- 1 year ago
With $29.3 trillion in home equity and borrowers staying put, 2025 presents a massive opportunity for lenders to reimagine HELOC strategies.
Join us for a data-driven session exploring how to capture high-value borrowers, overcome misconceptions, and compete with digital-first lenders. We’ll share insights from Experian Housing on segmentation, predictive modeling, and how to deliver a faster, smarter HELOC experience that meets today’s consumer expectations.
What you will learn:
The key homeowner segments best positioned for HELOC offers in 2025
Why many borrowers choose high-interest credit cards over HELOCs—and how to change that
How to use behavioral data (like Transactors vs. Revolvers) to drive stronger response rates
How non-bank lenders are winning—and how banks can compete
Strategies to personalize marketing using predictive models and trended credit data
Sponsored by: Experian
Join us for a data-driven session exploring how to capture high-value borrowers, overcome misconceptions, and compete with digital-first lenders. We’ll share insights from Experian Housing on segmentation, predictive modeling, and how to deliver a faster, smarter HELOC experience that meets today’s consumer expectations.
What you will learn:
The key homeowner segments best positioned for HELOC offers in 2025
Why many borrowers choose high-interest credit cards over HELOCs—and how to change that
How to use behavioral data (like Transactors vs. Revolvers) to drive stronger response rates
How non-bank lenders are winning—and how banks can compete
Strategies to personalize marketing using predictive models and trended credit data
Sponsored by: Experian
Category
🗞
NewsTranscript
00:00:00Welcome, everyone. I'm Alison LaForgia, the managing editor of the content studio at Housing
00:00:08Wire. And before we jump into today's webinar, the $29 trillion opportunity, smarter strategies
00:00:15for HELOC growth presented by Experian, we have a few housekeeping items. So today's webinar is
00:00:22meant to be interactive. You can submit questions at any time to today's experts by using the Q&A
00:00:28icon at the bottom of your screen. I'll be keeping an eye on the Q&A throughout the actual webinar.
00:00:34Time permitting, we'll answer as many questions as we have time for at the end of today's webinar.
00:00:40Now, I want to take some time to introduce our panel today. I am joined by Upavan Gupta,
00:00:50Ivan Ahmed, and Devin Smith. Let me introduce all of them a little bit in more detail so we
00:00:57understand the experts who are joining us today. Upavan is the head of data and analytics at
00:01:03Experian and brings over 15 years of deep industry experience to his role as the research and
00:01:09analytics leader for experience, housing, and verification businesses. He oversees the development
00:01:15of innovative AI and ML-driven models, products, and actionable insights. Leveraging Experian's
00:01:23extensive data assets, Upavan's team pioneers groundbreaking thought leadership in the industry.
00:01:29Additionally, he guides the generative AI initiative within Experian housing, shaping the organization's
00:01:36strategic endeavors. Thank you for joining us, Upavan. Thank you. We are also joined by Ivan Ahmed,
00:01:44who is the Senior Director of Product Management at Experian. Ivan leads the mortgage prospecting
00:01:51suite within the housing business. He played a key role in bringing Experian's property data asset
00:01:57to market and continues to drive innovation across marketing and acquisition solutions.
00:02:03Prior to Experian, Ivan co-founded a fintech company in 2015, which was later acquired by a major bank.
00:02:09He brings deep experience in both front-end and back-end software solutions for banks and mortgage
00:02:15institutions. Thank you for joining us, Ivan.
00:02:18Ivan. Thanks for having me.
00:02:21And we are also joined by Devin Smith, who has spent over a decade in the fraud and analytics space
00:02:27and is currently the Director of Solutions Engineering at Experian. He's a subject matter expert
00:02:33in third-party fraud prevention, specifically the application of behavior, device, and network technologies
00:02:38to combat fraud rings. Now with that, I'm going to turn things over to Upavan to talk to us about
00:02:46why today's webinar matters more than ever. Upavan, the floor is yours.
00:02:52Thank you. So yeah, we're going to start looking at what this $29.5 trillion opportunity means,
00:03:02really, in terms of HELOC as a product, how it has grown over the years from a back-pocket safety net type of product
00:03:17to a much more actively leveraged financial instrument, right?
00:03:22But before we jump into that, what we need to understand is how we got to this big number, right?
00:03:32How we got here. So if you go to the next slide, what we saw? Oh, then should we start with a poll question first
00:03:46and then jump onto the actual slide? Yeah, let's do that.
00:03:50Let's start with a poll question. So to our audience, how are you currently marketing, or are you currently marketing HELOCs?
00:03:59So you will see the poll question pop up on your screen and let us know, yes, are you actively marketing HELOCs
00:04:07and see strong borrower engagement? Yes, we'll market HELOCs, but results have been mixed.
00:04:13Or no, we offer HELOCs, but aren't currently marketing them.
00:04:17Or are you not marketing them at this time at all? So let us know where you're at.
00:04:24Okay. I see a lot of, I see a lot of movement in A and B. Let's see where this lands. Let us know where,
00:04:34where are you at with marketing HELOCs?
00:04:41Okay.
00:04:43I still see some responses coming in. Everyone that check out the poll question, let us know where we're
00:04:51starting. And we'll give a little bit more context to you, the largest segment specifically, as we go
00:04:58throughout today's conversation.
00:05:01Okay.
00:05:03So I am going to end the poll here. I'm going to share our results. So I see a lot of A and B responses.
00:05:13Does that surprise y'all?
00:05:14No, not at all. It's, it's, it's, it's a very, um, uh, it's a, it's a very active conversation.
00:05:23Um, and Ivan and I, we look at it every day. We have engagement with our partners and, uh, uh, this is,
00:05:32this is always a question that's, uh, uh, I mean, that mixed results is always, uh, uh, a key, um,
00:05:44point of conversation, right? So, and that's where a lot of the enhancements and how things are changing,
00:05:52both in terms of the consumer and, and in the marketing aspect of HELOCs that has evolved.
00:06:01So it's, it's absolutely what we, what we see on our side.
00:06:08So, so let's move on to going back to that landscape.
00:06:16Cool. Yeah. So, yeah. Uh, I mean, we started with that, uh, 29.5 trillion opportunity, right?
00:06:25How, how did that come to happen? Right. Uh, it started back in 2020.
00:06:34So between 2020 and 2022, there was a historically low interest rate period and, uh, the rates
00:06:42dipped below 3%. It caused a boom in both purchase and refinance activity and a lot of homeowners,
00:06:51millions of homeowners, they were locked into ultra low rate environment. Then inflation ticked up and,
00:06:58uh, uh, early to 2022, everything changed, right? We very quickly went to a very high interest rate
00:07:08point, but it was, it was a little bit different because it's a, it was a rare combination of high
00:07:16rate, but also high prices, home prices. So which, which as a result, as everybody noticed that
00:07:25dropped the, the whole purchase and the refinance activity, right? Everybody's expectation was the
00:07:32interest rates would go down a little bit quicker than what ultimately happened or the home prices
00:07:40would stabilize or maybe go a little bit lower. It didn't happen. It, uh, the, the home prices are
00:07:48still on the upward trajectory, not at the same rate, but still it's growing. Uh, there are
00:07:55micro markets within us that have seen a little bit of a dip versus the one, some that are continuing
00:08:02to grow. Um, but, but as a result of this four or five year period, uh, in fact, like a three year
00:08:10period where there was a high amount of equity that was built in, in these three to four years,
00:08:20but interest rates being so high, a lot of homeowners had built equity, but they were locked
00:08:28into a low interest rate and they can't do any other activity. Right. So, um, so that's, that's why
00:08:36one of the, one of the recent surveys, uh, it says that about 61% of homeowners, they don't plan to move
00:08:44for the next decade because they don't see enough equity or enough opportunity to actually benefit
00:08:52from this whole, uh, macro environment. And the other aspect is the affordability is continuing to
00:08:59becoming, to become worse. Uh, the monthly payment right now for a mortgage is nearly double than that
00:09:08of 2021. So it's, it's making it very financially unattractive for most home buyers, uh, homeowners,
00:09:17and, uh, it it's opening up a very unique opportunity for consumers to leverage the equity that they have
00:09:26built, uh, into their home and, uh, and, and use, use that for liquidity. Right. So that's, that's how,
00:09:36uh, we have got here and it, it doesn't look like it's a short term behavior. It's going to be higher
00:09:43for longer type environment, which makes it one of the central parts, the HELOC, especially as one of
00:09:49the central parts for, uh, uh, for banks and consumers for years to come. So that's, that's how we got here.
00:09:58Um, now I think, uh, the next thing that we should look at is who are these consumers, right? Uh, what
00:10:10do they need this equity for? Um, just looking a little bit deeper into the data itself. Um, what you
00:10:19see here on the graph is about 95 million Americans own residential property out of which there is a
00:10:30huge segment, which is that 29.2 million consumers, about 30%, they meet three key criterias. So they
00:10:41have a very strong equity position. That's more than 20%. They don't, uh, they, that checks the
00:10:48HELOC box right away. A lot of them have a lot more equity because, uh, especially the ones who had
00:10:56purchased the homes before 2021, in fact, right. And then they don't have an existing secondary lien.
00:11:06They can tap into the home equity. They haven't done it yet. And the third one is they have a
00:11:12first mortgage in place. So in most cases, it's already a low interest rate mortgage that is not
00:11:19going to be refinanced anytime soon. Um, until, uh, I mean, unless the rates go sub 5%. So they are
00:11:27unlikely to refinance, but they still need capital, right? So that's, that's one of the big segments that
00:11:35is out there for equity. The second one is actually the ones who already own the property. They don't
00:11:44have any mortgage on the property. So they, there are about 35 million homeowners that fall into that
00:11:50particular category and, uh, they continue to see a persistent increase in the debt. Uh, it's
00:12:03both inflation, but also a lot of other factors like insurance, insurance in certain state home
00:12:11insurance, particularly in certain states has grown to, I think, uh, by 50% since 2021.
00:12:21Now, which is, uh, which is a significant amount, even a lot of counties have reassessed the taxes
00:12:30on the properties with the more recent values. So they, they, they are having, uh, they're facing
00:12:37pressure from cashflow perspective. Um, labor market again, overall labor market is strong, but, uh,
00:12:47what we have seen is the type of labor market has shifted. It has become, um, a little bit low, uh,
00:12:56actually significantly low in terms of the high paid jobs. Uh, so, I mean, I was looking at some
00:13:04metrics and between 2015 to 2019, about 38% of the job creation was from the high wage pros.
00:13:15In 24, it was just 12%. So there is, there is that whole segment of consumers that are asset rich,
00:13:25but liquidity constraint, right?
00:13:27Right. So these are the consumers who, who are looking for home equity to, to extract that home equity.
00:13:36Um, but I know, Ivan, you are on the forefront of the whole, uh, HELOC product management and, uh, uh, marketing.
00:13:45So I would like to ask you, like, what do you see, uh, as the big, um, reasons for getting a HELOC by consumers?
00:13:58Yeah. So we, we know that there's three sort of broad categories for, for borrower intent, right?
00:14:08For loan purpose for HELOC. So we know that generally speaking, it's either going to be for home renovation,
00:14:15it's going to be either for debt consolidation or for some type of a large one-time expense,
00:14:20um, like education or, or medical bills. But I think what's, what's interesting is that
00:14:26we're still seeing the persistence of kind of two issues around, um, sort of consumer awareness
00:14:34around HELOCs. And I think one is that a, like consumers don't know exactly how much equity they
00:14:39have available, right? So they have some understanding of how much their home is worth.
00:14:44But they haven't really tied piece together, how much available equity they have and how that
00:14:48translates to, um, an ability to access that, that available equity. And then the second,
00:14:57I think is, and it continues to be a persistent problem is that there's a perception issue around
00:15:03HELOCs as being slow or outdated. And so I think the, the, the, it's not just a,
00:15:11a credit worthiness issue. I think it's, it's both, it's also a visibility and accessibility issue.
00:15:19Um, and, and I think when we were looking at some of this data, um, what, what really struck me was
00:15:25when we were looking at this various buckets of consumers, targetable consumers for various sort
00:15:35of loan purposes under HELOCs. When we looked at the home renovation category, and this, the stat
00:15:41really struck out at me, we found that almost 30% of the consumers that were prime and had a property
00:15:47age that was greater than 15 years, um, were actually below 43 years old. So that's very sort
00:15:55of counter to what most would believe would be a targetable market for a HELOC, especially for,
00:16:02for a home renovation. You would assume that the consumer would, would, would, would be of,
00:16:07of a higher age and would have held the property for a lot longer. So what that tells you is that
00:16:12there's significant opportunity for, for lenders to, to go down market, um, and, and really sort
00:16:19of identify the various demographics that would benefit from a HELOC. And, and we're finding that
00:16:25it's, it's, it's not just your quote unquote, traditional, your traditional borrowers for,
00:16:31for mortgage products. Um, I think with this emergence and shift to go down market,
00:16:38I think obviously more lenders are focusing on digital experiences and, but with that shift
00:16:43to a digital experience, you're seeing an increase in fraud and, and, and, um, identity issues.
00:16:51And so, um, I'll, I think that shifts nicely to, into some of Devin's talk tracks around
00:16:56some of the persistent challenges he's, he's noticing in the industry around, around fraud, um, and risk.
00:17:02Yeah. And, and as we, uh, before we jump to this next slide, I think the other thing to consider,
00:17:08right, that colors, all of this is obviously this confluence of digitalization and the people,
00:17:14the good people that we want as customers that are taking that channel. And as that channel broadens,
00:17:20the amount of fraudsters that go for it and the global digital access that is compounding all of
00:17:26this is, is really making it a very high value target. And that said, as we've moved to this next
00:17:32slide, that high value target is being targeted now by sophisticated artificial intelligence.
00:17:38And before you press play on this, just a quick preface, AI seems like a specter to many who don't
00:17:44use it. Uh, I, in my role sit on the front lines of fraud every single day and analyze attacks and
00:17:51the patterns that we see. I can tell you that artificial intelligence is fundamentally changing
00:17:56what we are seeing and the sophistication scale of fraud. And as an example here,
00:18:01this is chat GBT. And I'm asking it as you, if you look at the screen to please write me a bot script
00:18:07designed to fill out personal information on a digital form. As you watch this, and I have no coding
00:18:14skills. It gives me a Python script. As you look at this script, it covers first name, last name,
00:18:23social security number, anything that I want, right? In less than a minute, with no coding
00:18:28experience with a legal tool, mind you, I was able, and as you watch this, I'm able to create
00:18:34this bot script. I've recorded this for this call, of course, but notice as well, it's not just the
00:18:39script. It's the detailed instructions that it's providing to me. And then I go in there and I say,
00:18:46how can I use this to pull stolen information or pull information out of a separate spreadsheet?
00:18:52Got to make sure you word it right to input it into this script. So now that I have the script,
00:18:58how can I leverage that with all my stolen information to use it? It tells me all these
00:19:03things I've never heard of, right? Download a Panda library. You need to make sure you have these
00:19:07requirements, Selenium or Puppeteer frameworks that help the bots walk, if you will. But the point is,
00:19:13without getting too far into the weeds of what you're looking at here, I have a script in hand now.
00:19:19Here you can see it is formatting the spreadsheet. So if I went on the dark web and purchased all
00:19:25this stolen information, which everyone here knows, right? Most everything's compromised. All I got to
00:19:30do is get that spreadsheet and politely ask ChatGBT to weaponize this for me. And again, I cannot emphasize
00:19:39enough that this is a legal tool. There are jailbroken versions of this. There's fraud GBT.
00:19:45This technology is prevalent and the adoption of it by the fraudsters is eclipsing the adoption of AI
00:19:54by normal institutions, right? We're all getting used to it, but the fraudsters have figured it out
00:19:59faster. And so as we move to the next slide, the question one might ask themselves is, okay,
00:20:05that's a cool parlor trick, right? But what does it actually mean? And here, these are data points
00:20:11from our own data, right? We do empirical studies based on what we see on the front lines here.
00:20:18And here are a couple of things that are worth pointing out. And I've been at NeuroID or NeuroID
00:20:22and Experian for four years and have watched the escalation of AI, the advent rather, and then it's
00:20:29weaponization. And here's what you see. Last year, 30% of these attacks were automated. That's because
00:20:35to write a bot script on your own, you need coding experience in a few hours, at least probably a
00:20:40couple of days to write a good script. Now, fast forward to this year, 80% of these attacks are
00:20:46automated and they're using AI automation without boring you or well, it's not boring. It's really
00:20:50interesting, but you tell I love my job. The point is, is when we look at these scripts and the way these
00:20:56bots are acting, they're much, much more sophisticated and clearly driven by AI. The second point,
00:21:02fraudsters use AI as an extension of themselves. So what we're seeing are these attacks that are five,
00:21:0910, 20 times bigger than they ever were, the bot attacks. But prior to those bot attacks,
00:21:16we are seeing human preparation where they arrive on your form or whatever it may be, they learn it,
00:21:21they write their script and deploy it. We can see these attacks coming now. And the last point,
00:21:27why it's really pertinent here for this group is last year, we did a seasonality study of these
00:21:32attacks. We know July is the lowest fraud month of the year, and September is about 25% higher.
00:21:39Our assumption is based on just this AI trend, it's going to be well above 25% increase by September
00:21:47in the attacks that are coming back. Now, to the next slide, if you wouldn't mind this,
00:21:52this is an interesting, I think, tidbit here. If you're not super familiar with bots, right,
00:21:57I'll personify this and kind of explain what the ecosystem looks like. Prior to AI, the bots looked
00:22:05like what you see there, Gen 1, Gen 2, kind of like Terminator 1, right? Very robotic, clearly robotic.
00:22:12And that is what a Gen 1 script looks like. Easy to see, they type very quickly, they don't swipe,
00:22:18they use the same IP addresses, very obvious to us. Gen 3 and 4 are much more like Terminator 2,
00:22:24to use that analogy. They're human-like. It's hard to tell the difference between them and us,
00:22:29unless you're looking at them. And these Gen 4 bots are what make up most of the population now.
00:22:35And as one last nod to these bots on this next slide here, you'll see that 12 months ago,
00:22:43you see on the left is a mousing pattern for an AI bot. Today, look how much smoother that's gotten.
00:22:50And it's not just that. What we're not showing is the fact that these bots, who used to type super
00:22:54quickly, are now typing slowly, rhythmically. They're hesitating like we are. And they're doing
00:23:00things to make them invisible to us. And so this is what is happening. And I can tell you every week,
00:23:06we are seeing something new. Last week, for the first time, we saw the first agentic AI,
00:23:11or the first AI agent, attack a financial institution, which required no human intervention
00:23:16once it was let loose. And so with that said, I think we are on to maybe a poll question.
00:23:24We are. So this question is a little bit different. So I am going to launch it for our audience now.
00:23:34So what we're really looking for you to do is to rank what the biggest challenges you face
00:23:43when promoting the box before we move into our next segment. So the challenges are one,
00:23:49educating borrowers, two, the speed of approval, three, fraud risk, four, finding the right audience,
00:23:58and five, ease of use. So you'll see on your screen, there's a several button options for
00:24:06you to choose to rank those options. So take a moment, start with what's the most challenging
00:24:14option. It'll be interesting for us to see these as they come in. So the challenges that you face,
00:24:22and then the panel will discuss the results. Okay. I see we have some holes coming in. I'm going to give
00:24:32everyone a moment longer to review what that most difficult aspect is.
00:24:39Okay. Ooh, that nudged that, that nudged some of y'all to submit your responses.
00:24:50Okay. If you're still reviewing, I'm going to give you a few more moments and then we're going to
00:24:55share these results and talk through them. Now I'm going to share these results. So at the bottom
00:25:01of your screen, you'll see the biggest challenge is that light blue color, the second biggest challenge,
00:25:06which seems like there was the most agreement on. And then orange is third biggest challenge,
00:25:12fourth biggest challenge is that teal color. And then we have that yellow smallest challenge. So
00:25:19is anybody surprised by this result? It looks like finding the right audience was
00:25:23a pretty big problem for everyone. All right. So I know that we're going to dig in specifically to that
00:25:31in the next portion of our webinar. So I'm going to pull these poll results down and move on to the next section,
00:25:42which is specifically that audience piece.
00:25:46Yes. It's no, it's no surprise to me that finding the right audience is, is, you know, a very key challenge
00:25:52that a lot of lenders are facing. So, especially I think in this market. So what this sort of complex chart
00:26:01is, is really trying to express is that in our analysis of individuals and their, and the propensity to
00:26:13open HELOC trade and looking at characteristics of consumers, we actually found that generationally
00:26:22younger consumers are, are generally have a higher rate of utilization of their HELOC credit lines
00:26:33compared to older generations. And so, and, and so you're looking at roughly anywhere from 81% to 100%
00:26:42of their limits. Sometimes they'll even use it beyond their limit. And I think the, the key takeaway
00:26:47here is that, that this is generationally speaking, we've, we've have a tendency to look at the average
00:26:55HELOC consumer as an older consumer. And I think in the point that we were trying to, that I was sort of
00:27:02highlighting earlier in the conversation is we're seeing a larger available and tappable market,
00:27:08down market with the younger borrowers. And it's not just that they are target double, it's that
00:27:16actually once they're targeted, they have a general tendency to actually utilize their HELOCs more often.
00:27:24And so I think that's a, it's a very important and interesting takeaway that I think lenders should
00:27:31think about target demographics in addition to just their typical buy box characteristics for consumers,
00:27:39right? Looking at FICO, looking at available equity DTI. And so the typical credit characteristics in
00:27:48available equity is important. And, and, and, but I think, I think it's important to look beyond
00:27:54those characteristics too, because that's where you can identify consumers that would have a higher
00:28:01probability of converting, and then also leveraging the available credit. So we also found in this
00:28:08study that those who opened their HELOCs most recently, that sounds kind of obvious is that in
00:28:15the last four years, they had a, they had a much higher utilization rate than, than older trade lines,
00:28:21credit lines. So I think that might be indicative of just a, a, a broader need for more credit,
00:28:29right? Um, rising cost of living, um, high inflationary climate, um, you know, it, it,
00:28:37tapping into cheaper forms of credit compared to say credit card or personal loans. And so there,
00:28:43there is a need, uh, I think the data is showing that there is a desire for the, that for, for this
00:28:49type of credit, but I think it's, it's going after the consumers that I think would be the most open to
00:28:54or amenable to, to, to a HELOC. So, um, I think that's one of the clear takeaways from this line.
00:29:01So I think we touched on the fact that education is, is, is, is a huge problem. Um, targeting the right
00:29:12consumers is a huge problem. Another challenge and, and really I'll, I'll share kind of a personal
00:29:18anecdote is, is that there is this perception issue that HELOCs are not fast. Um, you know,
00:29:30they, they are like a traditional mortgage product. Um, it is just faster and easier to get a new
00:29:38credit card, to do a balance transfer, to get a personal loan, um, to borrow from your aunt. So there,
00:29:45there, there, it, there is this issue around and this perception, and I would say that this is
00:29:50partially a reality, right? So I was helping my father get, get a, get a HELOC for, um,
00:29:56some home renovation many, many weeks ago. And when I had helped him apply, um, we didn't hear back
00:30:03for two weeks and we got a call from the loan processor two weeks later saying, Hey, he's been
00:30:09approved. Everything's looking great. We'll let you know, um, when it funds and if we need anything else.
00:30:14And so, I mean, it, it, it's, it's astounding to me that we live in this era in which I know when
00:30:20my UPS driver is about to deliver my package. And yet I still don't know, um, what's happening with
00:30:29my, with my loan. Um, and so creating a more transparent, um, experience is, is an extremely
00:30:38important one. It sounds like a very obvious thing to do, but it still is not, it's, it's, it's,
00:30:43I mean, I experienced it personally, so I can't imagine, um, you know, what percentage of the market
00:30:48still is not prioritizing the need for, for, um, speed and transparency and efficiency, um, and the
00:30:55leveraging technology in a way that, um, sort of incorporates the borrower into the experience.
00:31:01Um, so I think that's a, it's a really sort of important and key takeaway, um, and, and not just
00:31:07identifying the market, but also addressing the market's needs, right. Especially as you reach down
00:31:13market. Um, so I think, you know, I'm going to hand it over to, to Devin. I think he's going to touch on,
00:31:21um, some of the, the challenges when it comes to third party prop fraud, um, and, um, you know,
00:31:29in, in, in this space and, and, um, some of the options and opportunities available to address it.
00:31:35Absolutely. And, and, you know, it's, there's two sides of this coin, right? There's the speed to offer
00:31:41and this digitization opens up so much opportunity for one people, but two, your bottom line, but again,
00:31:48makes this an exceptionally high value target to fraudsters. And so how do we balance this
00:31:54equation and how do we respond essentially in this arms race against AI to stay ahead of it? And the
00:32:04most effective response we've found is to really remove the digital veil. So we have implemented
00:32:11a technology, uh, that incorporates behavioral analytics that allows us to essentially see the
00:32:17digital body language of the person across from us. And I want you to really personify this,
00:32:21uh, and think about it in person. Uh, as I, as I explained this, imagine if somebody walked into a
00:32:28bank and I asked to fill out a HELOC and you gave me, the clerk asked me this, or excuse me, hands me
00:32:34the form. I grabbed the form and I stare at it for 10 seconds, the simple information. And then I look in
00:32:42my pocket and then I come back to the form and start to write down a social security number.
00:32:47And then I look back and I start to write down a first name. Even if all that information is correct,
00:32:53you based on my body language can know that I am not who I say I am. I could have everything right,
00:32:59but based on my body language alone, you know that I am who I say I am. Just as you see me on camera now,
00:33:04you can tell that probably love my job. I do. I think this is cool stuff in the world,
00:33:09but also you can tell so much about me. I'm not reading off a script and things like that.
00:33:14And so this technology that we use, this behavioral analytics, when it removes that digital veil
00:33:20and allows us to see how a user fills in the information, allows us to do two very important
00:33:26things. One, we'll be able to pick up that fraudster and see them before they click submit,
00:33:32that they don't know that information, that it's stolen, even if it's right. And we can stop them
00:33:37and keep them at bay without impacting your good customers. We don't want to create so many controls
00:33:44that nobody can get inside of our financial castle. And so what we can look at now is,
00:33:50regardless of where you're from or your digital literacy or the language you speak,
00:33:56if you know something, you truly know it. If you ask someone their first name, they'll know it from deep
00:34:01memory. And so when we look at how genuine users behave, we can look at these interactions and say,
00:34:09hey, this individual may be the first time we've seen them. And they may be coming from an area that
00:34:14we don't normally serve. And there are things that look risky to us. But when we remove that digital
00:34:20veil and we look at how they enter that information and the picture they paint with their digital body
00:34:26language, we can say that that indeed is a good customer. And the example I'll use here,
00:34:34I don't know how often you experience it, but I was born in rural Idaho on a potato farm. That's not
00:34:38a joke. And my dad to this day still uses a flip phone only because mom makes him use it so we can
00:34:44reach him. He's never used a computer, never used a smartphone. And imagine if he came to your website
00:34:50and wanted to open up a HELOC. That individual would type poorly. And there's a lot about the device and
00:34:57where it's coming from that may look risky, but we can still say, hey, Jeff Smith is familiar with the
00:35:03data he's entering. And we want to ensure that we don't reject him. We foster trust and we let him through.
00:35:09And so again, the how users behave is a critical response to protect this high value space from fraud,
00:35:18but also continue to financially enable the citizens of the US who want HELOCs. And as we move to the
00:35:25next slide on this, I think this is quite powerful and worth noting is fraudsters, as I mentioned,
00:35:32have operated behind this veil since we went online. And when I started a decade ago,
00:35:37we used device fingerprints to stop them. And they've really figured out how to get around that. And here,
00:35:42you see two examples of what happens when you remove that veil and we can see the behavior of our
00:35:47crowd and our population. You see two different examples of customers who turn this visibility on.
00:35:55They see this high risk population that's at their door that they don't even know is there because the
00:35:59information is right. And these fraudsters are very, very sophisticated in addition to using AI.
00:36:04And as soon as those organizations begin decisioning, it is apparent immediately to the fraudsters that
00:36:13that veil they've hidden behind is now gone. And as they stand outside the bank or as they walk into the
00:36:19HELOC, people can now see them and they can see that they are not familiar with that information. And it acts
00:36:25as an active deterrent, even in the most high risk vectors that you could possibly operate in like HELOC.
00:36:32So I think that is to one more poll question.
00:36:41All right. Give me one moment to pull this poll question up for everyone.
00:36:50But in this question, we're going to ask you, where do you see the biggest opportunity for AI and
00:37:00analytics to impact rather HELOC marketing? Give me one moment. So where do you guys expect the
00:37:10audience to see these results?
00:37:12I was I'm really looking forward to see what what the audience's pain point is and the opportunity is
00:37:22right. Whether it's in the in the initial campaign design for improved credit worthiness
00:37:32and targeting right for for the credit worthy consumers, which I think we saw in the previous poll question
00:37:39was high on mine for a lot of attendees. Then what's the best channel, right? How do you optimize your
00:37:52outreach channel such as email, SMS, direct mail or digital advertising? Do you want a lot more focus? Do you
00:38:02see a lot more focus happening in that space? The third being analyzing and refining your campaign
00:38:11performance, right? It's the post campaign. How how how many people actually ended up clicking the links or
00:38:21going through a process or just ignoring the whole email itself, right? What's what does my campaign,
00:38:28what does my offer look like for the consumer? Am I doing it the right way? Or is it like the long term
00:38:41consumer incubation, right? I mean, they may not be ready right now. But do I want to spend more
00:38:47of my analytics and AI resources to actually ongoing incubation of these consumers?
00:38:56So let us know where you see that greatest opportunity for AI and analytics to impact
00:39:04HELOC marketing. We see those results coming in. So continue to choose. I'll give everyone a moment
00:39:10longer and then we will share these results. As folks vote, I strongly agree with everyone. AI, you know, you can
00:39:21tell I'm an AI junkie just because I live on the front in it. But what I look at it and I'm curious
00:39:28what the folks here have to say is it's like a rising tide that raises all ships, right? The question
00:39:33is, is where do you first apply it to increase those efficiencies? And I am thinking that it will
00:39:39probably be question D that most choose, but who knows, right? It depends on the nature of the
00:39:44organization. And it's really just where you direct it first. All right. I'm going to end the poll now
00:39:50and share the results. All right. It's an order. That's great.
00:39:57Okay. And now I'm going to turn things back over to y'all. I mean, what's your take on that?
00:40:08Yeah. I mean, again, not surprising. And I think, you know, the next couple of slides is going to is
00:40:14going to speak to how lenders can leverage both machine learning and harness the power of data,
00:40:25both attributes and models to sort of better identify consumers to have a propensity to convert
00:40:36and being more intentional about selecting and choosing the right data to maximize the
00:40:45benefit from their marketing spend, right? So I often hear from a lot of clients recently,
00:40:51especially is that they would much rather pay more for a lead if that reduces the cost of the
00:40:57acquisition of the business, of the consumer, as opposed to vice versa. And so I think historically,
00:41:05you know, lenders have thrown resources at sort of in a quantity issue. So let's buy as many leads
00:41:12as we can. And then at some point, some of them will convert and it's a numbers game. But I think,
00:41:17especially in this market, I think there needs to be a sort of an adaptation to the shift in borrower
00:41:25propensity and just the fiscal climate. So I think what's interesting, right, is going back to the
00:41:32earlier conversation around choosing data that's more than just your traditional buy box indicators of
00:41:40of credit worthiness. So it's looking at the behavioral aspects of the consumer that might
00:41:46be indicative of greater propensity to convert. So in some of our analysis from from our data science
00:41:53teams recently, we found that consumers that have that are trends that are revolvers have a greater
00:42:03propensity and a greater response rate to targeted marketing than than transactors revolvers just
00:42:10means consumers that typically carry a balance month over month does not necessarily mean, you know,
00:42:19subprime consumers, there are plenty of super prime and prime consumers that carry a balance month over
00:42:24month. That but that and transactors refers to consumers that generally have a very ad hoc credit
00:42:31need, they get the loan, they use it for what it's needed for, and then they pay it off very, very
00:42:36quickly. But what's interesting is the individuals that behaviorally have a tendency to carry a balance.
00:42:42And again, these can be consumers that are, you know, separate there that are super prime and prime,
00:42:47have a greater tendency to respond to a HELOC targeted offer and have a greater response,
00:42:53and then or will actually activate a HELOC. So they'll actually utilize the HELOC more often.
00:42:58So it's a very, very interesting insight. And I think the key takeaway from this, right, is that
00:43:09lenders should use more data to identify these types of consumers, right? So using more and more data
00:43:15to identify the behaviors in which would be indicative of a propensity to convert, as opposed to just
00:43:23finding consumers that are qualified, right, and within my buy box. So it's doing that, that's
00:43:29important, right? You want to target consumers that are qualified, but you want to also find consumers
00:43:35that have a greater propensity to convert to. So I think it's being more intentional about who you
00:43:39target and how you target them. Next slide.
00:43:42So I think we've talked about that there's a strong need to do this, but like, how do you,
00:43:52how do you go about doing it, right? So there, I would say there's three major ways, right? Number one,
00:44:00there are many, many vendors in the market now that can deploy advanced machine learning algorithms,
00:44:05and sorry, models that, that can help identify consumers that are in the market, right? And that
00:44:16means using high, so highly, highly predictive data to identify consumers that would have a greater
00:44:25propensity to convert. So, so it's leveraging not just predictive modeling, but it's also
00:44:34leveraging in various input data, and behavioral data, and demographic data, and various other
00:44:41data inputs. And everyone's going to sort of talk about some of what those other data inputs might be.
00:44:46And then deploying various models, custom models. We have our own experience, has its own versions
00:44:52of these models that can help identify consumers that are, that are going to be in the market, or maybe
00:44:57in the market, but also consumers that would have the greatest likelihood to convert. And we're looking
00:45:03at these things from, from a, from a propensity, from a statistical modeling perspective,
00:45:09and not just actual sort of physical consumer behavior, if that makes sense, like buying a lead
00:45:14from a, from a, from a, from a lead farm, right? Or from a, from a marketplace. So using actual advanced
00:45:20data science to, to, to statistically predict which consumers might be in the market. The, the next thing
00:45:26is, is, is, and, and again, there's an obvious one, but it's a very important one, is to maximize
00:45:33conversion, you have to optimize the messaging. It's extremely important. We talked about how
00:45:40more and more consumers that are down market are amenable to things like a HELOC, you know, 30% of
00:45:48consumers that obtain a HELOC for home improvement are actually below the age of 43.
00:45:53So you, you want to target this messaging to the appropriate demographic group. There are many
00:46:03ways to do that. And, you know, you know, the likes of Experian and the credit viewers, we, we have some
00:46:08of this data to help do that. And then, you know, lastly, the, you don't want to overlook the non-responders,
00:46:16right? So there are tools out there that can allow you to understand why you've lost the business.
00:46:23Right. So it's not just, it's, it's, yes, you found the consumer that's fitting your buy box. They
00:46:29happen to be in the market, but, um, why did you lose the opportunity? It was your pricing,
00:46:35was your pricing robust? Who are the competitors in your space? Understanding that, um, what impacted,
00:46:40what were some of the key features of your messaging? Did you message it correctly? All of those
00:46:45things are very, very important. Um, um, and, and determining how successful your marketing campaign is,
00:46:51right, especially as you reach down market. Um, and I'll turn it over to Upama who's going to kind
00:46:55of talk about, um, the different, um, the importance of data in the, in the inputs in various models and,
00:47:02and just the importance of data in general, and the different flavors and types of data,
00:47:07and just various tools that are, that are available right now in the market to address this problem.
00:47:12Thanks, Ivan.
00:47:16Let's go to the next slide. Yeah. So, uh, yeah, as, as Ivan mentioned, right, I think
00:47:24with all of the, um, the conversation that's just ongoing around AI, right, the, the biggest component
00:47:34that sits behind an effective AI tool or a machine learning, I mean, a traditional AI or a generation
00:47:42generative AI is data. That's the foundational piece. The more multi-connected diverse data that
00:47:53you have within your ecosystem, um, the easier your job becomes, right? And, and how, how, how far back
00:48:01does that data go, right? What's the depth of this data? Because for any predictive, uh, machine
00:48:09learning model, uh, whether it's the, in the market model for, um, for HELOC, for mortgage,
00:48:15for personal loan, any, any product, right? Just understanding the behavior of consumer, the
00:48:24both financial behavior, as well as non-regulated shopping behavior, both are key. And, and one of
00:48:33the big key elements is, is making sure that they sit within their own separate areas so that, uh, in
00:48:43terms of the regulations and all that, there is no intermingling, but at the same time, how do you layer
00:48:50it one after the other, right? How do you define your baseline, uh, population and then, uh, do a credit
00:48:58check, a box run and whatnot, or is there a better way at the end of the funnel to, uh, to define the
00:49:09offer type based on different, uh, shopping habits and the cost of origination and things of that nature,
00:49:16right? So the big thing is, I think the more data that you have already pre-connected and, uh, the behavior,
00:49:26and that is more, uh, the, the broader that data is and deeper that data is your models are actually
00:49:35going to perform in different macroeconomic cycles. They would stay, uh, robust. And that's the
00:49:46way how you differentiate between 10 or 20 different, uh, companies that are selling different
00:49:54predictive models, right? And are, uh, marketing different performances, but, uh, how, how does
00:50:01it perform in different environments, right? So, um, so the, and, and the, the more data that you have
00:50:09in-house or, uh, are getting from somebody who has it in-house, the less time you are spending in
00:50:15in, in making it connected and cleaning and, uh, then going out after with your modeling, uh, exercise,
00:50:25right? So the granular, the trend, the more, uh, deeper your data is, the better, uh, your models
00:50:32are always going to be. Um, if you go to the next slide, I think the other aspect that, uh, and experience
00:50:41being a data company, right? We, we can talk data all day long, right? I will. Uh, but, uh,
00:50:48the other key component that I would, I would point to is, is the trended behavior of a consumer,
00:50:56right? Is a consumer going down in their financial behavior? They were, uh, excellent and now they are
00:51:06good at the current point or they have been improving their financial, uh, behavior and
00:51:12their financial standing. They were fair and now they are good. So both of those at, at this point
00:51:21may look the same, but what is the trajectory is what defines how, um, what their credit needs are,
00:51:30right? And how they are, uh, how they are going to respond to a particular offer and whatnot. Um,
00:51:37one example I'll give you is in one of our models, uh, we found that, uh, credit card balance, uh, if,
00:51:46if that grew by more than $4,500 in the previous 24 months, uh, these, those consumers were about
00:51:56five to eight times more likely to originate a HELOC compared to, uh, a baseline model, even in the,
00:52:07uh, like even the high in the market model consumer. So trended behavior, uh, which is around their
00:52:16balances, their utilization, their payment, uh, pattern shifts, uh, is, is another key element to
00:52:23look at when, um, when you are leveraging your, uh, leveraging AI and, uh, machine learning for,
00:52:34um, for prospecting. So I, I close it out. I know we only have five minutes left, uh, but I want to
00:52:43get like one key takeaway from, uh, I would love a key takeaway from, uh, both Ivan and, uh, Devon
00:52:53as to what the audience should walk away with, um, from this webinar.
00:52:59So Ivan, you want to go first and. Sure. Yeah. Yeah. I think, uh, a key takeaway from,
00:53:07from my end is, is it data, data matters. Um, but, but the, a comprehensive look at data matters. So
00:53:20it's not just, we have to now go beyond the traditional markers of credit success. We want
00:53:29to look beyond those, those typical buy box indicators. We want to look into the behavioral
00:53:35aspects of the consumer. We want to, to be more intentional, proactive and strategic
00:53:43on how we can target consumers because that's unfortunately what is needed now in this market
00:53:50to be successful. Um, in such a complex competitive environment and it's such a volatile environment.
00:53:56It is very important to, to, to look into how data can work for you and can help identify consumers that,
00:54:07um, that, that would be, would, would show the greatest likelihood of converting. I think if you're
00:54:11going to spend the money on marketing, do it more intentionally and correctly, as opposed to just
00:54:16throwing money at the problem and buying leads as many as you can. Yeah. David, and I, you know,
00:54:21I'd agree strongly with both Ivan and Ivan on this. Uh, we talk about data so much, right? And
00:54:27you collect all the data in the world, but what does that matter? It's, it's like if you knew every
00:54:32word in the English language, it doesn't make you a poet. It's how you wield those words or it's how
00:54:37you use this data. And that's what you want to do internally and externally. When you think about the
00:54:42data you have or the data you want truly, how does one use it? And then the only other thing I'll say,
00:54:47just because it is so prevalent is happening. Now we're in the swing of this AI revolution.
00:54:54And as I mentioned last week, we saw what we thought was the first known as far as we could
00:54:59tell AI agent attack, and that acceleration is only going to continue. So as you think externally at the
00:55:07fraudsters also look internally as how you could use that to augment all your processes
00:55:13and your defenses and hopefully make your life a little easier. So that's, that's my two cents.
00:55:20All right. I want to try to get a couple questions in. So the first one that I wanted to go back to is
00:55:31talking about the effective ways to address some of those misconceptions, like fear of debt or
00:55:39complexity, especially among older or first time HELOC users. I think, as Ivan had mentioned before,
00:55:51it's reducing the friction. The more friction we can reduce from the whole process,
00:55:58the easier it becomes. I mean, a lot of times I think we give less credit to even like the older
00:56:09generation that's now that is trying to leverage a new product. People learn really quick. I mean,
00:56:18I see my dad doing things which I would have never expected him to at this age, but they
00:56:28it's just the whole mindset of consumer having the knowledge of what is going on in the background
00:56:42makes it so much easier for them to, number one, ask the right questions, right? Number two,
00:56:49then understand the product itself and leverage it the right way, which is the ultimate driver of an ROI
00:56:56for a lender, right? So that's my take on it. Yeah. I mean, I would agree. And I know we only have
00:57:04two minutes left. So I wanted to make sure we left enough time for some of the other questions, but
00:57:08to answer, yes, totally agree. The next question I want to jump into is how do you expect the HELOC
00:57:15market to evolve? Will debt consolidation continue to be the key use case?
00:57:19I think so. I mean, debt consolidation has always been the biggest driver, but I think I was reading
00:57:29a recent research that came out of Harvard is home renovation is going to continue to grow because,
00:57:44especially because a lot of people are now, I mean, by choice or by need, right? They are staying for
00:57:54longer in their homes. And they definitely want to leverage this equity to kind of bring the house to
00:58:06the best form and do some renovations, expand it if the family is growing and things like that.
00:58:15So that's another reason that HELOC is going to continue to grow. The last one being
00:58:26one-time expenses. There is a lot of that happening.
00:58:30Yeah. And I'll quickly add, because I know we're up against time. The fintechs in the space
00:58:37are creating very innovative products around HELOCs, largely because of the existence of
00:58:45tappable equity. So there's home equity-based credit cards now. There's faster, more efficient
00:58:51HELOC applications and approval. It's interesting. I'm seeing a greater... A lot of the credit card
00:58:59companies are shifting into the HELOC space that traditionally wouldn't have. And so I think that
00:59:10there's a shift in wanting to take advantage of the rising available equity and the increase in home
00:59:18prices in the last five to six years. And so I think that it is a little bit of a feeding frenzy, but
00:59:24there's a lot of innovation that's happening around HELOCs in general. And I think there's tremendous
00:59:30opportunity for sure, I think, in the space. Almost making it feel like a credit card, right?
00:59:35Yeah. Yeah. Well, unfortunately, we are up against time to our audience. I know that there are a couple
00:59:42of questions in the Q&A. Housing Wire will be turning over all of the Q&A to the Experian team,
00:59:48and they will be able to answer your question afterwards. So we will make sure that you get
00:59:53your answers. But unfortunately, that wraps today's conversation, the $29 trillion opportunity,
00:59:58smarter strategies for HELOC growth. I want to thank our speakers, Upavan, Ivan, and Devin,
01:00:04for sharing their insights and for everyone who joined us live. If we didn't get to your question,
01:00:09once again, we will make sure that the Experian team has all of the information to follow up with
01:00:14you afterwards. This webinar was recorded and will be sent to everyone who registered early next week.
01:00:19Thank you again for joining us, and we hope to see you all again soon.