and we also see really small companies come in and really get a lot of traction out of it, startups included. So very big range of companies are really using this data, and I think that’s exciting. It does give everybody the opportunity to play.
Jen Gold:
Yeah, that makes sense. I mean, obviously in programmatic there are, it’s programmatic specifically, and across digital advertising, there are so many different ways to get your message in front of customers, and there are ways it does level the playing field. The cost of entry is very low. You don’t have to spend millions or tens of millions of dollars just on generating creative, and you can get your foot in the door. You can test, optimize, run across many, many channels, work with different partners, test data, get results very, very quickly and iterate. So that makes a lot of sense. I like that idea of leveling the playing field, testing what works on a small scale, and then scaling once you’ve optimized the balance of your spend and the results. Makes a lot of sense. Alright, so we’ve talked a lot about the challenges and the opportunities of telematics data and how it can be used by marketers, but in practice, how does this all work, Fred, as the product owner? Can you explain how marketers can work with Arity to harness this kind of data for their campaigns?
Fred Dimesa:
Sure. So we offer a few different solutions. We offer a private marketplace, which is how we began this whole exercise and what that private marketplace basically is, this is where we collect the data from the consumer, so the consumer’s continuously seeing feedback on their driving. We are scoring that driver and then we’re giving the opportunity for the insurer to come back in and create specific offers based on how that person drives. You can actually use that in pricing. It’s a contextual ad, it’s very powerful. But what we learned pretty quickly was it’s got some limitations on scale like all private marketplaces do, unless you’re talking about Facebook, right? Most wall gardens are fairly small, but they have a really high value use. And then we said, look, let’s actually take this same data and push it out to where people are already spending. Let’s actually enable your marketing today with the data.
And so what we did was we actually pushed that out through LiveRamp. It’s available in any DSP. [Note: Arity no longer works with the LiveRamp DSP as of 2023.] It’s actually live just sitting in several DSPs for you to use. You can’t use it to price in that scenario, but you can absolutely use it to target, but you can actually see it. And if you can’t see it and you’re using a different DSP that we don’t have signed up, we can actually just push a custom audience to you. And we’re learning as we hear feedback from our customers, we’re creating custom solutions for them, we’re helping them design tests, we’re helping them to decide where to use it, where not to use it, and we’re collaborating to build better product. We’ve only been working on this for four years, and it’s really only been live at any scale for two, and some might argue one. So it’s a new product, it’s a new way of doing this. This is how you’re going to be on the cutting edge of what’s happening in marketing, but we believe it’s going to change how insurance is sold in the country.
Jen Gold:
Great. So just to be clear, the way that insurance marketers can actually activate these kind of data segments and target people based on risk is by working with LiveRamp, working with their DSPs. Can you dig into that a little bit more?
Fred Dimesa:
Well, you can work with LiveRamp or you can come directly to us. We operate our own private marketplace. We have a portal that you would come through, you would sign up to use that portal. We would authorize, you authorize your users and they could use the private marketplace.
Jen Gold:
And then there are segments…
Fred Dimesa:
And there are segments within that. Okay. Fred, we also push to LiveRamp.
Louisa Harbage-Edell:
I have a question here that rather than hold to the end, I want to make sure we get clarity on. Can you tell us what exactly a DSP is?
Fred Dimesa:
Yeah, it’s a demand-side platform. So a demand-side platform is something like Google DV 360. It’s the trade desk. It is Media Math, places like where Jen was. So this is where ads are placed digitally, whether that’s a traditional digital display ad or it’s video or it’s OTT, it doesn’t really matter. You can still use the data in any of your buys that are digital. And every carrier seems to have its own preference for a DSP. And so we just want it to be agnostic and say, look, if you want to use that, we will make it happen. We’ll make the data show up to where you can actually deliver against your target
Lisa Jillson:
And OTT meaning over-the-top television – to not use as many acronyms for those of us that are still getting used to all the acronyms
Jen Gold:
To build on what Fred said. Thanks. Yeah. Having come from the DSP side, myself, Fred, a few Trade Desk, Media Math, DV 360, there are a bunch, if you’re an insurance carrier, your marketing team likely has a relationship with A DSP already. They are sort of the partner that marketing teams use to run digital advertising across digital media programmatically, which is automatically. And so if you’re not sure if your company works with the DSP, I’d say if you’re on the marketing team, you can talk to your digital marketing folks or ask your marketing team, but you likely already have a relationship in place if you do any kind of online advertising.
Great. Well, thank you, Fred. That was really, really helpful. I’m going to start wrapping things up because I want to make sure we have time for, we’ve got quite a bit of questions I’d love to get to before I wrap up and list out some key takeaways. I just want to ask the panel if there’s anything else that we missed or any other comments or questions that we had on the panel. If not, I will start to wrap things up and then move into the q and a portion. I think this has been a really great discussion with a ton of information. And for me, there’s a lot to latch onto. So I wanted to mention a couple of things that really stuck out from this conversation. And I’ll be sharing, we will be following up and sharing these key takeaways in an email follow-up later today or I believe tomorrow.
But some of the key takeaways for me are really that the best predictor of future driving behavior is past driving behavior. I think that makes a lot of sense. Secondly, the telematics data is really incredibly powerful for understanding how people drive and move throughout the world. I think we can all agree with that, but until now, marketers really haven’t been able to harness the power of telematics data to target drivers based on their driving behavior and risk level. It just hasn’t been something that’s available to them. I think we heard a lot from all three of our panelists about how a really critical metric to measure for marketing campaigns is that ratio between a customer’s lifetime value, LTV and the customer acquisition cost or the LTV to CAC ratio. And we’ve seen with the William Blair study and just historically and economically that companies that focus on and measure success based on this ratio are growing more profitably than others.
And finally, with Arity marketing solutions, marketers can now leverage real-time driving data to target prospects with the best lifetime value upfront and optimize campaigns using that LTV to CAC metric. So I think those are some good key takeaways and I’ll be sending out, I should say we’ll be sending out a follow-up including a recording of this webinar that you can share with colleagues who weren’t able to attend. So I know one of the questions we got on the Q&A was, will this be recorded and shareable and Yes, absolutely will. We would love for you to share with your marketing teams, your product and pricing people, your agency and DSP colleagues. If you do work with agencies and or DSPs demand side platforms, please share this recording with them. We hope that this information will be relevant to them. I think it really is. But now with that, let’s open things up to Q&A. We’ve got a few ready and if you haven’t submitted a question, please feel free to use the Q&A feature to submit and we’ll answer them. We’ve got about a few minutes left, at least 10 minutes left. One of the questions that was submitted here, let me just take a look. How closely is telematics data correlated with retention? So I think this is a really interesting one. I think Fred or Louisa…
Fred Dimesa:
Yeah, I’ll take the first crack at it, and Louisa will give you the more detailed answer. There is a correlation, it is not the most important correlation. The correlation is with risk and retention is an element of profitability here. But yes, better drivers do tend to retain more. We have some specific data. I’m not sure if we’re able to share that, but we do have some specific data around that. Louisa, if you want to add to that?
Louisa Harbage-Edell:
I mean, I think there is absolutely a correlation there. As you’ve said, we see that, I think even in general, books of business, right? Tier one drivers, the premium preferred drivers retain longer than non-standard drivers, and that translates directly when you look at telematics tiering as well. So your top tier drivers retain longer if you’re running a usage-based auto insurance book. Obviously we also see increased retention with UBI programs as well, but even if you’re not, you’re just using this for pure marketing, there is absolutely a correlation there. To Fred’s point, I don’t know that we can share the exact numbers that we see in our customers, but there is definitely a lift, and of course that will impact your customer lifetime value and increase profitability longterm and reduce your marketing spend for having to acquire new customers to replace the old ones.
Jen Gold:
Great.
Lisa Jillson:
I’m going to add just one thing onto this, and that is early on in this webinar, Louisa talked about how incredibly important telematics is to pricing the opportunity. And part of the reason why so many insurers have a program in-house for telematics is they want to understand how they use that for pricing. While what we’ve been talking about mostly today is using telematics for targeting and marketing, which is different than pricing. When you start to get into retention, that gives you the avenue. If you knew upfront that somebody that you acquired was an extremely valuable target, one is you may want to change the price, which will affect retention because frankly, they’re getting a better deal from you than they would from other carriers. So it absolutely has the opportunity not just from a pure, better drivers tend to retain longer, but it also has the opportunity when you can goose it with pricing on top of it or retention, renewal pricing on top of it, it really could affect retention in a big way.
Jen Gold:
That makes sense. Thank you all three of you for answering that. So we’ve got another question, which is about the private marketplace. How many drivers are in the private marketplace that you can actually reach?
Fred Dimesa:
Currently, there are about 23 million drivers in our private marketplace that you can reach [as of 2020], and there’s a similar number in our audience network. We can reach pretty much all of those drivers, and through either channel, you’re going to see them more often in what we call our audience network, which is our LiveRamp integration with the private marketplace. You are limited to when people open that up and when they go to where the ads are actually running. So it limits the potential of the audience in terms of scale, but it definitely gives them a much more holistic experience, much more contextual.
Jen Gold:
Great. Thank you. There have been a couple of folks interested in taking a look at that William Blair article, which I think is great. I would encourage everybody to take a at it. It is really rich with very detailed quantifiable information. We can share information about how to, we can’t share their study, but we can share information about how to access that study after the webinar, either as part of our follow-up email or we can reach out to the individuals that asked about it afterwards. So thanks for asking about that. I have another question here, which is an interesting one. So doesn’t using this data cost more money upfront?
Fred Dimesa:
I’ll jump in on that. There is a fee to use it. It doesn’t have to, we talked about this a little bit earlier. It really is all about how do you use the data to bid and then how do you use the data to select channels? So in some cases, we’re actually delivering very, if you’re a lower funnel advertiser, we’re delivering very low cost per click and cost per call and cost per quote. So in the upper funnel, we’re hitting metrics that people are wanting to see, but it has an expense to it. But that expenses a can be made up in better channel selection, better bidding, and actually at the end of the day, it’s all about the LTV from our perspective. Even if it cost you a little more to get someone who is 3.3 times more profitable and you paid 10% more to get them, that’s a pretty good deal. Most of us, if they had to pay 10 cents to get $3.30, we would spend the 10 cents. Right? And that’s what we’re talking about.
Jen Gold:
That makes a lot of sense. Okay. Looking through the questions here. So Louisa, you mentioned a figure of 3.3 x profitability. The best drivers are on average 3.3 times more profitable than average drivers. Question here is how do you guys know that? How do you calculate that number?
Louisa Harbage-Edell:
Yeah, so we’re very lucky. We have a lot of very smart data scientists and actuaries at our company, and they actually model this out for us. So they look at, as I said, we not only have driving data, we also have risk data attached to that. So we can look not just at how many miles people drove, but also frequency and things like that. So we look at the actual risk of the drivers on the road and compare that to the average population for those segments.
Jen Gold:
Got it. Great. We all have access to the q and a. Does anyone else on the panel see any questions that I missed that any of you might want to answer? I think I’ve gotten them all. We’ve got, again, a couple more people chiming in wanting that William Blair study, which is fantastic. Anything else? Well, I think we’re good. As I mentioned, we’ll be following up with some key takeaways and some more information. Link to the recording. Please. Like I said, share this around with your colleagues internally and any agencies or DSPs that might be interested. I think they’d be very intrigued by some of the information and the data. We really appreciate it. Thank you guys so much for joining us. This has been a lot of fun. Thank you all panelists, Lisa, Louisa and Fred, this has been a great discussion and I look forward to continuing in future webinars. So thanks everyone. Have a great rest of your day.