Track 5 – Tech Stack Must-Haves & Wish Lists For Lenders & Servicers

Lead. Loan. Closed loan. These three primary customer service objectives across originations and servicing ultimately drive all tech stack strategy. Your tech stack must cover these areas, but tight budgets often limit all-in tech investment. So what's your best approach? This panel of lenders and servicers shares how they're making tough need vs. want decisions.

Transcription:

Young Pham (00:07):

Great. Well welcome everyone. Hopefully you're very well nourished after lunch. We are pretty excited for this last of our panels today that CIT has been hosting. I'm Young Pham, the Chief Strategy Officer. This part of the conversation is really I think the most exciting part, which is really the tech must haves. And so with the challenging environment we have going on, we have three of the best technical minds out there and so I'd love them to introduce themselves. So with that, Paul?

Paul Akinmade (00:41):

Sure. I'm Paul Akinmade. I'm the Chief Strategy Officer for CMG. I've been there about eight years, marketing BI and technology roll up to me.

Daniel Cowles (00:52):

Dan Cowles, Senior Vice President of software engineering over at Guaranteed Rate. I'm responsible for a majority of the engineering teams over at Guaranteed Rate. It encompasses the mortgage engineering, so a lot of the consumer and the point of sale systems. Prior to that, I was at Loan Depot and I worked on some of their mellow products over there.

Norm Steeg (01:16):

I'm Norm Steeg, one of the Founders of Loan Depot original founders. Today I run our purchase lead platform. I run our real estate company, Mellow Home, and I'm the business lead on all of our non-mortgage products like our solar and home security offerings.

Young Pham (01:33):

Alright, you ready? Norm, we're going to start with you first. So each one of you work in multi-channel shops as part of that, and when you think about your technology strategy, how do you approach that? Are they specific to the different channels or do you look at more of a unified ecosystem for that? Why don't you take it off?

Norm Steeg (01:55):

Yeah, so if you look at the history of Loan Depot, and I'd say that's probably been the biggest change that was unanticipated by me, which is we started as a call center and getting technology really to work with large call centers was the challenge. Trying to connect dialers into the LOS and have that all be seamless, a great consumer experience. Then we had retail and that's its own beast right now. We had to go from a highly centralized to decentralized model. Then we have JVs, it's its own challenge, and now we have a real estate company, non-mortgage products, we added servicing. So yes, if you're now trying to, after you bolt all those things on, try to take a step back and say what is the consumer experience? That is the thing that we're, one of the big things we're solving right now is how do we increase that engagement, improve that engagement, because really all of our technology over the years was born out of a need or a concern or a frustration through the user or the consumer. So even the mellow product, which a lot of people here worked on, really good start with loan officers saying this is not working for us. So that's how we do approach that stack is say, given our customer and all the different channels and all the different experiences, how do we create the best customer experience including the consumer, the loan officer, et cetera,

Young Pham (03:23):

Dan?

Daniel Cowles (03:24):

Sure. So at Guaranteed Rate, we've got, our primary channels are retail, JV and a little bit of the consumer direct side. We focus a lot on a core set of technologies and platforms that we've implemented custom. Other things that are say ancillary that are utilized by the various channels, but our core technologies are all around our custom solutions that we've done and it's not a one size fits all as we know. And so what happens is those core technologies, we change some of the experiences a little bit, tweak them a little bit, change some of the workflows for some of those channels.

Paul Akinmade (04:08):

At CMG, we have, we're distributed retail first, so that's kind of who we cater to. We have TPO with wholesale and correspondent. We have a broker dealer division and we're launching consumer direct, which doesn't seem like the best time to do it, but we're going to be doing it anyways. Our stack, it's really the system of record controls everything, and then we build off of that to differentiate for the different segments. I think when we're looking at consumer experience or LO experience, we cater to the people we have to cater to. So for correspondent obviously it's just, it's a bid tape, lowest cost wins. But that is where we probably do some more interesting things for cost reduction automation, things that require RPA auto indexing because those processes for TPO or more centralized, we're able to really take advantage of that and that's where we invest our time there. But when we get to the distributed retail side, it's really trying to create as much flexibility to account for the decentralization that occurs in that market, whether you're catering to, we also have JV too, so whether you're catering to a joint venture loan officer, a retail distributor loan officer, or a call center loan officer, they all have different kind of flows and workflows of how they do their business. So you really have to account for that.

Young Pham (05:15):

So Paul, shifting a little bit, you've touched on the different channels and they're very distinct on their own. When you think about the customer life cycle and really moving from that originations to servicing and how do you think about or that unified experience so that the end customer actually can have that connected experience? What are some of the strategies you're using there?

Paul Akinmade (05:40):

So it is a long journey. I'll tell you that. I put a slide in front of our CEO four years ago and that put me unfortunately in charge of technology and it was talking about the on-ramp. You never get off. So I think as originators we do a really, really good job from application to funding and then you lose the customer. If you sell it, you've lost it. If you're sub-servicing it, they're angry. If you're servicing it internally, they may be angry at you, but there's not really a good system to kind of create that connective tissue. One of the things that we really focus on at CMG is not just education but nurturing leads early in the process, making that process sticky. So what I would define it as from a lifecycle perspective, you've got your in-process, nurture period. Those are things that you can do to move really high in the funnel of attracting a customer, home search, home buyer education, things that go ahead and where you create value for them, where they trust you and we have home funded, which is a crowdfunding platform, then you get into that app to funding process.

(06:32)

You just got to be fast and direct, get 'em through it, but I think everybody always drops off when you get to servicing. So that's one area too that again, we kind of saw what I'm doing on the screen, if you guys were in that presentation, what the LO can do, we're going to surface that data to the consumer in our side of our borrow portal. Now the big challenge that I found is because there's so many disparate systems, taking over management of the identity is key for that because if you're taking over management of identity, you can create a hub that you can push people into different experiences with different partners and different technology partners, which is again, from that perspective when I'm talking with some of the providers here, first question is, you guys have, right? And how's that set up? What can I pass? Do you have deep links? You not have deep links, things like that. So sorry for a long-winded answer, but it's an area that I'm semi passionate about.

Young Pham (07:13):

Yeah, just a follow up question to that, Paul. Is it different now that we're in a purchase market say as opposed to a refi? When you think about that servicing component and how you're treating customers differently there?

Paul Akinmade (07:25):

I think in either case you have to create value because if you're creating value, you create loyalty and that gives you an opportunity, whether it's getting a person early in the nurture process to be a purchaser or to get somebody who's a past customer to be a refinance customer. So when you're thinking about your points of entry, they don't need to start as an originator as in the nurture side. They can start in the origination side from a referral. Additionally, I'm looking at ways and techniques that we can take somebody who was a customer from a different mortgage company but has a relationship with the loan officer and bring them into that funnel as well too. And you just put them on that wheel where hopefully we do enough to go ahead and earn their business and create loyalty where they become a customer again in the future.

Young Pham (08:02):

And Dan, what are you guys doing at Guaranteed Rate?

Daniel Cowles (08:04):

So we're not really big on the servicing end. So what we have done to make that, I'll say transition from origination to funding and to servicing through our consumer, essentially our consumer portal, they'll go in there. We've implemented a lot of single sign-on so that at least the login and going through to the servicing partner is seamless. They're not having to remember a different login or anything like that. It's just they log into our portal, they click into the servicing area and they just go the transition to the servicing portal. So we do try to do that into your thing about engaging with the consumer. One of the things that we've done over the last year and a half or so is that again, we're doing personal loans besides just standard mortgage products, we're doing personal loans. And so that brings those customers back in when they log into their consumer portal. If they funded with us, there'll be references to that. There'll be references to insurance services. So we do have an insurance service division so that they can get homeowners insurance quotes, also any other property insurance quotes. So we do do that and then we are working on some initiatives and there'll be an announcement in the next week or two about some product launches to further engage our consumers and have them come back in.

Norm Steeg (09:23):

Yeah, same thing. Top of mind in my world. And we really want to get at least a portion of our platform to go beyond the transaction based model. And there's nothing wrong with transaction based model. And once you get that done efficiently, obviously then you have a platform that works. What we want to do is get to that next phase where we have more relation with the consumer through their home ownership journey. And that's been a lot of my work with Alec Hanson over in our chief marketing Offer officer. We think there's really an opportunity on the marketing side of that to bring awareness to the consumers of what we can do more. And we think that our me home platform, for example, doesn't constrain us to be about loans. We can really talk about real estate, we can talk about real estate ownership, but still be part of the overall ecosystem of Loan Depot. So yes, bringing servicing in I think gives you at least an opening into that window. And if you do it right and you provide engagement on that platform, obviously it's easier to leverage into these sort of conversations and somebody's going to solve this at some point. I know we've been talking about this a long time as lenders, but I know all of us really want to get beyond a transaction every five to six or seven years.

Young Pham (10:51):

Norm, just following up on that, how important is the marketing technology in terms of now how you're thinking about your stack and really starting to augment that component?

Norm Steeg (11:01):

That is becoming a big part of what we're doing. So we just started a campaign called Mellow Minute, for example, that's going into social media. We're doing some learnings on that. That's been very interesting. Alec brought up the point that we really have somebody sitting there engaging with a customer right now as they're responding back on social media. Then we said, well, why don't we just give them access? They're online access to just go ahead and click something and engage nothing. I mean, it literally killed the conversation. So that might be a little tech lesson for all of us, but you have to meet people where they want to be met. And obviously they told us pretty clearly that they still want to engage and they're not ready to go to that next step, but all those sort of efforts are going to mature that platform. We're going to gain learning through what the consumer wants.

Young Pham (11:54):

So falling on that question, and we're going a little deeper, gentlemen, so whatever you feel comfortable in, this is really the confession in front of the group. So let's talk need versus want. And in that question it's really where is your tech stack really strong and adoption is great and really performing well versus where their opportunities are, certainly where you're a little weaker and Norm touched on a little bit of that, but would love to hear from you, Paul, just following along that thread there.

Paul Akinmade (12:27):

Sure, So we're earlier in our journey of rebuilding our point of sale. We just recently got to a place where we feel like we've got good parity with that, but we've been able to really focus on things that drive the loan officer to be more efficient. So self-disclosure is one of those. We talked about our opportunity funnel there. I think where we're really, really strong in addition to that is from marketing technology. So our ability to customize and personalize, I would say we try to rival other verticals where that's key e-commerce. So retargeting kicks in. One of the big things for me is understanding the customer journey. I'm a data freak and again, BI rolls up to me. So for me, I want to know when a customer accesses a website, what do they engage with? Am I able to feed that back to the loan officer when we're in servicing, how many times do they log in? How many times do they view an email as they move through the funnel, how are they behaving? And I think when you have those types of insights and you can pass 'em back to the salesperson, you're positioning 'em in a way to not only be more effective, but if you also take advantage of that from a development perspective, you can curate a very personalized experience based on how they behave. So that's the area that I think we're strong in, very strong in and where we're making more investments.

Daniel Cowles (13:42):

Dan and I was going to say, one of the things that and areas that we can improve on is that what you just touched up on a little bit there is the customer touch points. Just like about lot of organizations, there's a lot of siloed systems and they're keeping track of customer information and they're keeping a track in a non-standard way. So if young, we wouldn't know if you went to one of our sites, one of our applications. So all of that correlation and deduping is very complex and very difficult. And so we're working right now to try to combat that a little bit and then have this unified model of what the consumer is doing, how we're touching them, how we're communicating them, how we're sending out marketing information. Because right now it's very siloed and we're on just, I'll say in the last year, we're starting that journey and it's not going to happen overnight.

Young Pham (14:32):

Norm, what about you guys?

Norm Steeg (14:34):

Yeah, we've always been, I think pretty darn good at lead acquisition, lead distribution, that world being born out of that kind of call center environment. So I'd say that's pretty darn good. Our point of sale system has good reviews and good adoption from loan officers, so I think that's pretty strong. I think there's opportunities in efficiency in the rest of the process when it comes to operations and we've got a few things in flight there. I think servicing is probably too new to rate. It's a fairly new platform, so I can't really comment on the opportunities there, but I think the front end actually, we do a pretty decent job and I was kind of born with a series of very long meetings with loan officers over months and months and months and months. And that product team myself, I was in a lot of them, when you build something that you get that level of user input from, you usually come out with a pretty good product, right? Then it's just about execution. You've gotten all the data points. So we continue to develop, we continue to get feedback on it, but I think generally since that group had such a say on what they needed, that's been pretty well adopted.

Young Pham (15:49):

Yeah, Dan, I want to go back to you a couple of the key themes that sort of between you and Paul and Norm touched a little bit, the data piece. Where are you guys on that journey? And I think people are out there, everyone sitting around, has anyone cracked the code? Is everyone just where I'm at? Am I at the 50% level or hey, this is going to be the next 10 year problem that we continue to solve? To be truthful for us, again, we're on a journey at this point on the data end, Guaranteed Rate's been around for 25 plus years, a lot of legacy live data all over the place. And so we are working on that. We've got initiatives around that, but it's not going to happen overnight. And we're integrating, we're doing custom systems. We just launched, officially launched our point of sale. That's a custom application and so it allows us to unify some of that experience and some of that data and then going forward, we will continue with that.

Young Pham (16:48):

Paul?

Paul Akinmade (16:50):

I don't think you're ever done, but I don't think you need to be perfect. I think the big challenge I always hear when people are talking about big data is what are you going to do? Just throw it all on Snowflake and see what shakes out and then you'll go through the unstructured data, find stuff. I think you need to be more purposeful. I think you need to be able to say, hey, in this experience or this opportunity or this engagement, what is it that we want to keep? And if we keep it, how are we going to use it? And I think you need to go through that exercise across the board. And I think that exercise needs to be recycled quarter after quarter after quarter because things are going to change, needs are going to change. So I'm not of the opinion that everything should just be thrown up into some gigantic data environment that you can just extract everything from because I don't think you, I think waste a lot of money moving data, but I do think you still need to retain it because you'd be surprised what you find four years from now that you wish you had.

(17:41)

So we have a very good retention policy, but when you're trying to activate it for analytics or actions, I think you really have to be purposeful of how you collect it and how you use it.

Daniel Cowles (17:49):

I was going to add Paul, that one of the things that you have to keep in mind is if you can go for the big bang and then say, okay, we're going to analyze all of our data throughout our whole organization and then 2, 3, 4 years ago from now, we're going to provide our solution. Generally those big bang implementations don't work. They fail. And it's more of an iterative process where you start taking bite-sized chunks and going through quarter by quarter by quarter, year by year, and then you start breaking things off and then improving, I'll say. And so it's very iterative to your point, a process versus let's go and burn everything down and then bring something up.

Young Pham (18:37):

Is that a fair statement in terms of the approach and how you guys are doing technical Implementations, It's constant Incremental builds.

Paul Akinmade (18:47):

It is somebody, well, it's funny that you mentioned what you mentioned because I go to conferences as well too outside of our industry to try to learn where I'm the dumbest person in the room completely. So I was at one a few years back with a CDO Chief Data Officer and was the chief was the sixth data officer for the company in the last seven years, and she was on stage talking about the needs, the reports, and my first question is how have you validated ROI and how do you present that data to sales and what actions do you take with that data and the kind of, well, we're still figuring that. I'm like, you're going to be another eighth CDO coming up here. I think from my perspective, you're looking for this big bang, you're hiring a bunch of data scientists hoping to pull something off, and if you throw enough people in the room, you're going to expect something that's magical. I think you have to walk in with a purpose. And I think when you're also developing technology, it's the same approach. Somebody described as to me as making components of the Mona Lisa, so go ahead and showcase the component that you can develop today and iterate from that point and just get something to market so you can test your hypothesis and see if it holds water.

Young Pham (19:48):

And Norm, you guys have been Successful in some of this, what true on all accounts or just in certain areas?

Norm Steeg (19:54):

No, I think that's a truism right there. If you take two or three or four years to develop something, I think if we would've done that three years ago and now facing this environment versus the environment we were three years ago, you just run the risk of the business turn around looking at that, and you're constantly having to stop halfway through that three-year project and reset your priority. There's just too many changes. So yeah, a hundred percent on that.

Young Pham (20:25):

Okay, So Dan, you're going to be My Guinea pig. This would Not be a tech stack must have, unless we brought up Generative AI. So hype Here to stay, there's some valid use cases. What do you think?

Daniel Cowles (20:42):

Definitely here to stay. It's just trying to figure out how that they fit into our current technology stack, our current implementations. It was interesting, the demo that they had with the AI chatbot. So we've been playing around with that honestly for a while. We've got a whole small team, not a huge team. We have a small team that all their background is AI and machine learning and so we've been going through different, I'll say proof of concepts, test cases and stuff like that. At this point we're testing things out and we're trying to figure out where those things will work and there's a lot of interesting things that we're demoed out there and we're looking at 'em at this point, but we're not wholeheartedly jumping in at it. We're trying to figure out how they fit in our processes.

Young Pham (21:32):

Norm, What about you guys?

Norm Steeg (21:33):

Yeah, it's not a fad. It's hard to call something a fad if lawmakers are meeting trying to say how do we control this? That being said, it's a ways off. I think there can be some use cases that maybe you could deploy and probably in the near future, but we're not talking to business leaders and I ask them about AI and I ask them to describe it to me. I'd say three quarters of the time if not more, they're describing either machine learning or OCR technology. So if you're trying to go to your executives and say, Hey, we got to get in this space, or we have this application and they fundamentally don't understand it, there's going to be a tendency to put it off. So I think it's going to take a little more time to get any sort of wider use. And we also have to take our counterparties into account. We have regulators, we have investors, we have the agencies, and they're going to have to go through their own sort of adoption, whether it's culturally, mentally risk, all those sort of fun things. So definitely an opportunity, but it's going to take some time. And I've been in the industry 30 years and there's everything like this that has come, always took longer than I thought it would to be a major part of what we do and it's because of all those counterparties

Young Pham (23:03):

Norm. Just following up on that, And before I come back to Paul, Paul entered this concept for that other conference. How do you fund something that is a technology that may not be vetted versus some of the mainstays that I think are much easier to describe relative to your business cycle? So you talk about the issues with the regulators and everything out there. So is it as you build these proof of concepts Or Ideas around that or is there any approach that's slightly different based off the justification of it's here to stay?

Norm Steeg (23:35):

We've taken different approaches. I've been a part of let's say a lab type concept where we took a handful of us created kind a microenvironment and just start testing things and see if we can prove them out and then bring out to the business and say, here you go. I think there's some things that maybe if you're good enough at building a deck and explaining what you want and it's fairly clear, you can make the case, especially with ROI. So it kind of depends. I mean I've done it multiple ways, it's worked different ways, but if it's new enough and hard enough to understand, I like to try to prove it first maybe in an environment that doesn't impact the business.

Young Pham (24:15):

Dan, anything different on your funding cycles for it?

Daniel Cowles (24:18):

What we do again is a proof of concept and then what we do is we will sandbox it, look at the results of what's coming out of the proof of concept, compare it to the real world results, and then once we're comfortable with it, we could beta test it in a small little group through either AB testing, through a Beta testing, through some of our small JVs that don't get as much volume and stuff like that. So that's what we've done at this point is just test things out, see what works and then come back and then we'll continue to refine it.

Young Pham (24:51):

Then Paul?

Paul Akinmade (24:52):

So one in this area, I am a little bit of a nerd where I dive in myself. So when we're talking about natural language and when chat GPT came out, I was using it for my marketing department the next day I was vetting everybody. I was doing ad variable testing and stuff like that because it's really great at coming up and creating ideation. Same with the visual imaging. You can go ahead and create something very, very fast when you're trying to create a concept or get an idea in your head. So from my perspective on this, it's an amazing co-pilot, right? It is something that can inspire you to move faster, you can get operational efficiencies out of that. I've seen that with my marketing department already. Now when you're applying it to the technology side of the experience before Chat GPT, we've been experimenting with chatbots and AI and what we've tried to do, we took three mortgage 101 books and had writers turn 'em into question and answer format and then loaded those in and then we behaviorally modified based on answers that we liked versus didn't like, and that was our patent app right now, that same application as an API into Chat GPT.

(25:54)

So now the natural language picks it up better. So if you're looking at it for doing education use cases, it's amazing If you're going to use it for decisioning, I don't think it's there yet. I do think that you could leverage it in a certain way where it's read the appraisal report, tell me if it's okay, and you could give prompts to it to say, here's the things to look out for as a negative. So from a QC process, I see that as something that could be deployed. Now effectively, I do think you would need to trust it, and that's why I like it. It's a QC process first because if a QC process works over and over again, that's your staging environment. You've also justified the cost by going ahead and eliminating some of your risk, especially if it starts catching stuff. And then the better that gets, the better that opportunity is to deploy it as an actual solution instead of QC.

(26:33)

But I do think it's going to take time. But again, I was watching CNBC this morning, they're talking about Chat GPT version seven. So this is going to evolve fast and I think from perspective, I know it's going to be painful. You're the first to the jungle, you're going to get smacked with branches in the face, but if you're going to be willing to carve that path, you have a chance of getting that first, but it's going to suck. It's going to be painful and it's going to be expensive. So find the small use cases and look for your wins in very small areas.

Daniel Cowles (26:57):

For your Chat GPT. What we've done to try test things internally is we ended up implementing a chatbot for self-serve it, reset my password, those kind of things that aren't that big. We're just refining the learning through there and Internally and so we can take some of those learnings and then potentially use them later on.

Young Pham (27:18):

Yeah. What was interesting, you said Paul, and I'd welcome your opinion across the board, you talked about the co-pilot for marketing, any of you guys using it relative to software delivery itself. So outside the QC or the loan process, how do I deliver technology faster?

Paul Akinmade (27:34):

So I'm not going to give you a shout out, but CINT, you guys do have a co-piloting software that you use with your developers that I'm trying to get my developers on. So if you were willing to share that, we'll make that deal now on stage.

Young Pham (27:44):

I think we can do that.

Paul Akinmade (27:46):

All Right, cool.

Daniel Cowles (27:47):

Yeah, for our development staff, it's more of a test at this point. Individual teams are testing out different things with some of the stuff out there for, I'll say the code analysis that's out there. So we've been testing things, but that's all we're doing at this point.

Norm Steeg (28:06):

Yeah, we're kind of that same. I like the term co-pilot. That's Probably in the arena we're at today.

Paul Akinmade (28:12):

We did have a developer try to cheat on one of his reviews by submitting a chat GPT code, and we were unsure, so we asked chat GPT, did you write this? And it told us it did. So that's the one thing you don't know about chat GPT. You can ask it if you believe it's right or wrong and it's answer and it'll answer you. Was this a good answer? No, it's pretty shitty. Can you give a better one? Let me try. So the challenge with chat GPT, when you guys are using it in its native format, it's hallucination setting is to always give an answer whether it knows one or not, but you can ask it. Was that a good answer or how would you change that if you were under these circumstances?

Young Pham (28:43):

Yeah, yeah. So final question, future outlook of tech. We've covered some of the things that are strengths, weaknesses, generative AI. What else are technologies that you guys are really excited about that you either think are just emerging or in some essence a bit of a dinosaur that may fall off the technical consideration? So Let you think about it. I'm looking at who's my first victim. I'm going to go with Dan first.

Daniel Cowles (29:17):

For me, there's a lot of vendors out there coming up with a lot of different solutions to speed up and automate the mortgage process through the whole origination, underwriting. Everything else, from what I've seen, we're not quite there. There's a lot of good things that's being developed, but for the full, I'll say effort, seamless experience from a consumer's point of view, where we'd like to get is again, just like if you're doing a credit card application online where someone starts their application and then it's funded and improved in 15 minutes, 10 minutes or something like that, that's where we'd like to get to at some point. I think the technology's getting there. We're not there quite yet, but we're getting there. And so that's where if you look at the industry as it was even five years ago, we weren't even close to that. And so that's where we can get pretty exciting. Obviously the thing that's a little scary about that also is that if we make it that easy, the fraud part of it is also, and the identity theft is also something that we have to keep in mind to make sure that a consumer applying for this loan is the consumer they are. And so we are looking at those technologies too.

Young Pham (30:34):

Yeah, Norm.

Norm Steeg (30:36):

I'd say there are definitely things that are improving the consumer experience on the front end. The things that we were really talking about four or five years ago I think are getting developed to the point like OCR and machine learning that things that probably would've made us pretty nervous about decisioning right off of a document and whatnot when these things are around early on, I think are going to become pretty standard here pretty quickly. And I think that'll change a lot of the game. So some of the timeframes that we talk about trying to get to X stage within a certain amount of days for approval, whatever it is, you're still dealing with someone at the other end consumer that's having to provide you information. And a lot of times that information causes you to have to redo essentially the whole thing. They give you a document, you said, oh, we're really not this, we're now this, and you have to restart that process.

(31:34)

So most manufacturing processes don't have that challenge. If you're building a car, someone doesn't jump in the middle of that production a consumer and say, oh yeah, by the way, I really wanted this instead of this, and suddenly you're sending that back. So it is a complicated process. I think everybody here that's really involved in all this, it's amazing that we have the technology we do to do what we do. It is complicated and there's so many variables. I wonder if you could mathematically get to a number of how many variables we have to do. But I think things like that could really, even though it might seem like a process thing, I think for the consumer, if we can make those changes quicker, get the decisions quicker, then you get to an answer quicker. And that's going to impact the consumer experience, I think.

Daniel Cowles (32:24):

And following up on that Norm, you talked about how your comfort levels change. How much of that is the technology's available? It's our maturity as an organization to adopt it versus the tech's just not there yet.

Norm Steeg (32:38):

Yeah, it may be 80% on our side, and part of it is the operational culture. You talk to a salesperson, they'll say, this is the way sales works. You talk to a processing lead, this is the way processing works, the underwriting this way. You don't understand this is the way underwriting works. And the people feel pretty strongly about that. And so that impacts adoption and you have to get all leadership on board to say, this is going to be fundamentally different and your world is going to change in this way and we're now going to have to change our leadership style into this new thing. And so even if you get the buy-off at the highest levels, you're still kind of coming down the pike into the different verticals that you also have to say, this is our direction, this is where we're going. So I think that has been one of the biggest hurdles.

Young Pham (33:36):

Paul, Question to you.

Paul Akinmade (33:37):

So I guess I'll start on the things that are the gloomier parts. So for me, every single origination system that is out there today is just legacy. It feels like whether you're, I can't tell you how many people I talk that are switching from a power to encompass, encompass to empower and so forth like that. And the challenge that I see is you've got those well-established systems that have this high maturity curve and high network effect where they've got all the integration partners and stuff set up. And then you look at these up and comers and they're facing major headwinds. They're not on the enough on the maturity curve. And depending on if you're doing vanilla products or if you're with us, we are like the TGI Fridays of mortgages. So every flavor you want it, we'll figure a way to stick it in the system.

(34:21)

So having a highly configurable system for somebody who's going to have that type of efficiency, it's tough to move up that maturity curve and also have the network effect with all the integration. So for me, that gap is a challenge and it's one that I get worried about. So when we're building our technologies, we're really trying to think through how do we go ahead and build a system that can go ahead and move from LOS to LOS for when those days do come because they're going to come. Things I do get excited about are the opportunities where we can go ahead and reduce friction. So Mark said something earlier, he said then LO wants to hit a button and make money. He just wants cash to pop out. And in my head, it's just go ahead and reduce the amount of touchpoints, right? And if you're good at that process, you're able to find ways to reduce the manufacturing costs.

(34:57)

I think the challenge though I see is there's a lot of things out there that go ahead and make these promises of, Hey, we're going to automate this, but for me, unless you've actually reduced your labor cost, then you haven't realized there's gains. You've just incurred more technical costs with no change in labor. So those are the things that I really look closely, especially when I talk to different vendors. The first question I ask is, in your case study, have you looked at what your HR cost has gone from, what your loan funding has been? There should have been a spread or a closing of that gap based on how effective your technology was. So those things are things that I think about and what I would say just from anybody who's working on reducing that friction, making sure they can plug and play into multiple systems very quickly with open APIs, those are the things that we're thinking about.

(35:37)

Because as we have those questions, the first question I'm asking is, what's your API documentation? I don't want to see interface because that just means I got to stick another system that everybody gets to jump into. I want to automate it into my system. So I think those are the things I do get excited about, but it's really where what I'm hopeful is we get to a place where you're able to go and deploy technology and very quickly take the KPIs, and that was a good investment. When we're doing automation, the people I target is are my outsourcers. Those processes are well-defined. So if we're outsourcing something to another country, it's a well-defined process. I'm measuring it by how many people I still have working in that other country, did our manufacturing automation work or not work. So those areas that I get excited about.

Young Pham (36:14):

I want to thank Paul, Dan, and Norm for participating in the panel and all of you for attending. This concludes our session. Please feel free to come talk to us, anybody if there's a follow up. We also have our booth at 105 right out front if you want to have deeper conversations. And again, Thank you.