Upcoming IP Investor Dialogues: Spirax Sarco, FAST, COST, CPRT, POOL, SNOW, DDOG - Register your interest at help@inpractise.com 

Anonymous

Disclaimer: This interview is for informational purposes only and should not be relied upon as a basis for investment decisions. In Practise is an independent publisher and all opinions expressed by guests are solely their own opinions and do not reflect the opinion of In Practise.

{audio:0:00:20} Randall, can we take a step back to when you first joined Cardlytics? What was your role and responsibility when you first joined the company?

{audio:0:00:26} I was the group president for the advertiser side of the business, working for the two co-founders. Cardlytics has an advertiser and banking side, and I ran the advertiser side with sales, client service, data and analytics; all the various functions supporting the advertisers and their programs on the platform.

{audio:0:00:49} What originally attracted you to the business?

{audio:0:00:52} Cardlytics had an interesting value proposition for marketers at scale. There are few marketing activities which drive significant top line growth at a high return on ad spend, because of the scale of the platform. There was a handful of interesting things under what they had built. They had the ability to use the anonymized purchase behavior data from credit and debit card transaction data of 161 million monthly average users. They could use that for targeting and could bring in second party data to match to that. They could deliver the cashback offers through the banks and e-banking applications.

{audio:0:01:52} That could be either the bank's app or their online banking website, which meant a pristine channel free from bots and fraud, all behind the bank firewall, a safe environment for marketers. The third thing was their closed loop measurement of the actual sales impact of the programs they were running. When you do those three things at scale, with a US audience of 160 million monthly average users, that is an interesting proposition for a marketer.

{audio:0:02:32} How would you describe closed loop measurement?

{audio:0:02:36} Think about the gold standard of efficacy in the pharmaceutical world. When a new drug is being tested, they use a double-blind test giving the drug to a group of people. Another group of people get a placebo, and nobody knows who is getting what, other than the people running the test. They compare the results between the two and, in essence, closed loop measurement does something similar to that. Cardlytics would typically run a campaign and 10 million people would see the offer, but several hundred thousand people – that were just like the people who saw the offer – were not served the offer. Over the course of the campaign, you can look at how the purchase behavior differs between these two groups.

{audio:0:03:33} That is considered the gold standard of measurement in advertising and media. It is very hard to do and few players can do it, but it is the single best way to measure the efficacy of advertising and promotion.

{audio:0:03:53} How would you compare closed loop measurement with the multi-touch attribution on Facebook, Google or other walled gardens?

{audio:0:04:04} At a high level, multi-touch attribution modeling measures a range of marketing activities. For the most part, they use multiple regression to understand how each of these independent marketing activities are having an effect on some outcome variable, typically sales or revenue. It is a statistical model which allows you to understand each marketing activity's comparative impact on the outcome variable. It is not as precise or definitive as closed loop measurement, but you can see how the activities compare on a level playing field.

{audio:0:05:08} One of the issues Cardlytics has, even with their closed loop measurement, is that if they offer a three to one return on ad spend, how does that compare to the other activities? They do not have closed loop measurement for all the other activities, so there is no simple way to do a direct apples-to-apples comparison of all the activities.

{audio:0:05:46} So an agency or marketer might spend $100 on Facebook and get multi-touch attribution measurement, but they cannot measure it against Cardlytics' closed loop. Could they see a higher return on Facebook or on a dollar amount that could distort the comparison which makes it hard for marketers to switch dollars to Cardlytics?

{audio:0:06:06} As I said, it is hard to make a direct apples-to-apples comparison. Having said that, in market mixed modeling or attribution modeling, for each activity you do, you will see for every dollar I spend, I get X dollars back. If I was doing display advertising on a particular platform, for every dollar I spend, on average, I got $1.23 back. Cardlytics gave me a five to one return on ad spend, but they are different measurement systems so it is simply an approximation. Companies like Nielsen, where I used to work, have done attribution modeling for years and have built up normative databases about the typical returns on different kinds of activities.

{audio:0:07:07} The normative data they have for TV, display or online video advertising, are less than $2 for every dollar you spend. If I’m a company like Cardlytics and I can prove I can deliver $4 or $5 return on ad spend, that looks good in comparison to what most people get from TV, online video or display advertising.

{audio:0:07:39} How do marketers compare returns with no true apples-to-apples method?

{audio:0:07:47} Some simply make an imperfect comparison. There are market mixed modeling providers who enable you to take closed loop measurement results or A/B testing, and include them in the market mixed modeling as what they call a prior. In some cases, you can do a comparison across things, but only certain companies do that today. It depends on the marketer and who they use as a marketing mix or attribution modeling provider.

{audio:0:08:27} What could improve the situation for Cardlytics?

{audio:0:08:41} Comparison will always be a challenge until Cardlytics allow individual personal level data outside their platform. Banks are uncomfortable with them sharing that data, even when anonymized and no PII is involved. Without doing that, it will stay where it is today. Historically, Cardlytics delivered the offers and measured the sales lift and outcome with their closed loop measurement system. For many marketers, especially those who did not spend large amounts of money on the platform, that was probably okay.

{audio:0:09:34} There were also marketers who, on principle, said, you cannot measure yourself. They could not spend lots of money on your platform, have you do the targeting, deliver the offers and measure yourself. It was not that they did not trust you personally, it was simply that, in principle, that was not the way they wanted to run a business. The same thing is true of the big media platforms. If I am a big brand and I spend a bunch of money on YouTube, there is a third party there to measure what I got, which is typically Nielsen, ComScore or others. One thing we did when I was there was to bring Nielsen in for third-party measurement. If a marketer spent above a certain amount of money on the platform, or if they were a new marketer on the platform, we would use Nielsen as a third-party measurement company to show them what they got, based on measurement by Nielsen who is a trusted third party.

{audio:0:10:33} One argument for why the Cardlytics' closed loop system is so powerful is that it is still not clear the return you can get on Facebook or Google. How do you look at that argument that Facebook and Google are also walled gardens?

{audio:0:10:58} Marketers are trying to measure what they get on Facebook and Google. I am talking beyond search; I’m talking about display and video advertising. In some cases, marketers are able to push their own data into a clean data room provided by the walled garden, do some analytics in combination with the walled garden's data, and bring back out aggregated results. That is complicated to do currently and there are few I am aware of who are doing that. In addition to that, Facebook, YouTube and others have started to allow third parties to measure brand lift, audience delivery or sales impact on their platforms.

{audio:0:11:53} For example, Nielson measures sales lift impact on walled garden platforms for CPG companies. In that sense, you could take the results from that and compare that with Cardlytics, which Nielson have also measured, but you cannot do that across every touch point today.

{audio:0:12:15} Could we walk through a hypothetical process of how advertisers would historically build a campaign on the Cardlytics platform, prior to the new ad platform and manager which they have recently rolled out with US bank? If a restaurant were to spend $100 on the platform, what would be the first step?

{audio:0:12:35} The first step would be sitting down with the marketer and understanding their overall business and goals and what they are trying to accomplish. Typically, the applications fall into three general areas. A restaurant would target those who have never dined there before but live within a certain radius, so those are bringing in all new diners. Cardlytics could find those people based on their past purchase behavior and deliver offers to them. A second application would be more of a loyalty play, where there are occasional diners who dine with competitors, and I want to build dining frequency with existing diners because they spend significant amounts of money with competitors.

{audio:0:13:39} A third broad application is most of these clients have their own first party data and they know they have people who dine there a lot, so they typically treat those important customers the same. Cardlytics can see that, in that group of very heavy diners who dine at your restaurant, you have almost 100% wallet share of a few, who do not use the competitors, or if they do, very infrequently and do not spend much. There is also another group of heavy diners who spend a lot of money outside your restaurant. Those are the people you want to target to try to continue to increase share. There is no reason to spend promotional money against the people who are already coming to you almost 100% of the time.

{audio:0:14:36} Those would be the three basic applications and if you were to do all three applications all year long at full scale on the platform, many advertisers could grow their top line by 2% to 4%. Having one marketing activity to grow top line at 4% at a five to one return on marketing spend is pretty good. Having said that, there are always interesting variations of these, and I will give you a great example. In the retail world, stores are constantly being opened and closed. If a competitor is going out of business, you could go to a retailer and say listen, we can find all shoppers who shop at the locations which are closing and which are also within five miles of one of your stores, and target and incent them to come to your store.

{audio:0:15:35} If you do not do this, that volume will simply distribute on a fair share basis to all the existing players. Whereas if you target them, you will get more than your fair share. That is kind of a vulture strategy. Conquesting brand new diners, making current diners more loyal and finding the heavy diners who still have more head room to spend more with you, are the three core applications.

{audio:0:16:06} This seems like ground breaking targeting capabilities. I don’t know of any other channel or way to specifically target on a user and region level.

{audio:0:16:25} It gets even more interesting. Think about the dining space for a second. The credit card debit card transaction data tells you how much, when and where a person spends. The when they spend is also really important. You could look at that data and group the dining transactions by 6AM to 10AM. Those are people eating breakfast, so you can go to a big casual dining restaurant that is implementing a new breakfast menu, and offer to target all the people who are eating breakfast at their competitors. You can start to target by dining occasion and all kinds of things. There are many variations you can play with, but the core of having the purchase data around who, where, when and how much they are spending, is pretty powerful.

{audio:0:17:29} What is the process of building a campaign on the application today?

{audio:0:17:39} Historically, the client service team worked directly with the marketer to define their goal, the target audience, and how big a campaign they wanted to run for how long. The operations team within Cardlytics would build and launch that campaign. The new self-serve platform Cardlytics has been developing and beginning to roll out, will be more automated. The advertiser, marketer or their agency can login to the platform and do that on a DIY basis.

{audio:0:18:23} How will the new self-serve element change the adoption of Cardlytics?

{audio:0:18:30} Anything that reduces friction and makes it easier for marketers to run campaigns is always a good thing. Cardlytics will evolve from having small rectangular tiles that say 10% cashback at Starbucks or some other merchant, towards display ads or maybe even video ads. They will still have embedded offers but those will start to look more like brand advertising. As that happens, it becomes important to serve agencies who are working on behalf of marketers, who will want to build and integrate campaigns on Cardlytics in combination with other things they are doing for their client.

{audio:0:19:35} Historically, Cardlytics went direct to the end marketer without dealing with agencies. This platform helps redirect some of the activity and focus on agencies who serve marketers.

{audio:0:19:48} What is different for agencies in the way they look at a channel like Cardlytics?

{audio:0:19:56} Purchase-behavior-based marketing platforms are relatively new, at least the way Cardlytics do it. It was always a challenge to get new clients when I was at Cardlytics, and it is about educating clients on what it is and how it works. Historically, it was a cashback offers platform only, which people tend to think of as a bottom of funnel promotional activity, as opposed to a brand building top of funnel marketing capability. They need to see that this is more of an advertising vehicle which can simultaneously build brand and drive sales.

{audio:0:20:53} I saw the demo of the new UI they are trialing with US bank, and it was completely different, with clear tiles and great imagery. How would the new platform change adoption by advertisers or agencies?

{audio:0:21:24} Traditionally, Cardlytics had three different groups who bought the service. One would be the team who run CRM loyalty programs. A second group would be performance marketing, and a third group would be traditional advertising, including digital and media. The largest groups Cardlytics worked with when I was there, was the CRM loyalty group. Having been a CMO and served many other CMOs, those groups do not have gigantic budgets. The challenge Cardlytics always faced was to get clients to use the platform across all the different targeting opportunities, at scale and all year long. The amount of money to be spent was way in excess of what these groups typically had.

{audio:0:22:23} That meant they had to get money elsewhere, which presents a bigger challenge. As Cardlytics starts to change the user interface and deploy traditional display and video advertising, it will be much easier to get money from the performance marketing budget or the advertising and media budget, because they have capabilities which are more apt.

{audio:0:22:51} How are those budgets structured in a typical marketing organization? Are they really siloed between performance, brand and CRM?

{audio:0:23:02} Most clients are fairly siloed and, in most cases, the chief marketing officer was not going to get a lot of extra funding to increase spending on a particular activity. It becomes a budget reallocation, which comes back to the question of measurement. How do you have an apples-to-apples comparison to know that Cardlytics' five to one return on ad spend is better than the three to one on YouTube or the two to one on Facebook?

{audio:0:23:41} So Cardlytics to grow billings from $300 million to three billion, they have to take spend from digital ad budgets and in store promotions? Where will the transition come from?

{audio:0:24:04} Having units which look more like traditional display will make it much easier, because there are typically larger budgets there. For companies with significant promotional spending, it is moving money from promotional activity to the kinds of activity Cardlytics does.

{audio:0:24:27} Does it come back to this measurement question?

{audio:0:24:32} When you sold the platform did you mainly target the CRM and loyalty team? Was there a specific director below the CMO who you dealt with?

{audio:0:24:44} You find your way in wherever there is an opening, but that was the place which was easiest to land and make progress with, because those people already had a view of what customers were doing and how much they were spending on the platform. Cardlytics is really powerful in that you can tell prospects who already know about their current customers, what those customers do outside their store or restaurant, and where the opportunity or challenge is. You can walk in with an educational story and tell the prospect something important about their business which they never knew, which presents either an opportunity or a risk they can take advantage of by using the Cardlytics platform.

{audio:0:25:36} How would you take Cardlytics to the next level?

{audio:0:25:53} There are several things Cardlytics needs to do, most of which they are doing. They need to bring in display and video-like promotional units, which they can get higher pricing for. In the regular world of digital advertising, video gets a much higher CPM than display, and the same thing should be true on the Cardlytics platform. They still have closed loop measurement on those; it is not just branding, so that should drive real sales outcomes. They also need to continue to get more sophisticated about targeting.

{audio:0:26:42} The better you get at targeting, the higher price you can charge. They need to get as granular and as fast as possible, to enable real time location targeting. They need to bring in first and second party data sets and match those to do additional more granular targeting. Cardlytics is tailor made for real time optimization using machine learning or some form of AI. As the offers are being used or not by customers, machine learning should gradually change the value and location of the offer, based on what it learns.

{audio:0:27:46} Historically, the platform was never designed to easily do that, and there was almost no campaign in-flight optimization which happened, so they really need to enable that. Banks allow Cardlytics to provide analytic insights to clients using the platform, but they do not allow Cardlytics to sell those insights to non-clients who do not use the platform. There is a sizable opportunity for Cardlytics to create a self-serve data analytics and insights platform for clients to use to get advanced insights above and beyond some the basic things Cardlytics provides today.

{audio:0:28:41} Nielsen, the company I used to work for, had a multi-billion-dollar business where they were getting purchase data from grocery stores, drug stores and mass merchandisers. They brought it together, standardized it, cleaned it up, and sold it back to CPG companies so that they could measure their business and understand what to do to perform better. That is almost exactly what Cardlytics could do but in different verticals. They could bring all this data together, go back to casual dining or retailers and offer them a full market view, where they are gaining or losing share and potentially some of the factors which are behind that. They could create a SaaS data and analytics business on top of their cashback offer platform.

{audio:0:29:30} Would they have to pay a royalty to the bank for that data given they are effectively integrating with them for the purchase history?

{audio:0:29:40} In almost all the situations with banks, the original conception was a cashback offers program. Anything different would mean Cardlytics would have a conversation with the bank to figure out how this would work in terms of revenue share, but it is a gray area.

{audio:0:30:01} The concept of having purchase history and the closed loop measurement seems so purely logical that it should work in the long run, but what are the risks that it does not scale?

{audio:0:30:31} The biggest risk they face is that the banks view the user interface on their app and website as their own real estate, which it is. Cardlytics depends on that real estate to run their program, and the banks are not particularly fast at upgrading their own user interfaces. They typically do it a few times a year, but they have priorities other than cashback programs. Cardlytics is always at the mercy of the banks in terms of them doing what Cardlytics wants.

{audio:0:31:28} The banks are Cardlytics' biggest advantage because of the scale they have with all the banks and it is a two-sided network. The more banks, the more audience you have and the more attractive you are to marketers. But they are also the biggest disadvantage to Cardlytics and that is one of the reasons you see them acquire other cashback offers programs. That provides them more flexibility about where they could go longer term if they do not make progress with the bank partners they need to.

{audio:0:32:18} How does this relationship with banks change at scale? Cardlytics currently has 50% of all card transactions in the US, but how does the dynamic change when it reaches 75% or 80%?

{audio:0:32:31} They are currently as big as they need to be in the US. There are only three left, Citibank, Cap One and American Express, but with 160 million monthly average users, they are in the same territory as YouTube and Facebook. They will continue to acquire more bank partners and get an even bigger audience, but they are big enough. What is more interesting to me is that they have three mega banks and a bunch of regionals. There is Bank of America, Chase, Wells Fargo and everybody else. In that environment you will likely see those three banks exerting influence on where they want the program to go, and you may end up with three different approaches to the way things are done, at least on the surface.

{audio:0:33:35} If I were one of these big banks, I would be thinking about this as not just a cashback offers program but, because Cardlytics have all this data, how could it be used in a privacy compliant way for me as a bank to achieve some my goals, whether that is bringing in new or retaining customers, being smarter about the way I do certain things as a bank. The banks were beginning to think about this when I was there. It was mostly future oriented and only thoughts at the time, but you could start to see each of the banks formulating their own views of what they might do with this program over time beyond cashback offers. It will be interesting to see where that goes.

{audio:0:34:26} This is where it differs from Facebook, Google or YouTube who own their own real estate and the demand, whereas Cardlytics is built on the bank's real estate. As this scales, will banks continue to leverage their power and direct where this goes?

{audio:0:34:55} The three largest banks will continue to leverage their size and scale, but the risk of any of them leaving the network is relatively low. Cardlytics have never lost a bank.

{audio:0:35:21} What was the biggest challenge in driving adoption with the agencies?

{audio:0:35:32} When I was there, we were not focused on agencies for the reasons we covered, because most of them were focused on advertising and media and we never had an offering which met that need, nor did we have a good platform to service them. We were more focused on the end marketer, and I could talk about some of those challenges. One which we've talked about was education, what it is, how it works and why it works. The second challenge was proving it works. Most marketers who have run promotion events around a certain date, tend to see a big bump in their business because of whatever the activity is. You are running along, you see this bump and then it goes back down again.

{audio:0:36:30} With the Cardlytics system you are running the offer across a period of time and consumers are seeing the offer at different times. The incrementality of sales is spread thin. It was very interesting because the first time marketers used the platform, they felt nothing. With Cardlytics, the results are spread out over the four-month period you run the campaign, so educating clients around how the system and promotion works while demonstrating the return on marketing spend was a big one, and third-party measurement was key.

{audio:0:37:27} So advertisers not get the dopamine hit they do from other channels?

{audio:0:37:35} Yes, because it looks like a bottom of funnel promotion so that is what they expect. I heard this great quote which was “effective advertising is like how the grass grows in your yard. You never see it grow but you have to mow the yard periodically.” It is the same thing; you do not see it happening but, over time, it works. Training clients that it does not work like a promotion but more like advertising and media, was the key thing we needed to do.

{audio:0:38:14} How did you navigate the measurement question?

{audio:0:38:19} Closed loop measurement was demonstrating, through our data science team, as to exactly how and why it is done the way it is and how we chose the test and hold out groups. If there was still doubt, we also had third-party measurement available from Nielsen.

{audio:0:38:39} Did having Nielsen make a big difference?

{audio:0:38:42} It made a difference because some customers would not use the platform without third-party measurement, whereas others were okay spending some money on the platform, but they would spend $400,000 a year. Those who spent $40 million a year wanted third-party measurement. There were also others who simply believed us.

{audio:0:39:27} With the transition of Cardlytics from a purely cash rewards player to more of a complete ad network with display, video and other paid media, how does the sales process evolve so that Cardlytics can sell into different budgets of a marketer's unit, away from loyalty into the digital ad or performance marketing spend?

{audio:0:40:01} There are several parts to this. One, is educating the advertising and media group at the client that Cardlytics is more than a bottom of funnel promotional activity, that there are newer units which are more brand building in addition to driving sales. Two, is having the platform for agencies to be able to use, plan and activate off the Cardlytics platform. Three, is enabling agencies to programmatically use this as part of their standard media planning tools. Currently, Cardlytics does not allow individual level data to travel outside their platform, so there is no easy way for agencies to buy and activate programmatically with Cardlytics; it is still a standalone platform.

{audio:0:41:04} I could programmatically plan and execute a digital campaign across multiple sites and see what reach and frequency I have in total and across various digital platforms, but Cardlytics would be separate from that, because you cannot share the individual ID to the platforms to match them. You cannot say the people who saw my ad on Facebook also got an offer on the Cardlytics platform, because there is no way of connecting the two. That will be a challenge for them, because even though the agency has access to the Cardlytics platform and can build and run campaigns, they cannot integrate it with other campaigns they are running.

{audio:0:41:55} Is that a major barrier to agencies adopting Cardlytics?

{audio:0:42:00} It is an issue because most marketers and agencies want to drive the highest level of unduplicated reach against their target audience. If you cannot measure Cardlytics' reach in combination with the other sites you advertise on, you cannot know whether that reach is net incremental or duplication. What Cardlytics do different to the other guys is they have offers embedded in. They will never do advertising in the bank channels, because the banks will never agree to that, nor should they. It will always be a hybrid of offers embedded in display and video, plus the closed loop measurement is different enough that agencies treat it separately. It becomes more like the walled gardens who do not share individual ID level data outside their platforms, but are big and powerful enough that they can do, and people put up with it.

{audio:0:43:12} Do they need to build a differentiated enough solution that still draws the agencies, even if they cannot integrate it with the other data they have?

{audio:0:43:25} Yes, and that is where having a separate analytics platform with insights becomes powerful because you can only have that if you are a client using the platform and spending over a certain threshold. Banks will never allow you to monetize that to non-clients.

{audio:0:43:48} Is there any way the banks could allow the integration of that data?

{audio:0:43:54} Banks could easily be comfortable with you sending individual ID level outside the platform, but it becomes a question of how much risk they are willing to tolerate. Their concern is the remote possibility that someone reverse engineers the IDs to figure out who the people are, and it is also a public relations thing. The last thing they want is for clients to understand they are selling their credit card transaction data and making money from it. They are on very solid ground by allowing customers to opt out of the program if they want to.

{audio:0:44:47} If they stay in, Cardlytics will anonymously use your credit and debit card transaction data to deliver cashback offers which are relevant to you. There is a give and take; you give up access to this data, but you also receive offers so the banks feel that is completely reasonable, especially when customers have the ability to opt out if they want to. But when you start to take their data and let other entities do things with it on an individual rather than an aggregate level, they get a lot less comfortable about that.

{audio:0:45:32} What are the economics of a dollar of spend on the Cardlytics platform?

{audio:0:45:57} There are several pricing models. The dominant model which I believe they still use is they charge the advertiser or marketer a percentage of the sales which happen among the exposed offers during the campaign. Let us use a simple example. If a marketer runs a campaign and across the campaign period, among all the consumers who saw the offer, they spent $100 with the client. Cardlytics would typically take 3% to 12% as a fee from the marketer. It depends on the category and other variables, but the customers who saw the offer during the campaign spent $100, you pay me $5 as a marketer. The marketer does not pay for the cashback offer, Cardlytics pay for that and a revenue share to the bank and keep the balance.

{audio:0:47:07} Do advertisers not lay out any dollars?

{audio:0:47:12} Not until they pay the fee based on the percentage of sales which happened during the campaign among exposed viewers.

{audio:0:47:19} So you set your target audience and budget and Cardlytics run the offers. The customer goes into the store, spends the money. Cardlytics debits the cashback to the consumer and the bank's share of that, and also collects the fee from the advertiser based on the result. Is the percentage set based on the quality of conversion?

{audio:0:48:10} Yes. It depends on the target audience, but the highest return on ad spend is almost always among heavy diners or shoppers, because they are inclined to use you anyway. The lowest return ad spend are the people who have never shopped or dined with you before. It typically starts with what the desired return on ad spend is, then you work back. It is also about the internal margins Cardlytics are looking to deliver in their business model.

{audio:0:48:46} Is that return a revenue number? If I spend a dollar on advertising, I want $5 in revenue?

{audio:0:48:56} Yes, for every dollar you spent, you got $5 back in top line sales or revenue.

{audio:0:49:03} How do you optimize consumer incentives at this stage for Cardlytics?

{audio:0:49:08} That was the first question I had when I went to work there; what is the optimal offer? The short answer is that Cardlytics have run many campaigns and have a good idea what the optimal offer amount needs to be; they work from past experience. Having said that, as they move towards machine learning driven optimization approaches, offers will change over time. Not necessarily to the same consumer because the banks would have a problem if today, I see a 10% offer from Starbucks and tomorrow it was down to 5%. You could see the reverse and you could also imagine campaigns where a certain group of people are not initially targeted.

{audio:0:50:08} The machine learning calculates this group responds well, and targets Randall with an adjusted offer. The machine learning approach allows for optimization at scale which humans could never do. The incremental return on ad spend or the lower cost to Cardlytics by optimizing offer value will be significant.

{audio:0:50:42} Shouldn’t Cardlytics just maximize consumer incentive to drive engagement at this point?

{audio:0:50:50} When Chase started with Cardlytics, they created their own fund to top up offers. If it was a 10% offer, they might top it up to 15% or 20% in the initial period, to get people engaged with the program. It would not be surprising to see something like that.

{audio:0:51:22} How willing are the banks to share that cost?

{audio:0:51:26} Generally, they are not that willing to share it; Chase was an exception to the rule. The economics of a debit card are thin, so there are few rewards programs attached to them. They are attached to credit cards because the economics are much better than debit cards. Banks looked at Cardlytics as a way to deliver a value-added program to debit card customers that they otherwise would not have, and in addition to that, could get a revenue share from it. In the big scheme of things for the bank it was de minimis, but it was still a revenue share. They were getting paid to do this as opposed to paying from their own pocket. Banks initially saw it as a win-win.

{audio:0:52:51} How can they optimize or improve conversion of ad units?

{audio:0:53:09} There are several factors which have a significant impact on engagement and we always tried to actively manage and improve those, because you want the most engagement possible across consumers using the platform. The first one was the user interface. The easier and more visible the program was to use and experience, the better engagement you had. Every bank had their own user interface, and Cardlytics could do analytics across all bank UIs, hold everything else constant and see which performed best and why. Part of their job was to get banks to move towards the UIs which performed the best.

{audio:0:54:09} This is a balance because they never want their UI to look exactly like other banks. There are certain aspects of cashback offers programs we knew drove the best engagement. You want to incorporate and keep those things as you upgrade your UI, while keep other parts different to your competitors.