Prepared remarks
Hello and welcome to the Better Home & Finance Holding Company Fourth Quarter and Full Year 2024 Results Call. All lines have been muted to minimize background noise. Following the speaker’s comments, there will be a question-and-answer session. I will now hand over the conference to Hana Khosla, Vice President of Corporate Finance and Investor Relations. You may begin.
Welcome to Better Home & Finance Holding Company's fourth quarter and full year 2024 earnings conference call. My name is Hana Khosla and I'm the Vice President of Corporate Finance and Investor Relations at Better. Joining me on today's call are Vishal Garg, Founder & Chief Executive Officer of Better; and Kevin Ryan, Chief Financial Officer of Better; and Ryan Grant, Co-Founder & President of Retail Lending at NEO Home Loans. In addition to this conference call please direct your attention to our fourth quarter and full year earnings release which is available on our Investor Relations website. Also available on our website is an investor presentation. Certain statements we make today may constitute forward-looking statements within the meaning of federal securities laws that are based on current expectations and assumptions. These expectations and assumptions are subject to risks, uncertainties, and other factors as discussed further in our SEC filings that could cause our actual results to differ materially from our historical results.
We assume no responsibility to update forward-looking statements other than as required by law. During today's discussion management will discuss certain non-GAAP financial measures which we believe are relevant in assessing the company's financial performance. These non-GAAP financial measures should not be considered replacements for and should be read together with our GAAP results. These non-GAAP financial measures are reconciled to GAAP financial measures in today's earnings release and investor presentation, both of which are available on the Investor Relations section of Better's website and when filed in our Annual Report on Form 10-K with the SEC. More information as of and for the period ended December 31, 2024, will be provided upon filing our quarterly report on Form 10-K with the SEC. I will now turn the call over to Vishal.
Thank you, Hana, and welcome to our fourth quarter and full year 2024 earnings call. We appreciate everyone joining us today and for your continued support as we work towards our mission to make home ownership better, faster, and easier for our customers by creating a technology platform that transforms the home ownership experience. We continue to make good progress towards our vision where every customer can effortlessly buy, sell, refinance, insure, and enhance their home digitally, online, instantly. I would like to begin by highlighting some of our key achievements. We entered 2024 focused on growth and artificial intelligence to increase volume and revenue while maintaining efficiency, diversifying distribution channels, and reducing corporate costs. We executed on these goals, achieving a 19% year-over-year growth in full year funded loan volume, a 50% increase in revenue year-over-year, and a 26% decrease in adjusted EBITDA losses year-over-year.
Significant investments in Tinman AI and the launch of a distributed retail channel through NEO Powered by Better are showing early positive results. Despite a challenging market characterized by low housing affordability and high mortgage rates, we made strides in 2024 aligned with our initial roadmap. In 2024, we reached $3.6 billion in funded loan volume and recorded $108 million in revenue, alongside an adjusted EBITDA loss of $121 million. In the fourth quarter, our funded loan volume was $936 million, reflecting a 77% year-over-year rise, driven by growth in all three primary product categories—purchase, refinance, and second lien loans. Seasonally, funded loan volume saw a 10% decrease from the previous quarter due to typical slowdowns in the fourth quarter purchase market. Q4 revenue climbed to $25 million, compared to $18 million in the same quarter last year and $29 million in the previous quarter.
We are consistently working on strategies to enhance conversion with additional products and services while improving sales efficiency to boost customer engagement. Throughout 2024, we have increased revenue per loan through pricing strategies and optimization of marketing channels, resulting in year-over-year gains in sale margins from 1.95% in 2023 to 2.17% in 2024. Looking ahead to 2025 and beyond, our strategic goals remain aligned with previous discussions. Our primary focus is to continue to intentionally pursue growth, which we made progress on in 2024. In the fourth quarter, year-over-year loan volume growth was spurred by increases across our main product offerings, with home equity products and refinance loans being key drivers. The year-over-year growth in purchase loan volume was 25%, while refinance loan volume surged by 611% from a particularly low point at the end of 2023.
Our HELOC and home equity loan volumes saw a staggering 416% increase in Q4 of 2024. TransUnion’s recent reports indicate overall HELOC and HELON originations rose by 10% in Q3, while Better experienced a remarkable 619% growth in origination volume for the same period. If HELOC market trends remain stable or improve in Q4, we anticipate that our current rapid growth will continue to outpace the industry as we gain market share thanks to our superior offerings. On a quarter-over-quarter basis, refinance volume rose by 34% in Q4. Home equity loan volume slightly increased by 3%, whereas purchase loan volume decreased by 10% primarily due to the typical seasonal slowdown in home buying during the fourth quarter. Although the mortgage market improved in Q4 relative to the same period in 2023, 30-year fixed mortgage rates remained in the high sixes to low sevens, continuing to exert pressure on mortgage demand.
We anticipate that rates will stay elevated in the near term, which should sustain demand for our Home Equity products. Our second priority is to keep reducing expenses and enhance operational efficiency with a view toward achieving profitability in the mid-term. During Q4, total expenses remained approximately stable compared to Q3. However, they included around $17 million in non-recurring restructuring costs primarily linked to unwinding our UK operations, along with an additional $4 million related to lease terminations. Excluding these one-time costs, overall expenses decreased by about 24% quarter-over-quarter, thanks to management initiatives. We trimmed adjusted EBITDA losses by approximately $11 million or 28% in Q4 relative to Q3, even amid lower revenues due to the seasonal downturn. Specifically, we achieved a 28% reduction in loan origination expenses, a 21% decrease in compensation costs, and a 27% cut in marketing and advertising expenses, while only seeing a 10% decline in volume.
This demonstrates the effectiveness of our AI and automation initiatives that are fostering operational efficiency within the business. By leveraging Tinman’s capabilities, we have automated many time-intensive aspects of the mortgage process, resulting in a cost to originate that is over 35% lower than the industry standard. We believe that our ongoing investments in technology and AI will further drive down costs, enhancing our operating efficiency and improving the customer experience. Continuing our discourse on AI investments, we've seen significant benefits from expanding Betsy, the first AI voice-based loan assistant in the U.S. mortgage industry. Betsy utilizes large language models to guide customers through pre-approval, rate quoting, and rate locking autonomously. It can verbally interact with consumers to handle mortgage application inquiries and gather and verify necessary application data, functions that were traditionally the domain of sales staff, with the goal of enabling faster service, better self-service options, enhanced customer engagement, and improved sales efficiency.
Betsy’s performance has skyrocketed—from approximately 5,000 customer interactions in June 2024 to over 115,000 by February 2025, marking a growth rate of over 20 times while we continuously expanded its functional capabilities. In the last quarter, we showcased an early interaction of Betsy with a customer. We are now excited to demonstrate Betsy managing a customer through the entire mortgage rate lock process, which previously required multiple interactions with sales team members and manual data entry across several phases. With Tinman technology, this is now achieved entirely autonomously through AI, available 24/7, allowing customers to engage anytime and anywhere. I want to emphasize what sets Betsy apart from other AI bots you've likely encountered in financial services. Unlike typical FAQ bots or appointment schedulers seen with other mortgage companies, which generally perform a single function through chat, Betsy stands out because it is integrated into an end-to-end platform that combines several essential systems used in the mortgage process.
This integration allows the AI to maintain full context and communicate back and forth with consumers as a human would. Most other mortgage companies would experience significant delays if they tried to implement a comprehensive AI agent due to connectivity issues between disparate systems. However, Tinman's unified platform allows Betsy to efficiently guide customers from the initial inquiry to funding in one seamless interface. Additionally, we have scaled the application of our Tinman AI to automate functions traditionally handled by underwriters, especially with the qualification of income, assets, and credit to facilitate underwriting decisions. Our automated processes have allowed us to grow our one-day mortgage product to more than 70% of our volume, with average times from lock to commitment letter down to eight hours—revolutionizing the mortgage lending experience when traditionally such a timeline could take up to 45 days.
Recently, we have begun to leverage AI in ways that can reduce the process to under a minute in certain scenarios, which is remarkable—transforming the one-day mortgage into a sub-minute mortgage enabled by AI. Our objective is to elevate AI underwriting review to over 75% of our locks by the end of 2025, aiming to drastically cut fulfillment costs per loan and streamline critical areas in the mortgage process. We've also made significant advancements in mortgage underwriting by intelligently routing appraisal requirements and enabling AI-driven underwriting for title insurance. By enhancing this process, we are already facilitating instant title and appraisal for a small percentage of our locks, and we plan to expand this capability over the coming months. We believe that the prospect of fully underwriting a mortgage across credit, income, assets, title, and appraisal within minutes upon entering a contract is very close.
The core message is that we are intensifying our efforts to capitalize on the significant competitive advantage we possess with AI agents within our Tinman platform, especially with the recent regulatory environment becoming more favorable towards AI. There has been a noticeable shift in regulatory attitudes regarding AI in recent months, creating new opportunities for us. Just last month, we implemented Betsy as the primary point of customer interaction in one state for refinance loans, with human loan officers available as a support measure. We see the potential for Betsy to lower sales costs per loan by $2,000 and operational costs by $1,400, which could bring our expenses into line with industry standards, and we are just beginning this journey. To conclude, our third priority is diversifying our distribution channels by expanding our B2B operations. We are witnessing sustained interest in our technology platform from new partners with reputable brands eager to provide mortgages to their customers in a cost-effective manner or enhance the efficiency of their existing mortgage businesses.
I want to begin with an update on our collaboration with Ally Bank. Better and Ally have developed a mortgage partnership for over five years, which we are proud of, but at the end of 2024, Ally decided to exit the mortgage sector entirely. Consequently, we started to wind down our volume with Ally in Q4 and anticipate completion by the end of Q2 this year. Alongside the typical winter seasonality, this transition has contributed to a decrease in volume in Q4 compared to Q3 of 2024. Moving forward, last quarter we announced the launch of NEO Home Loans Powered by Better, and I am pleased to share that we are making significant early strides toward diversifying Better's distribution and leveraging Tinman to empower local loan officers by streamlining their fulfillment process and increasing their capacity to assist more customers. NEO Powered by Better will harness Better's AI technology and digital lead funnel to enhance NEO's loan officer teams, who have established reputations for customer service excellence in their communities.
Betsy, uniquely branded for each retail loan officer at NEO, along with the lead routing of early-stage purchase customers from Better's direct-to-consumer channels, is expected to markedly improve efficiency and conversion across both direct-to-consumer and retail distribution channels. Since commencing production in January 2025, we have onboarded around 110 NEO loan officers across 53 branches, and NEO Powered by Better has already helped approximately 220 families, resulting in $95 million in funded loan volume. For NEO Home Loans, we are observing an average gain on sale margin of around 365 basis points in comparison to our Better.com gain on sale margin of 217 basis points in 2024. I am enthusiastic about the unique opportunity to tap into key market segments that have historically been difficult for online originators without local footprints, particularly in the purchase mortgage sector and specific loan types like FHA, VA, down payment assistance programs, and buy-down programs. Joining me today is Ryan Grant, Co-Founder and President of Retail Lending at NEO Home Loans, who will share his insights on the exciting prospects we have ahead with NEO Powered by Better.
Thank you, Joel. Our team here at NEO Home Loans is incredibly excited to now be Powered by Better. This partnership is more than just a collaboration. It is a fundamental shift in the mortgage industry. We believe that together we are creating the most valuable mortgage platform, not just for our clients and our business partners but for every mortgage professional in America. NEO Home Loans was founded on a simple but powerful idea: changing the expectations of what a mortgage company should be. For decades, the mortgage industry has focused on selling debt, leaving many professionals questioning the real value they provide, and we set out to do more to guide clients well before they purchase a home and continuing to proactively support them for decades after, helping them build long-term wealth. However, we face significant challenges from an entrenched industry. Outdated technology made processes inefficient, scalability was costly and limited.
We struggled to get our message in front of enough clients, and for local mortgage professionals, the industry lacked financial transparency, which created a misalignment of priorities. These were all major barriers, but the partnership of NEO Powered by Better is solving them all. That's why we're so proud of this opportunity as it positions us to truly transform the industry. You see, for years as mortgage professionals, we found success despite technology, not because of it. And when we visited Better's headquarters, we saw Tinman and met Betsy, and we were stunned. Better had built technology that matches human level performance across a range of mortgage tasks, something that no one in our industry had seen before. And with this AI-powered infrastructure, NEO can now combine the best of both worlds: the speed, efficiency, and automation that clients want with the advice, strategy, and long-term commitment from a local mortgage advisor that they need and deserve.
This is an absolute game changer. Now our strategy at NEO Powered by Better is built on three key beliefs. First, we will drastically reduce loan costs and increase the scalability of our teammates, which is a major problem for most in the mortgage industry. With Tinman and Betsy, our teams can efficiently serve more families with more value and at a much lower cost. This allows us to compete with discount lenders without sacrificing value. Our second key belief is that we can help subsidize the local mortgage professional with lead generation that has become much harder in the past few years. The market has shifted, making it harder for mortgage advisors to find clients that need help. And when we learned that roughly 30,000 people per month are inquiring with Better about purchasing a home, we knew that our team of highly trained advisors could convert more of these curious prospects into actual homeowners.
Now we're going to begin scaling lead routing in April with a short term goal of 10% conversion, which is roughly a 500% increase over the current levels Better experiences in its direct-to-consumer channels. We're also excited to connect these home buyers with the best real estate agents across the country and help more of their clients as well. And between the increased lead conversion and working with more of the best agents in our local markets, we expect that the cost of acquiring a client can be drastically reduced. Now our third key belief is that by creating the industry's first truly transparent partnership lending model, we can empower local mortgage professionals to have the confidence, knowledge, and financial understanding to operate at much higher levels. And by doing this, NEO Powered by Better is creating a natural alignment of interests between each team member in the organization, as opposed to misaligned priorities that mortgage professionals have had to deal with for decades prior.
Lastly, we are excited to be able to share more about NEO Home Loans Powered by Better with local mortgage professionals across the country. We expect that when they see and start to understand this combination of technology, efficiency, scalability, and growth through lead generation and referral partnerships, the best and brightest will want to partner with us in our efforts to completely change and improve the mortgage industry. In 2024, our NEO team funded approximately $2 billion in mortgage volume while remaining profitable. Now Powered by Better, we're positioned to scale even faster, drive greater profitability, and deliver even more impact. Together we’re redefining mortgage lending, and this is just the beginning. Thank you. And with that, I'll turn it back over to you, Vishal.
Thank you, Ryan. We are so pumped to have the NEO team on our platform and as our partners as we disrupt the mortgage industry together. Looking now towards 2025 and beyond, the medium-term opportunity is very exciting. We remain focused on enhancing our go-to-market with growth being our North Star alongside continued expense management and channel diversification. We will continue to invest in Tinman AI to improve the customer experience and further drive down labor costs, making our platform more efficient and scalable, driving the business to profitability in the medium term. With that, let me now turn it over to Kevin Ryan, our Chief Financial Officer, who will discuss the quarterly performance and our financial strategy. Kevin?
Thank you, Vishal. As discussed in 2024, even through a continued challenging market environment, we've made great progress towards our goals of driving increased volume and revenue balanced with ongoing expense management and improved efficiency. In the fourth quarter of 2024, we generated funded loan volume of $936 million, revenue of $25 million, and an adjusted EBITDA loss of $20 million. Total GAAP net loss was approximately $59 million. Our fourth quarter funded loan volume was 81% generated through our direct-to-consumer channel and 19% generated through our B2B partner channel, with 62% from purchases, 18% from home equity loans, and the remainder by dollar volume from refinanced. In addition, we are experiencing early success with our UK bank, the Bank of Birmingham, with scaled loan originations over 10-fold from December 2023 to December 2024, as we have implemented our technology in the UK.
We expect to more than double UK bank originations again in 2025 as we deploy AI with the goal of building the leading AI-driven specialist mortgage bank in the United Kingdom. Turning now to our outlook for full year 2025. We remain focused on managing towards profitability in the midterm and we expect to drive growth through efficiency from Tinman AI, distribution channel diversification, and optimized marketing while balancing these growth expenses with further corporate cost reductions. For the first quarter of 2025, we expect funded loan volume to be down approximately 10% to 15% compared to the fourth quarter of 2024, given continued seasonal slowness and the wind down of our businesses, which as a reminder made up 29% of our full year 2024 volume and 19% of our Q4 2024 volume, and we expect to be only low double digits percent of Q1 volume before fully winding down at some point in the second quarter.
We are particularly excited that the NEO funded loan volume is pacing ahead of plan, and we expect to do over $90 million of NEO originations in March alone after February was the first full month of NEO on our platform. As another data point here to put our trends in the context of the industry, the Fannie Mae February housing forecast is overall Q1 market volumes declining 24% quarter-over-quarter, demonstrating Better’s outperformance of the market as a whole in Q1. For the full year of 2025, we expect funded loan volume growth in the low to mid double digits percent growth year-over-year, driven by tailwinds from growth initiatives including NEO Powered by Better, offset by continued macro headwinds and the loss of the Ally business, a roughly $900 million headwind. We expect this growth to come particularly in the second and third quarters of the year, at which point we expect NEO to be more fully ramped and to benefit from improved seasonal tailwinds.
We also expect to further decrease our adjusted EBITDA losses in 2025 as compared to 2024 due to a combination of efficiency gains as well as continued corporate cost reductions. Lastly, we are undergoing efforts to exit our non-core UK assets while continuing to focus on growing the bank. We expect the exiting of three smaller non-core UK businesses to start being a benefit to our adjusted EBITDA losses in the second half of 2025 as a result of their disposition. With that, I'll now turn it back to the operator for Q&A.
Questions and answers
Thank you. Your first question comes from Eric Hagen with BTIG.
The AI playback was actually pretty interesting. How does the underwriting and the AI technology adjust for the high cost and limited availability of property insurance? Like can the tech adjust for that in any way and do you even see that maybe creating an opportunity because there are folks coming into you online and just finding ways to fulfill that loan more efficiently? Thanks.
I mean, that is a really great question. I mean, what you saw there was not just a form-filling outer engine. There are over 15 different data points and API calls that went through 45 investors filtered down to five HELOC investors and running all the permutations across credit, DTI, LTV, cash-out amount, and actually insurance quotes closing costs across 3,600 counties in the U.S. We have an insurance engine built in where we deliver instant homeowners insurance to consumers while they're going through a refinance, or a HELOC process, or a cash-out refinance process. So we're talking about things that used to take a lot of people to do. So like, again, if we think that we're going to be able to get the 10-year treasury down, back in 2019 we went from $85 million in revenue to over $850 million in 2020, over $250 million of EBITDA, but the machine was about 50% automated. Right now, Betsy can do basically all of the functions that those refinance salespeople were doing back in 2020 at zero, near-zero marginal costs.
That means we don't have to hire 3,000 salespeople, we don't have to hire 5,000 processors and underwriters, and we don't have to hire 1,000 insurance agents that we used to have. So I think there's just extraordinary scale that we've now built into the product and Betsy is accommodating all of that, and we're really, really looking well-positioned in a way that we haven't been in many years for anything changing in the macro environment, including what you've outlined—homeowner insurance rates going up.
Really good stuff. Interesting. If the trend for profitability keeps moving in the right direction, how do you maybe think that will drive the amount of risk you take and how do you benchmark the amount of risk you're taking in a certain period, and even how you might price for things on the front end?
That's a really good question. I think what gets lost in the dialogue about us versus many of the other fintechs is we are operating a pure marketplace business. We do not hold loans on our balance sheet that have not already been committed to be sold to others. We have 45 institutional investors on our platform that are buying our loans, our HELOCs every day. And so, fundamentally, we don't make a loan unless we've got a buyer lined up. When you're locking that loan with us, we already know where it's going, and that's how Tinman is fulfilling the set of underwriting criteria for that particular investor to deliver that loan to that particular investor. So I think the path to profitability we're talking about is not one built on taking any more marginal units of risk. The path to profitability is really built around, like look like we lost like basically $9 million a month last quarter, right?
With what we're doing to shut down the UK businesses, that's like a million a month. What we're doing to improve the profitability at the bank, we think that that gets us another million a month. I think with AI driving down operations costs, I think we can scale up and save another $3.5 million a month. We've got a whole bunch of compliance legal costs from the DSPAC, the very aggressive CFPB era, all of that sort of stuff. I think we can scale that down a million a month. We got a bunch of legacy contracts that we still have from 2021, 2022, right, which we just finally got out of the office space that we had in New York that was like 45,000 square feet and downgraded it by 80% and moved to a cheaper space. So that's another two million a month. And then, you should add some volume growth, add some improvement in margin, add some profit from the NEO channel, and you're getting to breakeven.
And so we see a path to breakeven, again, built on efficiency, built on exiting a bunch of the legacy costs that we have from the 2021, 2022 days and the legacy, exiting the legacy businesses that we have and improving margins that you see us continuously doing. And again, without taking any more risk from a credit standpoint.
Yeah, I mean, Eric, I'll add to what Kevin and Vishal said: like I think 100% of the loans we did in the quarter were pre-committed to investors at the time of origination. If it's not 100%, it's 99%, right? So we don't take really any risk. The only way we think about risk is should we lean into marketing this month versus next month, depending on market conditions, right, but that is a very tactical, in-the-moment decision. And then on the expenses, if you look just Q3 to Q4, we took out $11 million of expenses, core expenses. The expenses look roughly the same because we took a $16 million charge on the disposition of the UK assets. That was a non-cash one-time expense. But when you stripped that out, we got expenses down about $11 million, or 12 million bucks in a quarter, or $4 million bucks a month. So we are, and in all major categories, I think, as Vishal said in the prepared remarks, we were down, corporate expenses were down a bunch, marketing was down a bit. So we definitely took a lot of expenses out in Q4 and continuing to do so in Q1.
Really good stuff here guys. We appreciate you. Thank you.
Thanks, Eric.
The next question comes from Jake Kooyman with Oppenheimer. Your line is open.
Hi, thank you very much for taking my questions. Can you walk through the saving opportunities from Tinman's application of AI as well as how that contributes? Thank you very much.
Totally, so I think the savings are when we think about your traditional loan officer and loan officer assistance, right? The bulk of their time particularly in the direct-to-consumer channel is spent servicing customers that are coming in via the online channel, chasing after those customers in the purchase market, chasing after the realtor whom those customers are using, and so there's a ton of effort on outbound calls, and then there's a ton of effort chasing inbound calls that you missed because you were on the phone with someone else. Now, again, you can staff up, you can have a 10,000 person call center to capture all these calls and make all these outbound calls like other mortgage companies do, doing 400 outbound dials a day. It's really inefficient and really grinds down the labor force. We have Betsy doing all inbound calls in the nights and evenings, so we don't miss a single call.
We used to miss 40% of calls that would come in because people would not be available to meet their loan officer because they were calling at 9:00 PM in the evening after they put their kids to bed. And they're looking at what they're doing for the home buying coming that weekend, or they're calling us on the weekend when they're about to go into contract on their home, and they want to make sure that the rate quote is still good and they want to refine the purchase amount. Now we had these tools online, but Betsy really dramatically reduces the cost, but also most importantly improves the customer experience because it's always on. And so I think that's been a game changer. I think there's the ability to take up $2,000 per funded loan in sales costs once Betsy gets fully implemented in the sales funnel, right? So we're doing almost 1,000 loans a month, right, and we're trying to scale that up.
With NEO, it's more than that, so we're getting there, right? That's some serious savings per month that we're able to generate as we implement this, not just for ourselves but for our B2B partners. On the automation side, we are pressing ahead. If you look in the earnings supplement, you will see the percentage of locked loans that are AI underwritten, and that's increased about 40%. The loans that are AI underwritten, we're saving $1,400 per loan potentially, right? And again, so you add those two things up: we're talking about a production cost that's already more than 35% cheaper than the rest of the industry, and now you're talking about for the full AI-driven loans, you're talking about $3,500 per loan in savings on top of that. Now that's all going to go to margin because we already have some of the lowest gain on sale and therefore the lowest price to the consumer out there. So all those AI enhancements will effectively drop to the bottom line. I hope that provides some context.
The next question comes from Bose George of KBW.
Hi, good morning, everyone, this is actually Alex Bond on for Bose. Thank you for taking our questions. Just to start with us now, almost at the end of the first quarter, I was wondering if you'd be able to give us an update on how gain on sale margins are trending quarter-to-date compared to Q4 in light of the decline in rates over the course of the quarter. And then also, as you mentioned in the prepared remarks, the gain on sale margin on NEO loans has been stronger to date than the 2024 company-wide margin, and you mentioned that there's potential to improve this further as efficiencies improve. Is this primarily from AI and other tech-related improvements or would there be other components that could be improved efficiency-wise as well? Any additional color there would be great. Thanks.
Okay. So Kevin, I'll start and I'll unpack that. There's a couple, I think, sub-questions in there, and Vishal probably you want to jump in. So on gain on sale in Q1, it is trending higher. I think we put in a release $90 million in yield loans already in Q1. We really just, they just onboarded in February. We crossed $100 million this morning. So we are over $100 million loans, and they are running much higher, 150 basis points higher on average gain on sale than the direct to consumer business. Now that is something we knew going in, we expected; we would've been disappointed if they weren't running higher gain on sale, just given their boots-on-the-ground business, their expertise, etc. So our aggregate gain on sale should trend higher as NEO is a bigger part of our production, right? I mean, practically we're replacing a billion dollars a year of Ally volume with call it $2 billion in NEO, let's just say this year.
And that will be a much higher gain on sale than what we would've reported on our Ally, so. And then on the direct-to-consumer business, I'll start, I'm sure Vishal will want to jump in. Yeah. Betsy and the AI allow us to—we've gradually increased our gain on sale in the direct-to-consumer business, right? We were sub-2%, we're now north of 2%. We're not at the 3.5% that NEO's at, but the business isn't built that way, right? It's an online business. As Vishal said, we have some of the lowest rates out there, but through our improvements and better customer experience through the AI, we've been able to gradually increase our gain on sale. And then the rate drop we've had, you know, we're around 6 and three quarters now on rate, right has definitely helped to bid as well.
Yeah. So I'll tell you what's contributing to the margin increase. Online, you know, a consumer submitting effectively a lead when they're online, they're shopping around, they're going to our site, they have a tab open with somebody else's site. They might be on one of the comparison shopping engines, and typically consumers would submit a lead and it would take us more than five minutes to get back to them, right? To call them to try to reach them, by which point they may have gone somewhere else. So the efficacy of our marketing was lower, but also they were shopping around. We've taken that five minutes and brought it down to 800 milliseconds with Betsy across the board—that's an improvement of 400x in speed—and so now we’re catching the customer faster than anybody else. We're catching the customer before they have a chance to go somewhere else, shop around, we're able to tell them about our closing guarantee.
We're able to tell them about the better price guarantee. We're able to answer their questions, we're able to convert them from a lead to an application. We're able to approve them, we're able to do all these things that before, just with a human-staffed call center, was kind of nearly impossible—and so then you end up competing much more on rate than you do on speed and service, and I think that's again, it enabled us to continue to get better margin while still maintaining our value proposition for the consumer.
Great. That all makes sense. Appreciate you taking the questions.
Your next question comes from Reggie Smith of JP Morgan.
Really encouraging the disclosures you gave around the potential savings from, I guess, a lot of your cost initiatives. My question, and I'm not sure if you guys have broken it out or even thinking about the business this way, but is there a way to contextualize contribution profit per loan or loan economics that way? I know you guys cited some savings potentials in the press release, but I'm curious how you guys think about and how we should think about like loan economics at the loan level. So like revenue per loan expenses, is there a way to attribute whether it's marketing or overhead to the loan origination process? And I have a follow up. Thank you.
Yes, and so it is Kevin, I'll make a few comments. Obviously, in the GAAP financials, you don’t see that. We run the business on a contribution margin basis, and the contribution margin in the mortgage business has been improving meaningfully in the last couple of months. I think we can go and we'll take away—to break it out for next quarter in a way that you’ll be able to kind of track it back to the GAAP financials. It’s something we talk about all the time. But I will tell you through these savings and the improvement in gain on sale, the cost savings via the AI, the contribution margin continues to get better on our production and continues to be into Q1. That’s something we are maniacally focused on, right? Because while we’re cutting corporate costs faster than we’re cutting costs in the mortgage business, we also need to lower costs in the mortgage business in order to really drive contribution profit that we can then use to cover what is good. You’re always going to have some fixed overhead. So we think about that all the time.
Yeah, Reggie, 2024 was a lot of changes. We moved from a salary-based loan officer and processor model to an incentive-based low base, high incentive model. We started recruiting experienced loan officers. We had to teach those loan officers Tinman. The ones that didn't understand or couldn't be productive on Tinman, we had to let them go, so we had some charges. So 2024 and then we did NEO at the end of the year. And then lastly, like we were hoping for the rate cut to actually in September to actually bring rates down and that didn’t. So that made a bunch of our marketing spend negative because we’re buying leads, and the consumers thinking they’re getting a 6.5% rate; by the time they get to lock, they’re getting a 7.25%, right? That consumer is not in the money anymore or that consumer isn’t able to proceed moving forward or is going to wait. So, we had a lot of challenges in 2024. We made a lot of progress.
But what we decided by the end of '24 was like we are going to focus maniacally on any growth that’s going to come with positive contribution margin, and we’re going to continue to expand the positive contribution margin even if that means forecasting a slightly slower growth rate. Now I think we can achieve fast growth and improved contribution margin, and that’s really possible not in a human-centric business model but an AI-centric business model, and I think that’s really what we’re leaning very, very hard into. So hopefully that can give you some context for the future. And of course, we’ll take it under advice that like we need to get out to make your job easier that to break out contribution margin next quarter.
I didn't know if I had missed that or what, like I said, I don't have a model for you guys, and so like, that was just something I was looking for. So I wasn't sure if I could sneak two more questions in really quickly. One, I love the demo that you guys showed, or presented during the call. Was curious how it's resonating with consumers, and my inclination, and I could be totally off, is that for younger people, it's probably a more natural way of doing things. I was curious any feedback you've gotten or insights you've gotten in terms of, are younger folks using the automated system better than older people? And how often do people opt out and say representative or something like that? And then my last question, you talked about, I guess approving a loan in one day. I was curious how quickly you guys can fund a loan. That's something that I've come to appreciate in the last month. I'm trying to sell my condo in Brooklyn, and so like people that have that are able to close quickly makes a difference. I was curious if you had any advantage there as well. Thank you.
Yeah, no, that's a great question. So we're seeing adoption of Betsy in general be quite high. About 18% of consumers ask to be transferred to a human loan officer as part of the process, right? So they encounter it, they realize that it is an AI and they want to move on to a human. So yeah, we're going to work to get that down and we’ve got to work to get the voices more humanistic. We've got to do a lot of work on continuing to add the functionality. So I would tell you it's early days. Like we launched the first version of Betsy only a few months ago. So you can sort of see we’re increasing functionality, increasing realism, and it is going to improve. The uptake has been greatest between the ages of 20 to 35 and then, honestly surprisingly, 55 and up, right? Because they're okay with something that goes a little slower in terms of talking and is talking clearly. We've seen some good stats around that.
Again, early days, we'll see how that all shapes up but definitely, young people love the fact you can talk to at any time. You can text with it, you can have a loan officer you can go to anytime. So we're just really leaning into that. And then the last question with respect to closing times. I think in New York, we're able to close a loan on average—let me just actually get the stats, but much faster than the competition. And, give me a second. I'm just going to pull it up for you. In New York, we’re doing—the industry average for closings is 46 days and we're doing 32 days. Now, of course, there's a lot of latency there with people having to figure out how to move and all that sort of stuff, but we're about 40% better than the competition in New York State specifically.
No, that's good and it is something I didn't appreciate until I went through the process myself. So glad to hear that's a really important selling point for buyers. And I'll tell you, like personally, I hate AI, but it's something I've got to get used to in terms of the automation. I'm one of those 18 percenters that's always representative as soon as I hear an automated system. But yeah, no, it's, it'll be interesting to watch that evolve over time. Congratulations on the quarter and good luck, guys.
Reggi, if you find a buyer in Brooklyn, tell him to go to Better.com; we'll preapprove them in a couple of minutes.
The next question comes from Michael Kaye with Wells Fargo. Your line is open.
If I look at the Q4 adjusted net income and adjusted EBITDA, there was no improvement in year-over-year profitability despite volumes being up 76% year-over-year. Can you just walk through why the higher year-over-year volumes and cost initiatives over the last year is not translating into better profitability? Maybe just walk us through some of the dynamics and if I'm missing anything.
You're doing year-over-year is what you're doing?
Yes, $38 million loss, adjusted net loss in Q4.
Yes, as Vishal said, we—marketing expense? We took marketing expense up, and that was the biggest difference, and we actually hired more people as well into—. As Vishal mentioned before, we hired more people into what we expected to be a declining rate environment in the second half of 2024, which bluntly didn't really pan out, and we've pulled back a bit on that. And so that—some of that noise you're seeing in there. Q4 of '23 was a low point of volume. It was also a low point on staffing within the mortgage business—the mortgage factory. So all the other corporate costs and everything have come down dramatically since Q4 of '23, but marketing expense was higher Q4 '24 as was sales and ops labor. As we're getting Betsy going, that will continue to come down on a unit basis, which I think was part of Reggie's question, which we're going to try to break out.
Yes. I mean, we were overstaffed in Q4. After implementing Betsy, we've been able to reduce staffing by about 250 people in the mortgage factory.
Shifting gears. And what's your level of optimism on spring home purchase season? Rates are around six and three quarters, more home inventory now available, probably some pent-up demand from buyers, though there's affordability headwinds. So just talk about how you're thinking about spring home purchase season.
We're seeing the volume of pre-approvals per marketing spend continue to improve dramatically. And I think that that’s a good leading indicator, right, of how many people are running out and just going shopping. We'll see how many of those people actually are able to find a house. But so far, it looks like the level of preapproval per dollar of marketing spend continues to improve pretty dramatically. So we're optimistic for what happens. And look, like, again, the noise from Washington is about deregulation and about getting the tenure down, so you put your odds on whether they can make that happen. If you believe that that can happen, I think we might be in for a positive surprise this spring and summer home buying season.
The next question comes from Jamie Friedman with Susquehanna. Your line is open.
Interesting presentation and demonstration, really quite helpful. Most of my questions, I was just wondering about the macro in terms of how you're characterizing the supply and demand dynamic in the end market. Where do you think we are in that continuum and what are you anticipating, if anything, for the year ahead?
I think what we always say at Better now is that we have unfortunately not been able to predict the macro environment for the past three years. We are optimists at hearts as any technologist will tell you, you have to be an optimist, otherwise, how are you going to believe? We do think that the supply-demand imbalance in housing is going to get rectified in the next year or two. I think there's obviously an impact of tariffs, but that means that the homes that are out there available for sale today are a relative bargain if those tariffs do actually continue and the raw material cost of building a home goes up substantially. So I think old houses are going to sell more. We're seeing demand for the HELOC product really explode because people need to fix up the houses—like the boomer houses need to be modernized to be able to sell to the millennial buyers, and so we're seeing a ton of demand for home renovation on the HELOC product. Before people look to sell their house or if they stay in in the same place, we grew our HELOC this 400% last year, fastest growing HELOC lender in the market, and we think we can grow it again pretty dramatically this year. And we've gone from basically nothing to being one of the leading HELOC lenders in the country. And so I think we're trying to build a balanced portfolio of loan types so that we can thrive in any macro environment out there.
Yeah, I think, we're, as Vishal said, it's been really hard to predict the macro and if you listen to earnings calls this season, right, we're in at the tail end, the word uncertainty comes up 20 times a call. That's just kind of the macro we live in right now. Every company does, and our industry in particular, that's important. We get it. But we definitely think we are way ahead of the trend and the inevitability around technology disrupting the legacy mortgage process in the US is probably taking a little bit longer than we thought, but the trend continues and we think we're definitely right on that trend and that'll play out over time. Regardless of where rates and the starts data is pretty good in February, but then people say consumer confidence is down. So what does that mean for spring season? We get all that and try to factor that in, and that does impact our market decisions on a day-to-day basis, but it doesn't really impact our tech roadmap at all. We know what we have to build.
Yeah, I'll add that, we're sitting on over two million pre-approvals over the past couple of years that we've issued where the people have not found a house. So, we don't know when the dam breaks, but when the dam breaks, we're going to be well-positioned. The number of people that come to Better.com and get pre-approved per month is percentage points of market share, in terms of the number of people that are shopping for a home. And the number that actually convert is basis points of market share. So and a lot of that has been availability homes. And so we are really hoping that if the tide turns on rates or home affordability or availability, we are in a position to meet that demand in the same way that we met the demand in refi in 2020, but without the staffing costs that we incurred in 2020. So I think that's why you continue to see us lean so hard into the AI.
The next question comes from Will Brunemann with North Coast Research. Your line is open.
So I wanted to ask, how long do you think it will take to get NEO back to its former run rate volume as Tinman gets ramped up, and you mentioned Betsy driving cost efficiencies going forward. When do you see the majority of those further cost efficiencies fully realized? And then I have just one more quick question.
Sure. I think we're looking for NEO to get back to their original volume in the next couple of months.
Yes, Q3, it'd be back to where they were and then Q4, hopefully be better than where they were.
Yeah. And I'll tell you, like we announced NEO, we were at some industry conferences, demonstrating Betsy, and the number of loan officers with large retail books reached out to me on LinkedIn to say, hey, can you join the number of people that reach out to Ryan and Chris and Danny were the principles at NEO is pretty long. So we're just making sure that we can fully broaden out the product sets that are in Tinman. We're making sure that we can fully serve all of these loan officers. We're helping them go from driving like a Ford Taurus in the existing infrastructure and mortgage lending that they're on to driving a Ferrari, and we just need to make sure they can do it, and then they're off to the races. So NEO is going to be back to its original loan volume in a couple of months, and then we're going to grow that dramatically from there. And then the other part about Betsy, we're going to keep you updated on the percentage of consumers and the percentage of consumers that are interacting with Betsy and therefore dramatically reducing costs from sales, and then we'll keep you updated on the percentage of loans that are being underwritten by Tinman.
You can compute just the cost savings from that. And then the remaining loans are still subject to the old cost structure. But we'll keep you guys updated on that. But the cost savings are pretty significant, and they're starting to show. They're going to show this quarter, they're going to show next quarter, they're going to show in the net third quarter. As I said, we think by the end of 2025, 75% of loans are going to be underwritten by Tinman AI. And with respect to Betsy, we've got one state that's all Betsy. So you take that, and then we've got to get to 50 states that are all Betsy.
Okay. Great. And then my last question, what do you see as the expansion opportunities within the broader distribution retail channel like NEO?
I think it's massive. I think the number of people that we have reaching out to us to scale. Look, like on direct-to-consumer we’re fighting against some pretty sophisticated books. We’re fighting against Rocket. We’re fighting against LoanDepot. We’re fighting against a number of people that have invested in technology. In the retail channel, we’re fighting against effectively loan officer staffing platforms that have benefited or mortgage broker platforms that have benefited from the lack of any technological sophistication within that universe, and I think there’s a pretty heavy tax that these platforms charge to the incumbent loan officer, and I think we can free them from that. I would say the closest example to that is what has happened in the RIA space, right, versus hanging your hat at an old wirehouse and getting a percentage, a small percentage of revenue. You could go to the RIA platforms that are tech-savvy and private-labeled and give you basically everything and keep a larger percentage of your profits. I think that’s the disruption that is going to happen in retail, and I think we’re going to lead that disruption.
That's awesome. Thank you, guys, for taking my questions.
The next question comes from Brendan McCarthy with Sidoti & Company. Your line is open.
Great, thanks, and good morning, everyone, and thanks for taking my questions here. I just wanted to start on the corporate cost reduction side. I think you mentioned you mentioned corporate cost reductions benefitted 2024. Just curious as to where you see the biggest opportunity for further reductions in 2025.
Sure. So, there are a couple of areas we're going to continue on comp and benefits now, and on the GAAP line item, right, everything's blended together, whether it's corporate or sales UK, everything's going to blend together, so we can, we'll do, we'll break it out going forward. You're going to continue to see benefits there. We are sitting in an office building right now that's 20% the size of the one we were sitting in on the last earnings call. We're saving $10 million in lease expense over the next couple of years just by moving offices. So you're going to start to, we're dramatically renegotiating vendor contracts. Some we got done in the second half of ‘24. So you'll see a full year benefit in ‘25. Some will get done over the course of ‘25, so you'll see a full year benefit in ‘26. And so it is really on all, and it's on all areas. If you look at just even G&A, right year-over-year was down 70%, and G&A quarter-over-quarter was down $2 million Q3 to Q4. So we're in professional fees would be the other. Those are kind of the four big areas. There’s obviously comp and benefits the biggest area. G&A, some of that is in technology expense, some of that's in G&A—that the vendors—and then obviously you’re, most of the lease work's already been done, but you're going to start to see the run rate benefits this year.
Great. That's helpful. That's helpful. Then what opportunities do you see for additional B2B partnerships in the market, just given the strength of the growing technology stack for your company?
Yeah, so I'll give you an example. Three examples. We launched Betsy in October. In the past couple of months, we’ve gotten a call from the CEO of a top-five servicer saying I want to build a recapture business, and I want it to be all AI. Can you private label Betsy for me and therefore private label Tinman for me and help me do that, right? So we're in term sheet discussions with them. We've got a top three financial services lead-gen company, right? Where again, the CEO called and said, I've got a Board meeting in four weeks and I need to show them, you know, AI needs to be implemented, right? And so, what can we do? So we said, sure, happy to do it. We’re going to be out there with their solution, and it’s going to be live and it’s going to be proprietary to us. We've got a large community bank that wants to, again, do the same thing and have it built around non-QM loans, bank statement loans, and non-conforming loans, which, again, are very hard for your traditional mortgage software to do, and we're in late-stage discussions with them.
So we're seeing the B2B side of it explode in a way that we haven't seen since sort of 2018-2019, when people were coming to us to help them build refi engines to address the refi demand. So, but we're not going just with the big guys, which was our strategy after Ally, Amex, and others that were on the platform. We've now taken the technology that we built for Ally and Amex and allowed us to get up and running with a B2B partner in three weeks, and that's a lot faster than what even a sales contract life cycle is for signing up with the traditional players in the industry like Ellie Mae or Black Knight or others. So we're aggressively pursuing that. You’re going to see us hopefully announce some big things in the coming months ahead.
Got it. Thank you. And last question for me. What do you expect the impact of losing the Ally business? I guess, how do you kind of gauge that impact? And what are you doing to ultimately offset that loss for the business as a whole?
Yes. So if I just look at—well, two key metrics. I'll just start with two and then let me know if you want to expand. On volume, you're just short of $1 billion that was in last year's numbers that you're not going to see in this year's numbers. There will be some Ally volume in Q1, but it really tails off after this quarter. NEO more than replaces that, and we guided up to low to mid-double digits growth in volume this year, so we're going to grow volume despite losing $1 billion. We're confident in that, and NEO is a big part of it, obviously. And on EBITDA, it's really a neutral event. Net-net, the way we had structured the fee relationship with Ally worked extremely well in a very good low-rate environment. But you have a fixed amount of people you need to put against a large B2B partnership like that, and as their volume came down along with the industries, they pulled back on marketing spend, etc. We totally understand why they did it, you were running basically an EBITDA-neutral business, and so there's no negative impact to EBITDA because we've obviously addressed the expenses associated with Ally as the revenues come off.
Thanks, everybody. Yes, absolutely. That's helpful. Congrats on the progress.
This concludes the question-and-answer session. I'll turn the call to Vishal Garg, CEO, for closing remarks.
Thank you all for all your great questions and for continuing to support us as we build America's leading AI mortgage platform. And in doing so, help consumers get a mortgage, get a better rate, have a better process, which lets them have a better house in a better school district with a better commute and a better backyard. And we started on this journey eight years ago; the past three years have been really difficult for us, but we're playing offense and playing offense hard again, and we're looking forward to driving the business and being able to share more positive news with you in the quarters ahead. Thank you so much and thank you for believing in us.
This concludes today's conference call. Thank you for joining. You may now disconnect.