In our Hot Topic this week, we have Briana Ings, Vice President of Product at Snapdocs, to talk about a new product that will reduce errors and QC costs for lenders by automatically QCing documents for errors on every closing and packaging the documents for investor delivery, specifically in the context of how important increasing operational efficiency is in light of the current macro mortgage trends.
How Lenders Combat Recent Margin Compression Trends With Briana Ings
It is March 28th 2022. It’s good to have you here with us. This show is created by mortgage professionals. It is four mortgage professionals. We’re so grateful to have you here. Our commitment is to bring you timely information that you can tune in to anytime anywhere. I was in Fontainebleau, South Beach, outside of the war zone. There’s some gunfire going on. We had a reception in South Beach, and guns are blazing. Police are all over. It’s crazy what goes on with these spring break crowds. We were at the Fontainebleau. You can only imagine what that looked like with a whole bunch of spring breakers and convention goers all mixed in together. That was interesting.
Also, I want to say thank you to TotalExpert. I had a chance to sit in several sessions where TotalExpert was presenting. What they have built and what they’re doing with their purpose-built CRM and customer engagement platform is nothing less than amazing. Now you’re able to start tracking what’s going on with borrowers that you have not funded loans that you have leads on, and looking at certain aspects of their credit and when they’re ready to go. They have a credit monitoring feature that they just turned on.
I’m very thrilled to be here with you.
There’s so much going on in the industry and what you guys are doing with your eClose Solutions and everything that you’re creating is creating a buzz in the industry. When we were at the TMC conference, there was a lot of buzz about you guys. First, we want to let our audience get to know you, Briana. Tell us about yourself and your journey to where you’re at now.
I’m excited to meet everyone on the call. As David said, I’m the VP of Products at Snapdocs. My career has been in building products. I got into the mortgage industry four years ago at Snapdocs. I’m a data nerd. I love solving complex problems. I met our CEO, Aaron, and he showed me more about the industry and how complex it was. I knew I would love to solve those problems. Everyone knows, even from the talks earlier, how complex things are. It has been such a fun journey building our digital closing product. I’m excited to share more about the new products we’re launching.
A little bit more about me. I live in lovely sunny San Diego. I don’t know if you’re in a cold environment, but it’s nice and sunny and 70 here. That’s the life I live and I love every type of sport. I’m thrilled that baseball is back. I’ve got tickets to opening day with the Padres. It was touch and go there on how much the season would take off. If you’re into sports, you got to be watching March Madness. My brackets were completely destroyed. If you picked the Peacocks, you are a unique one, making it all the way to the eight here. I wish my alma mater, the University of Virginia, didn’t have such a bad season and could have participated in the tournament. It’s been fun. These games have been amazing to watch.
It’s fun to get to know you a little bit and also your love for sports. I always ask for more engagement when we get into the sports topic. Tell us more about Snapdocs and what you do there specifically.
We’re a digital closing provider. Our whole company’s philosophy is built on this hypothesis and theory that a lot of the closing process and the mortgage transaction overall is challenging because of the fragmentation. As a lender, you work with thousands of title partners. Those title companies work with thousands of notaries, and then you’ve got borrowers at the center of that.
That leads to these manual, opaque, dynamic processes that are inefficient and create a poor borrower experience. Our mission is to change that. We do it by connecting together all of these participants to create what we like to talk about as the perfect closing experience. That’s one that’s error-free, transparent, faster, more secure, and less work. We got hundreds of lenders and title companies that are using our platform to digitize their closing process.
With the interest rates rising and the real estate transactions declining, what impact do you see this having on the market?
That’s what everyone has been talking about earlier. As these interest rates are rising, everyone is seeing transactions decline. We’re seeing this higher mix of purchase versus refinance. With our customers, we’re seeing a few trends as a result of what’s going on in this macro environment. First, these refi-focused lenders are needing to focus their efforts and become competitive in a purchase market, which requires shifts in how they think about winning business.Lenders need to focus their efforts on becoming competitive in a purchase market, and that requires shifts in how they think. Click To Tweet
Across the board, it’s more competitive. There are fewer transactions to go around. Companies are focused on how they can create the most competitive borrower experience, boost referrals, and cross-sell into other lines of business. The biggest area is companies are looking for ways to reduce their cost per loan. How can they increase the team’s capacity? What are these full bodies of work that they can eliminate so they can get their people working on higher-value tasks?
We’ve seen those as blessings in disguise for lenders in trying to figure out how they’re going to solve these problems. It’s that with fewer transactions, they can come up for air after the craziness that was 2020 and 2021. They use this as a time to look at their whole tech stack, evaluate it, roll out, adopt new technologies, and focus on making sure that they’re getting the value that they need from technology across their organization.
It is challenging how to get the value because everyone is looking at where things are at. Jack and I were talking about how you can save money through that. Alice, let’s get over to you to ask questions.
I’m going to focus on operations. That’s a big part of my world. What specific levers can lenders pull to combat margin compression now to reduce operational costs?
When prepping for this call, I talked about it’s a great time to evaluate technology, but technology can also be death by a thousand cuts. When you’re evaluating technology providers and solutions in this world, especially in this market environment, you want to find solutions that are going to help you cut a lot of expenses, especially as you look at the way you operate your business. If you can move any of the fixed costs that you have and variable them, that’s an ideal world.
That’s a great place where Snapdocs and digital closings in general fit in. I talked about the manual comp is riddled with these inefficiencies that lenders have solved with teams of people. If you could replace that with a technology solution for each closing, it helps your operating costs. There are inefficiencies in the closing process in general with coordinating thousands of settlement partners that lenders work with. It’s also time-consuming. This is what I want to focus on with our new product. It is very time-consuming for the QC work to catch manufacturing defects and prep documents for investor delivery.
Digital closings, where a portion of the closing documents are electronically signed, is one way to reduce costs. We recognized that there was more we could do to work towards a world where we completely eliminate errors from closing transactions. That’s why we’re launching this new product, we call it Closing QC, that not only reduces the QC costs that lenders incur, but can reduce manufacturing defects overall. Therefore, the downstream risks and indemnification buyback risks occur when you have those manufacturing defects.
There’s no question that these defects get expensive and the best way to do a cure is like this. Jack, say that you were able to rejoin the dialogue here and you were part of that pre-call that we had. Jack, it’s good to have you back. Sorry to hear that your internet went down. It’s good to have you live via a cell phone.
I was about to ask Briana to delve in a little more about the quality issues that are occurring on a typical file, and what impact this has on lenders that see these repetitive manufacturing defects.
This is the problem we saw across our customer base. This hopefully resonates with a number of the people dialed in, but they’ve got a lot of time that they spend trying to prevent these manufacturing defects, and get documents prepped for investor delivery. We’ve seen a couple of things. One, they’re spending about twenty minutes per loan checking that every document was returned, and that key signatures are on the document.
Also, the correct documents were signed. Maybe they did a redraw and they’re worried that the wrong pages got signed, and they could be sending the wrong terms to an investor. They are then spending another fifteen minutes organizing the documents in the appropriate stacking order for investors, ensuring that they’ve got the appropriate documents that investors require. They got some technology to do that, but it doesn’t work for all of the pages, so they’re needing to sort through PDFs. That’s the time. They’re spending a lot of time on it.
Even still, because this is a manual process and manual processes aren’t perfect, they end up with these manufacturing defects and incur a deficiency rating. We’ve had some data customers using our functionality. In those QC transactions we’ve been running through our system, we’ve seen that as much as 20% of transactions have an important document missing that wouldn’t have gotten necessarily been found. The question is imagine you could save all those minutes and reduce manufacturing defects, what sort of impact and savings would you have for closing and combating the margin compression trend?
Jack, let’s get you to the next question.
Can you talk a little bit about Snapdocs and the place they’re at where they can help lenders succeed in an environment where originations are projecting down and margin is compressing? Why is Snapdocs at a place where they can be a benefit to lenders?
Let me first share the QC functionality that we’re launching and how it’ll solve those problems for lenders. We break down our current functionality into three parts. Jack, you were talking earlier that you’d like to get into the details. I’m sharing this with you so you can understand how it works. We collect the big giant PDF signed documents from the settlement company or notary.
What we first do is automatically classify every page using AI so we know what document it is. We make sure that the name is correct and standardized, and we deliver it to the appropriate folders in your LOS so that it’s ready for investor delivery. We then check all of those documents if there are any important ones that are missing. That will make sure that all of the documents were returned. We then come through and check and make sure that all of those documents were signed correctly. We scan the document for signatures, notary stamps, dates, and checkboxes that often get missed or incorrectly signed. We got those three main parts, and that’s just the beginning.
In the future, we’re also planning to check key loan data, and make sure it’s consistent within the documents like the name and the loan amount. It also matches the data that you have in your origination system to make sure it’s all correct. Also, automated workflows with settlement, notary and borrower to automatically resolve the errors so that we can ultimately guarantee that the documents coming back through Snapdocs are accurate.
I mentioned we’ve had this product live with some beta customers. We’re seeing some real results like increased staff capacity. Customers have realized 3X to 4X increase in closings per closer. Loan capacity and those closers are saving about three hours of time per closer per day, which is huge for being able to run efficiently. They’ve seen a reduction in errors. They’ve been able to catch more errors more quickly and get those resolved, which has reduced their deficiency rating, and time to release funds and sell to investors.
That’s the specific functionality with our Closing QC product that we’re launching. One aspect of Snapdocs that makes us able to deliver this type of functionality and drive these results for customers is this aspect of scale. Oftentimes, you talk to lenders and they can’t implement technology on all of their transactions. You don’t solve the efficiency problems because you’ve got a subset of loans that are going through some old slow error-prone process.
What we’ve focused on is helping lenders get and achieve scale so that you have this one single new process that’s more efficient. We’ve designed the whole company around how we do that. It’s represented in how we design our platform. We think about it not as a technology problem, but also network problem. We got the largest settlement network to ensure adoption, and then we support that with a team. We have an amazing team of experts who have implemented hundreds of lenders. They provide hands-on support throughout the entire journey. That’s what has made us able to deliver a product in this space.Snapdocs is all about helping lenders get and achieve scale so they can have one single new process that’s efficient. Click To Tweet
I have a question. You mentioned document checking. As we know, OCR can get fussy with non-standard documents. This is a two-part question. You mentioned that you virtually have a guarantee on the accuracy. That must mean your platform can support all these non-standard documents that can come in if somebody is perhaps delivering to multiple investors.
I should say the guarantee will come in a future version. The first version we have live now is checked, but we do support non-standard documents. The way we’re able to do this is to leverage technology that we built earlier on in the transaction. There’s another part of our product where we automatically digitize documents for eSignature. In that, we needed to solve for custom documents and non-standard documents. That has allowed us to collect a vast set of documents.
We needed to build algorithms that don’t need to be templatized and can be more dynamic based on what we see on the document. This workflow does allow us to have this automated solution for non-standard documents as well. Depending on the investor you’re selling it to, you can get the right checks regardless of how unique the documents are.
Now let’s get to some of the questions that are coming in from our audience. One of the ones that Alice touched on was one of the questions that come up several times from our audience. One of them has to do with the OCR air rate. Alice said it perfectly on the non-standardized documents. How are you handling that, what is the error rate that you’re seeing, and how does this compare to the competition?
What we’re seeing is we’ve got our AI running and it’s about 99% accurate. For that 1%, we have a human in the loop process. Especially when we introduced the missing signatures, we wanted to make sure that you were getting the report back from Snapdocs. You’d be confident that it was correct. We do have a human-in-the-loop process to review any of those documents that our AI wasn’t confident in, and to have that human review step in there to make sure that over time, we can even improve those models and feed new training data back into our model. This is a very high accuracy rate.
Many competitors, if they have AI technology solutions, they either won’t be able to get to that level of accuracy because they don’t have the specific training data as we have from the hundreds of thousands of closings, or they don’t invest in the human in the loop process to constantly retrain the data to make sure it’s accurate. With those two things combined, it has allowed us to deliver a highly accurate process.
That’s phenomenal, 99.9%. AI is a big component of that. One of the questions says, “How is she doing that?” The answer is AI and the human in the loop. We are always going to have to have humans in the loop as good as technology is getting. Talk more about the AI component. Several people are wanting to know if it is true AI. Is it machine learning? What level of technology is this operating at? AI can be a broadly used wide term. It sounds like this person has a strong technology background, so they want to know whether it’s AI or machine learning or rules-based. What is it?
I wish we had our head of data science on the call. He could answer much more precisely than I could, but it’s an algorithm that learns based on a set of training input data. It allows us to model. We take a set of training data, which is the known output, and the algorithm builds variables that can detect those trained outputs, and then we apply that to new data. It learns as new training data gets added.
When you say it’s learning, that’s AI, folks. Only AI is truly out there learning based on previous results. I love the fact that you have humans in the loop. One of the questions that came in says, “Talk about how you could achieve scale. Could you expand on that? How are you doing that? By the way, we’re Snapdocs users and we love them. They have been such a fabulous partner. Customer service is over the top. Kudos to you.” This is from one of your customers who happens to be a listener.
A lot of it comes down to both the product and the team. Thank you so much for the kind words, whoever wrote in with that. We’ve got a customer success team that are experts in how you adopt and roll out digital closing technology across your customer base. We designed the product such that it can run in the background. There are not a lot of point-and-click buttons you need to click in order to action things. It’s very much integrated with the loan origination system. When things are clear to close, it can kick off the closing getting created in Snapdocs. Once everything is done, they automatically go into the LOS.
We got this team of experts that then helps lenders strategize on how they roll out and adopt the platform across their loan base. Many will roll it out on a region-by-region basis. They might start early on by making it opt-in to just get the lay of the land and understand the new process. Once it works, the most successful thing we’ve seen in achieving scale is changing digital closings to be an opt-out type model, where lenders can use our system and we configure rules to know what type of digital closing it can be.
Any closer or loan officer is auto-opted into that unless there’s a specific reason that they want to opt a borrower out because they want to sign via pen and paper, for example. Those best practices that we’ve seen across our customer base enable our team of experts to work with lenders to strategize on how to successfully roll out the technology across their business, depending on the uniqueness of how it’s set up. If it’s very region based or if they’ve got central ops, they can partner with you to figure out the best way to build that confidence early and then get it to scale.
Alice, any other questions you have for her?
It sounds like you’re using this primarily for post-closing. Is there a use case for this for pre-closing as well?
Totally, especially as we add the functionality to be able to detect if there’s any error in the data of the document. It’s checking that the loan amount and the interest rate are right, and the names are consistent. We can be pulling those checks earlier on pre-closing so that we preemptively catch them before the documents are signed. There are certainly areas further upstream where we can be doing some of these checks to prevent redraws and things like that.
Good question, Alice. Jack, as you’re listening to this, as we get ready to wrap this up, any other thoughts?
It’s the repetitive question that I ask a lot. Briana, what are you at Snapdocs integrated into with the myriad of loan origination systems out there? Have you built a number of integrations so that the user community can leverage this technology?
We’ve taken a very API-first and tech-agnostic approach to our platform from the beginning. That was one reason why we built the AI initially to automatically digitize documents. It’s compatible out of the gate regardless of the doc prep that you use. We’ve taken a very API first. You can even google and look at our APIs online. We’re very much open and tech-agnostic from the beginning. That allows us to work with whatever loan origination system a lender uses. We form partnerships with origination systems. We can also work with lenders on a lender-by-lender basis to get an integration set up for them. It’s a key part of our overall company and product strategy.It’s good to take an API-first and tech-agnostic approach to your platform so you can be compatible with everything right out of the gate. Click To Tweet
I always like tech solutions that eliminate or mitigate, stare and compare solutions.
That’s so true. That came out when you were building the correspondent division there at Texas Capital Bank. We spent a lot of time talking about this very thing. You guys are leaders in the eSignature space. Any new products that are coming out or any developments there that you want to talk briefly about? I’m so excited about what you have done and built, and many others are. Where do we stand with the whole eSignature initiative of what you’re going on? Also, what is the rate of adoption? Is it getting any better?
In addition to this Closing QC product, the other big investment for us on the eClosing side is helping the industry adopt more and more eNotes. One of the big challenges that we’ve heard of is knowing where eNotes are accepted and which investors. What are their rules around acceptance from a remote online notarization perspective? What county, state, or title underwriter? Are they going to care? Are they going to accept it? We partnered with MISMO, an e-Eligibility Exchange, to collect this robust data from the industry.
It was also super time-consuming. Investors are telling all of their lenders what their rules are. Every lender is going to all their investors, so it’s inefficient. We’re collecting all the data together in a central platform in partnership with MISMO. We then bundle that into an eligibility engine that will tell you how E your loan can be. With that type of technology, we’re seeing high adoption of hybrids. That’s easier to know where it’s eligible. We’ve seen a bigger set of lenders that have implemented it and a small set of those that have scaled it. We’re starting to see that start to take off. This eligibility engine will help a lot to let you know and not think about where you can legally do it.
It’s very exciting. We could go on this topic. You guys are leaders in the space. I encourage anyone who has not gone out and requested a demo or spoken to someone at Snapdocs. I encourage you to do so. We had as our special guest, Briana Ings. She is the Vice President of Products at Snapdocs. Kudos. You have a great vision and the team does. I’m thrilled to have the opportunity to work with you guys on a regular basis. Thank you for all that you’re doing with the industry, especially an E everything initiative. You guys are doing so much to make this happen for our industry, big participants, and pushing this important initiative forward. Thank you, Briana, for being with us. I appreciate it.
Thank you so much for having me. It was a lot of fun.
Next episode, we have Jack Konyk joining us from Weiner Brodsky Kider. Jack and Brian Montgomery did a presentation at the Lenders One event in Phoenix. The comments that were made there just said, “I’ve got to get both of these guys on the show.” We’re working on it. We got Jack starting off. I just tease that conversation a little bit.
Jack said it. If the CFPB were to be running the highway department, your speed limit signs would read something like this. To find out what we think is the appropriate speed for this highway, check our website for recent enforcement action. From that, you should be able to draw a conclusion about what we think is the right speed limit for this highway.
How ridiculous would that be? There’s just not a great amount of guidance. This environment is one of the most action prone we’ve seen ever. Paramount to what we’re doing as far as litigation, we’re just seeing a lot of it happening. There’s a lot of concern that’s going to be kicking up, and then you can have the fraud factor that Alice mentioned earlier. It just opens the door for us.
Folks, you will not want to miss Jack Konyk on the show. It’s good to have you with us, everybody. I want to say a special thank you to our sponsors as we exit. I want to say thank you to Finastra, the MBA, Lenders One, Mobility MMI, Modex, Knowledge Coop, the Mortgage Collaborative, Snapdocs, SuccessKit, Lender ToolKit, TotalExpert and FormFree. It’s good to have you with us, everybody. Have a great week. I look forward to having you back here for the next episode.