Latest / Leaders in Lending / The $10B Insight Every Lender Needs

Transcript

The $10B Insight Every Lender Needs · Leaders in Lending

0:00
Speaker 1: It allows us to move from just predicting who will
0:03
Speaker 1: default to actually shaping the default. We're managing more than
0:07
Speaker 1: ten billion dollars in loans today, and while that might
0:09
Speaker 1: seem paltry compared to maybe some of the big five banks,
0:13
Speaker 1: of course it's a really, really large portfolio. If you
0:16
Speaker 1: think about all of those barbers who are delinquent, and
0:19
Speaker 1: you take the ones who are definitely going to pay
0:22
Speaker 1: you back whether or not you talk to them, well
0:25
Speaker 1: then you definitely shouldn't talk to them. Lms are, of
0:28
Speaker 1: course transformative. They're an unbelievable technology layer on our work,
0:33
Speaker 1: and I sort of imagine that they will eventually be
0:36
Speaker 1: diffused into almost all.
0:38
Speaker 2: Of our activities.
0:43
Speaker 3: Welcome back to Leaders Lending, where we explore what's next
0:46
Speaker 3: in the worlds of lending and financial innovation. I'm your host,
0:50
Speaker 3: Lynd Saderbil, and today we're talking about a part of
0:52
Speaker 3: lending that doesn't often get the spotlight. Servicing. It's where technology,
0:56
Speaker 3: operations and customer empathy all meet, and at ups art
1:00
Speaker 3: it's becoming a powerful driver of performance and borrowers success.
1:04
Speaker 3: Joining me today is Jesse Opink, who's the general manager
1:07
Speaker 3: of Servicing at Upstart. Jesse, do you want to introduce
1:10
Speaker 3: yourself for our listeners.
1:12
Speaker 1: Ellen, thanks for having me excited to be here. I
1:16
Speaker 1: joined Upstart a little bit over seven years ago, and.
1:22
Speaker 2: I came up on the product side.
1:23
Speaker 1: I was a product manager number two are our product
1:26
Speaker 1: management function was relatively late to develop, you know, and
1:31
Speaker 1: I've gotten a chance to work across our personal owns function,
1:33
Speaker 1: a little bit in growth, product, a little bit with
1:36
Speaker 1: machine learning, and and now for for the last two years,
1:40
Speaker 1: I've been the general manager of servicing, you know, which
1:44
Speaker 1: I think really combines my passion for great customer experiences,
1:49
Speaker 1: you know, with with my desire to run scaled in
1:52
Speaker 1: and highly effective businesses.
1:55
Speaker 2: Awesome.
1:55
Speaker 3: Well, I think the one thing I think, you know,
1:58
Speaker 3: we we do a lot of unique things here at Upstart,
2:00
Speaker 3: and we do I think a lot of them very well.
2:03
Speaker 3: And this is an area that's been a big area
2:05
Speaker 3: of focus for the past couple of years as you've
2:09
Speaker 3: become the general manager and have been really leading the servicing, product,
2:13
Speaker 3: engineering and operations teams. One of the things you've said before,
2:18
Speaker 3: and I think you may have shared this at the
2:19
Speaker 3: client summit with many of our clients and potential clients,
2:25
Speaker 3: is that servicing is really just underwriting with an operational twist.
2:29
Speaker 3: What did you mean by that?
2:31
Speaker 1: Yeah, you know, as you as you mentioned, I have
2:35
Speaker 1: been at Upstart for I guess seven years last Saturday,
2:41
Speaker 1: so just just past the seven year mark. And of
2:44
Speaker 1: course I had spent the first five years working, you know,
2:48
Speaker 1: very deeply in upstarts core business that is predicting who
2:52
Speaker 1: will default and then pricing them correctly, our our underwriting business.
2:57
Speaker 1: And so when I joined servicing a little over two
2:59
Speaker 1: years ago, I really tried to think quite a lot
3:02
Speaker 1: about what it meant for us to take our core
3:05
Speaker 1: competency and apply those skills to this relatively new space.
3:10
Speaker 1: And I guess the way that I would say is
3:12
Speaker 1: that in underwriting, your job is to take maybe all
3:16
Speaker 1: of these applicants. You're taking applicants, you're sort of lining
3:19
Speaker 1: them up in riskiness, you're spacing them not correctly, and
3:22
Speaker 1: then you're determining generally what risk based price is associated
3:27
Speaker 1: with the level of risk that you see for each applicant.
3:30
Speaker 1: And I think in many ways you really want to
3:33
Speaker 1: do the same kind of thing with servicing. You want
3:35
Speaker 1: to maybe take all of these loans, you want to
3:38
Speaker 1: line them up from least risky to most risky. But
3:42
Speaker 1: in underwriting you have this sort of unfair advantage that
3:46
Speaker 1: is that you can maybe take any person who's risk
3:51
Speaker 1: is too high, and if you don't like how risky
3:55
Speaker 1: they are, well you just don't offer them alone. And
3:57
Speaker 1: in servicing we actually have a little bit more operational complexity.
4:01
Speaker 1: We are forced to maintain our relationships with these customers,
4:05
Speaker 1: and so instead of being able to maybe just say, oh,
4:08
Speaker 1: this person is too risky and we're not going to
4:11
Speaker 1: work with them at all, we instead have to maybe
4:13
Speaker 1: look only at the range of operational levels that we
4:16
Speaker 1: have to change their riskiness today, to move from how
4:20
Speaker 1: risky they are today to how risky they might be tomorrow.
4:23
Speaker 1: And I think of this as a really really powerful
4:27
Speaker 1: part of our overall opportunity at Upstart, because it allows
4:31
Speaker 1: us to move from just predicting whole default to actually
4:36
Speaker 1: shaping little defaults.
4:38
Speaker 2: That's a little bit what I mean.
4:39
Speaker 3: Yeah, Now it's super I think super interesting is really
4:42
Speaker 3: kind of applying the the data science or behavioral predictions
4:47
Speaker 3: after origination, I'm kind of dynamically handling that relationship as
4:51
Speaker 3: you move forward with the customer. I like the idea
4:55
Speaker 3: of what you said of You know, if you identify
4:57
Speaker 3: a higher risk pre origination, you can than their APR.
5:01
Speaker 3: If you identify that after you have to still figure
5:04
Speaker 3: out a way to work with that customer and to
5:06
Speaker 3: maximize the returns that you're getting from them.
5:09
Speaker 1: That's something that we've been actually really focused on in Servicing.
5:13
Speaker 1: That is like, oh, we you know, not only want
5:16
Speaker 1: to be good at the risk prediction part of our work,
5:20
Speaker 1: but we also want to have a really wide range
5:23
Speaker 1: of operational levers. And so one way tactically that this
5:27
Speaker 1: has played out for us is that we really want
5:29
Speaker 1: to have collectors with a range of experience and persuasive capabilities.
5:37
Speaker 1: You know, you might imagine that maybe on the extreme,
5:40
Speaker 1: if you are collecting on very very small dollar amounts,
5:44
Speaker 1: it's it's probably actually not worth it for you to
5:48
Speaker 1: have someone really high end making that call. You might
5:51
Speaker 1: actually even want to stick with just digital only collections.
5:54
Speaker 1: But on the other hand, if you have someone who
5:57
Speaker 1: maybe owes you tens of thousands of dollars, maybe has
5:59
Speaker 1: a a secured asset that they really care about, you know,
6:03
Speaker 1: and and maybe need some coaching to get to a
6:05
Speaker 1: good a good outcome, well, that starts to look a
6:09
Speaker 1: whole lot more like artisanal sales. It looks like a
6:12
Speaker 1: really persuasive and engaged conversation, you know. And and so uh,
6:16
Speaker 1: I guess our work in the last couple of years
6:19
Speaker 1: has been both trying to build out our risk predictive capabilities,
6:22
Speaker 1: the ability to decide which loans are risky and which
6:25
Speaker 1: are not, but also a much broader range of operational
6:29
Speaker 1: capabilities from collections, hardships, settlements, uh, you know, and uh
6:34
Speaker 1: and recovery options.
6:36
Speaker 3: No, that's I think that's super interesting. And you know,
6:39
Speaker 3: and certainly the business that you're running has a really
6:41
Speaker 3: large scale, uh that it is a large business that
6:44
Speaker 3: we are servicing and across multiple products now versus just
6:47
Speaker 3: the the original personal loan uh fixed rate installment loan
6:52
Speaker 3: product that Upstart is known for. So maybe talk to
6:55
Speaker 3: me a little bit about like why is the servicing
6:58
Speaker 3: operation different, particularly for the size of the organization.
7:03
Speaker 1: Yeah, I definitely do think there are some parts of
7:07
Speaker 1: servicing that are really well plumbed, you know, and well
7:11
Speaker 1: known in the industry, you know, and then some parts
7:15
Speaker 1: that really give us some opportunity. And maybe I think
7:18
Speaker 1: of us generally as we scale as having the opportunity
7:22
Speaker 1: to do more and more things with novelty. That is, like,
7:27
Speaker 1: you know, we're managing more than ten billion dollars in
7:30
Speaker 1: loans today. And while that might seem paltry compared to
7:33
Speaker 1: maybe some of the big five banks, you know, of course,
7:37
Speaker 1: it's a it's a really really large portfolio, you know
7:40
Speaker 1: when you think about it at an absolute level, and
7:44
Speaker 1: that sort of gives us the opportunity to really make
7:47
Speaker 1: deep investments in our technology infrastructure. You know, we have
7:51
Speaker 1: a large engineering team. We have not that large, but
7:56
Speaker 1: very high talent and growing machine learning team, you know,
8:00
Speaker 1: and we have a set of vendor partnerships you know,
8:04
Speaker 1: in the in the low low dozens, and these are
8:09
Speaker 1: these are kinds of infrastructure and partnerships that you know,
8:14
Speaker 1: most lenders just can't justify on their own. If you
8:17
Speaker 1: think about, you know, maybe trying to set up this
8:20
Speaker 1: kind of a technology organization. You know, if you were
8:24
Speaker 1: managing a portfolio of maybe uh, you know, one tenth
8:27
Speaker 1: or one fifth our size, then trying to amortize those
8:30
Speaker 1: costs wouldn't make sense. And you know, we're really quite
8:33
Speaker 1: excited about the opportunity to build out that technology in
8:37
Speaker 1: a way that deliberately plans for scale and then use
8:41
Speaker 1: the sort of operational leverage of our scale to reinvest
8:45
Speaker 1: in more technology build out.
8:47
Speaker 2: Yeah.
8:48
Speaker 3: Now, I think that's a really interesting point, Like you
8:50
Speaker 3: think about the things that upstart is known for, and
8:52
Speaker 3: we do really well our core competencies, and then applying
8:55
Speaker 3: it to this area and really being able to do
8:58
Speaker 3: kind of the basic things that content acts, compliance, collections,
9:01
Speaker 3: all the specialty items you mentioned, but then really having
9:04
Speaker 3: that enterprise level infrastructure run like supporting it and learning
9:09
Speaker 3: from all of those repayment events that your team sees
9:12
Speaker 3: every day.
9:13
Speaker 1: To bring it back to that first point that I
9:15
Speaker 1: was sharing a little bit earlier, I think this is
9:17
Speaker 1: a perfect machine learning problem. That is, it's very high
9:20
Speaker 1: analytical complexity, you know. That is, Oh, what we care
9:24
Speaker 1: about in servicing is not really whether someone pays us
9:28
Speaker 1: just tomorrow, but whether they pay us over the next three,
9:33
Speaker 1: five or seven years depending on the loan term, you know,
9:36
Speaker 1: and deciding whether we should be really pursuing collections hard,
9:41
Speaker 1: whether we should be approaching someone with an offer of
9:44
Speaker 1: a modified payment schedule for short term relief or actually
9:47
Speaker 1: offering a much longer term and more permissive settlement. Well, well,
9:52
Speaker 1: this is this is something that has an objectively correct
9:55
Speaker 1: answer but is extremely extremely hard to predict. So the
10:00
Speaker 1: combination of high analytical complexity, massive personalization, and as we
10:05
Speaker 1: were just talking about this sort of ten years of
10:07
Speaker 1: data from Upstart's history and servicing, I think this is
10:11
Speaker 1: this is something that really gives us a great position
10:14
Speaker 1: to build out the kind of scaled program that it
10:18
Speaker 1: would just not be economical for smaller portfolios to deliver.
10:21
Speaker 2: Yeah.
10:22
Speaker 3: Absolutely, I think the scale is important. And you know,
10:24
Speaker 3: it's a lot of the reason that that partners work
10:26
Speaker 3: with us on the on the underwriting and customer acquisition side, right, Like,
10:31
Speaker 3: we have the very large marketing engine to to identify
10:35
Speaker 3: and acquire the customers that they will want to convert
10:38
Speaker 3: into members or or into customers and cross sell them,
10:42
Speaker 3: and we can apply that same scale in this area
10:45
Speaker 3: and really deliver a delightful experience while maximizing their returns.
10:50
Speaker 3: One thing that you said that that kind of caught
10:52
Speaker 3: my attention is that you've got in a third party's
10:55
Speaker 3: and vendors maybe in like the dozens that you're working with,
10:58
Speaker 3: and so you know that is certainly an operational challenge,
11:01
Speaker 3: and how do you think about all of that vendor
11:05
Speaker 3: management and like quality and oversight and more of kind
11:08
Speaker 3: of the core operational function of working with those sorts
11:11
Speaker 3: of BPOs and vendors.
11:14
Speaker 1: I think this is such a good question because there's
11:16
Speaker 1: really a strong tension in servicing and that is maybe
11:21
Speaker 1: number one that I want to invest our energy in
11:24
Speaker 1: doing things that we can do differently and better than
11:28
Speaker 1: other kinds of organizations. You know, and boy call centers,
11:32
Speaker 1: they've been around for a long time, They've been optimized
11:35
Speaker 1: in really really deep ways. And at the same time,
11:39
Speaker 1: you know, we really want to set a very very
11:41
Speaker 1: high quality bar and we want to make sure we're
11:44
Speaker 1: delivering better than simply industry standard. And the way that
11:48
Speaker 1: we think about that in relationship to vendors is really
11:50
Speaker 1: that it starts with pretty clear accountability. We have this
11:54
Speaker 1: dual ownership model where both our internal teams and our
11:57
Speaker 1: partners really share responsibility for these outcomes. And the way
12:02
Speaker 1: that we sort of materialize that in the world. While
12:05
Speaker 1: we run random audits, of course, we try to spend
12:07
Speaker 1: a lot of time with our vendors.
12:08
Speaker 2: We're doing a lot of site visits.
12:10
Speaker 1: Uh, you know, we have a pretty clear performance scorecards,
12:13
Speaker 1: you know, and then we want to make sure that
12:15
Speaker 1: actually those things are backed up with very clear quality
12:20
Speaker 1: incentives in our contracts with those vendors. You know, we
12:25
Speaker 1: want to make sure that our incentives are aligned and
12:28
Speaker 1: when we do well, they do well, you know. And
12:31
Speaker 1: and generally what that means for us is like, if,
12:33
Speaker 1: for example, an error rate it creeps above our target,
12:36
Speaker 1: then we want to make sure that there are contractual penalties.
12:39
Speaker 1: But if conversely, quality improofs the partner is going to
12:42
Speaker 1: win more volume, they're going to be able to grow,
12:44
Speaker 1: and they're gonna be able to sort of benefit from
12:46
Speaker 1: that relationship with us. So you know, in this regard,
12:50
Speaker 1: you know, I would say that managing vendors well and
12:52
Speaker 1: doing doing contracts, this isn't really the flashy part of servicing.
12:56
Speaker 1: It's really just about building trust and enforcing consistency and
13:00
Speaker 1: and really identifying root causes really clearly when things go wrong.
13:04
Speaker 3: Yeah, and I think so less scenario where you'd maybe
13:07
Speaker 3: focus on novelty and more about just being precise and
13:11
Speaker 3: having discipline and having good communication between you and the
13:15
Speaker 3: vendors as well to make sure that you're all I
13:17
Speaker 3: think agreeing on what quality is and how you're measuring
13:20
Speaker 3: it is certainly critically important.
13:23
Speaker 2: I love that you use that word novelty.
13:26
Speaker 1: There's a quotation from Angela Duckworth, who I think is
13:31
Speaker 1: one of the foremost thinkers on grit that talks about
13:35
Speaker 1: how you maintain focus in an area when it becomes
13:39
Speaker 1: a little less fun And I just remember so clearly
13:43
Speaker 1: her sort of saying, Oh, your ability to substitute nuance
13:48
Speaker 1: for novelty, to think about doing the same thing but
13:51
Speaker 1: with ever increasing nuance is highly predictive of your willingness
13:55
Speaker 1: to stick with something, you know, even once it's no
13:59
Speaker 1: longer flashy, you know. And I really think this applies
14:01
Speaker 1: to servicings so deeply that the best servicing models they're
14:04
Speaker 1: not going to be they're just really relentlessly consistent.
14:07
Speaker 3: Yeah, that's funny. That's one of my favorite quotes, and
14:10
Speaker 3: I'm probably going to paraphrase it a little bit wrong.
14:13
Speaker 3: Is that like a differentiator between people as people who
14:16
Speaker 3: are willing to sprint when the finish line is unclear,
14:20
Speaker 3: but are willing to sprint anyways.
14:22
Speaker 2: I'm going to take that. I love it.
14:24
Speaker 3: So, you know, we've talked a lot about like the
14:26
Speaker 3: machine learning pieces, and then some of the core competencies
14:29
Speaker 3: and the vendors and different things that you've been working
14:31
Speaker 3: on with the team that maybe talk to me a
14:33
Speaker 3: little bit about like the people side. So when you
14:36
Speaker 3: think about like identifying that differentiated talent, like, as you said,
14:39
Speaker 3: you know, if you're collecting on say a fifty thousand
14:42
Speaker 3: dollars loan that's passed you, you may need a more
14:45
Speaker 3: skilled senior collector. You may be spending a little bit
14:48
Speaker 3: less effort on something that's maybe a smaller dollar or
14:51
Speaker 3: maybe less likely that you don't need is skilled of
14:53
Speaker 3: a person in there. So you know, maybe kind of
14:57
Speaker 3: a couple of things differentiating talent, like identify people who
15:00
Speaker 3: are the right fit for the different roles and what
15:02
Speaker 3: those roles are. And then as you think about all
15:05
Speaker 3: of the work with LLLMS and automation, how you kind
15:09
Speaker 3: of bring humans into the loop and where the humans
15:12
Speaker 3: are there, and then where the automation is best the
15:15
Speaker 3: kind of forward forward point.
15:18
Speaker 1: I think these are both really really rich topics. Maybe
15:20
Speaker 1: we can start with sort of thinking about human talent
15:23
Speaker 1: and then talk about how we're incorporating ELMS, because of
15:26
Speaker 1: course that is the technology shift of our age. On
15:31
Speaker 1: the human side, I guess I really tend to think
15:35
Speaker 1: about the first and most important thing as being able
15:39
Speaker 1: to understand what causal role particular human is playing in
15:45
Speaker 1: a servicing decision. And what I mean by that is like,
15:48
Speaker 1: if you think about all of those barbers who are delinquent,
15:52
Speaker 1: and you take the ones who are definitely going to
15:55
Speaker 1: pay you back whether or not you talk.
15:57
Speaker 2: To them, well, then you definitely shouldn't talk to them.
16:00
Speaker 2: You know.
16:01
Speaker 1: There's no there's no reason to spend your time and
16:04
Speaker 1: money trying to reach out to somebody who is sort
16:06
Speaker 1: of gonna pay you no matter what. In that case,
16:09
Speaker 1: the human would have no causal impact. And maybe on
16:12
Speaker 1: the way other end of the extreme, if we take
16:15
Speaker 1: people who are definitely not going to pay you back
16:18
Speaker 1: whether or not you call them, well then don't call them.
16:20
Speaker 1: You're wasting their time and you're wasting your own money,
16:25
Speaker 1: you know, and again you're having really no causal impact.
16:28
Speaker 1: And so the first question for us has really been
16:31
Speaker 1: trying to isolate the causal impact of particular individuals in
16:36
Speaker 1: changing people's mind I often talk about the best collectors
16:39
Speaker 1: as being highly empathetic, highly persuasive, you know, and there's
16:44
Speaker 1: a there's a big analytical question in sort of how
16:47
Speaker 1: we attribute a certain amount of the collected dollar value
16:51
Speaker 1: to the exact person who was on the phone at
16:54
Speaker 1: a given time.
16:55
Speaker 2: This is one really really interesting question.
16:58
Speaker 1: But even once you have solved that pretty rich analytical problem,
17:02
Speaker 1: and we have not solved it, we're sort of making advances,
17:05
Speaker 1: but it's not all the way solved by any means.
17:07
Speaker 1: Even once you've solved that really rich analytical problem, well,
17:12
Speaker 1: then you can sort of think about, oh, well, are
17:16
Speaker 1: the kind of people who demonstrate the greatest causal impact.
17:20
Speaker 1: What kind of background do they have? You know, do
17:22
Speaker 1: we want to actually hire for that background? Do we
17:24
Speaker 1: want to train for it?
17:26
Speaker 2: You know?
17:26
Speaker 1: What are we actually willing to pay? You know, of
17:29
Speaker 1: course you can get generally a different kind of person
17:33
Speaker 1: if you're willing to pay a lot more. And you know,
17:35
Speaker 1: I aspire for us to maybe have some collections roles
17:40
Speaker 1: that are highly competitive with inside sales. You know, I
17:43
Speaker 1: think that generally, if you think about the profile of
17:46
Speaker 1: a collector, we really want that profile to look more
17:48
Speaker 1: like an inside salesperson and less like you know, a
17:53
Speaker 1: call center operator, you know, and then the third thing
17:56
Speaker 1: really how are you incentivizing them?
17:58
Speaker 2: What actions are you encouraging them to take?
18:00
Speaker 1: And and what are you rewarding them for and and
18:02
Speaker 1: so I think the whole structure on the human side
18:05
Speaker 1: is really really nuanced and complex in terms of combining, uh,
18:10
Speaker 1: the background of your collectors, the overall compensation, and then
18:14
Speaker 1: sort of how you mix based comp variable comp with
18:17
Speaker 1: specific collections based incentives. If we shift to the l
18:21
Speaker 1: M side of things, l lms are, are, of course,
18:25
Speaker 1: uh transformative, They're They're an unbelievable technology layer on on
18:30
Speaker 1: our work and I sort of imagine that they will
18:33
Speaker 1: eventually be diffused into almost all of our activities. That
18:38
Speaker 1: that is, like, we want them to be part of
18:40
Speaker 1: our software, We want them to be part of our
18:42
Speaker 1: internal tooling, you know, and of course we want to
18:44
Speaker 1: be part of our operational interactions, uh with our operational interactions.
18:51
Speaker 2: Uh with borrowers, you know.
18:53
Speaker 1: And the way that that's manifested today, well, we generally
18:58
Speaker 1: have already rolled out programs that allow for our own
19:03
Speaker 1: frontline agents to benefit from better M assisted tooling. We
19:08
Speaker 1: have LMS analyzing, analyzing callscripts, maybe helping to identify where
19:13
Speaker 1: there are complaints, helping to categorize where we have opportunities.
19:18
Speaker 1: You know, we have lms assisting in our training, trying
19:22
Speaker 1: to you know, role play with some of our incoming
19:26
Speaker 1: operational agents, and of course, as with others, you know,
19:30
Speaker 1: we really have started to put lms in a place
19:34
Speaker 1: where they're taking some of these ropes or routine high
19:38
Speaker 1: volume tasks like summarizing notes, triaging inbounded messages, or guiding
19:43
Speaker 1: income in some cases next best actions. And our frontline
19:46
Speaker 1: team has generally responded really really positively to this. That is, like,
19:50
Speaker 1: if you are a skilled and influential collector, you don't
19:54
Speaker 1: really want to be taking notes all day. That's not
19:56
Speaker 1: the part of your job that you enjoy, you know.
19:59
Speaker 1: And so some of the feedback that we've gotten from
20:02
Speaker 1: our team is that the system generally feels more organized,
20:04
Speaker 1: feels more understandable, feel more clear, you know. And our
20:07
Speaker 1: goal here is for AI to play a pretty central
20:11
Speaker 1: role in helping to surface relevant context instantly, which we
20:15
Speaker 1: think will let our best people focus on empathy, on negotiation,
20:19
Speaker 1: as I said, on persuasion, you know, and exercise that
20:22
Speaker 1: that human judgment, which we think is is still pretty
20:25
Speaker 1: pretty unique.
20:27
Speaker 3: How do you think about maybe then finding that talent
20:30
Speaker 3: so as you think, you know that is the role
20:32
Speaker 3: is changing. And I talked a lot about and I
20:35
Speaker 3: think it's a pretty uh maybe a unique idea that
20:38
Speaker 3: you want to think about. Uh, people who you know
20:42
Speaker 3: maybe are are currently paid more like contact center specialists,
20:46
Speaker 3: and in the future it could be paid and compensated
20:48
Speaker 3: more like an inside sale like a BDR inside sales representative.
20:53
Speaker 3: How do you think about identifying or and or upskilling
20:56
Speaker 3: people that are in these roles today, so as you
20:58
Speaker 3: think about helping them get comfortable with with AI l
21:01
Speaker 3: ll ms and and have the right skills and then
21:05
Speaker 3: and and that I think that mindset of that this
21:08
Speaker 3: is an enhancement, this will make my job better, not
21:11
Speaker 3: like this will just replace my job. Uh. And you know,
21:13
Speaker 3: focus on the opportunity and out the fear, Like how
21:15
Speaker 3: do you think about like identifying that talent in the
21:18
Speaker 3: existing pool, but then also you know, as you're as
21:21
Speaker 3: you're hiring and looking for that outside with with new roles.
21:26
Speaker 2: Yeah.
21:26
Speaker 1: Uh, it's it's such a good question because of course, uh,
21:30
Speaker 1: you know, if you're going to go out and and
21:32
Speaker 1: you sort of say, I bet there are a bunch
21:34
Speaker 1: of people who would be really good at this job,
21:36
Speaker 1: who would never think of themselves as good at this job.
21:40
Speaker 2: Uh.
21:40
Speaker 1: Well, that's it's a really really hard search problem. And
21:44
Speaker 1: I guess the way that we've thought about it is
21:46
Speaker 1: by really creating a virtuous cycle, a little bit of
21:49
Speaker 1: a flywheel. That is, Oh, first we put in place
21:52
Speaker 1: this measurement, trying to see how, uh you know, how
21:56
Speaker 1: each person is contributing to the overall impact of the team,
22:01
Speaker 1: and then maybe looking back at those people who are
22:04
Speaker 1: most impactful and trying to isolate which skills or behaviors
22:09
Speaker 1: they demonstrate that maybe other folks don't demonstrate, and then
22:13
Speaker 1: of course taking that both into our training and into
22:17
Speaker 1: and into our hiring, you know, so holding a higher
22:20
Speaker 1: bar for those skills and behaviors, you know, generally rewarding
22:27
Speaker 1: those skills and behaviors. And of course, as I mentioned,
22:29
Speaker 1: all of this is really tied into compensations. So if
22:32
Speaker 1: you're looking for either higher quality expression of the same
22:37
Speaker 1: skills and abilities or novel skills and abilities that maybe
22:41
Speaker 1: someone has and another person doesn't have, well you have
22:44
Speaker 1: to be willing to pay for that. And generally, my
22:47
Speaker 1: hope is that our collections team over time is better
22:50
Speaker 1: and better compensated because they are so good at helping
22:54
Speaker 1: barbers find paths back to current they're playing a really
22:57
Speaker 1: pivotal role in shaping people's minds, their behaviors, and ultimately
23:02
Speaker 1: the repayment on their loans.
23:04
Speaker 3: Yeah, and I think it's a unique way that we've
23:06
Speaker 3: talked about this in UH in in servicing here at Upstart,
23:12
Speaker 3: where you know, a lot of times folks can think
23:14
Speaker 3: about operations teams as more of a cost center and
23:17
Speaker 3: it's just you know, preventative or defensive in servicing UH.
23:22
Speaker 3: And I think the way that you've approached it and
23:25
Speaker 3: are thinking about it is more dynamic than that that
23:27
Speaker 3: this is a place where where there's returns to be
23:30
Speaker 3: to be had because there are people, as you said,
23:32
Speaker 3: who there's some people that will that will pay if
23:36
Speaker 3: we contact them, or will pay if we don't contact them,
23:39
Speaker 3: and then there are people who do the opposite, like
23:40
Speaker 3: we can contact them and they won't. But there's you know,
23:43
Speaker 3: I think you have a pretty strong belief that there's
23:45
Speaker 3: a subset of folks that maybe won't pay if we
23:48
Speaker 3: don't contact them in the right way or the right frequency.
23:51
Speaker 3: But our behaviors can differentiate that and can drive a
23:54
Speaker 3: better return. You know, maybe talk to me a little
23:57
Speaker 3: bit more. You know, we've already talked quite a bit
23:58
Speaker 3: about machine learning and service and kind of being a
24:01
Speaker 3: hard problem, and I don't think it's an area that
24:05
Speaker 3: really anyone else is spending a lot of time focusing
24:08
Speaker 3: on and using machine learning to solve some of these
24:10
Speaker 3: servicing challenges. Maybe just expound a little bit on like
24:13
Speaker 3: why is it so hard? And why is it so
24:16
Speaker 3: hard to take something that that we know how to
24:19
Speaker 3: do really well machine learning and apply it to this
24:21
Speaker 3: particular kind of set of problems.
24:24
Speaker 2: Yeah, it definitely is hard.
24:27
Speaker 1: And you know, I think there are increasingly folks who
24:31
Speaker 1: have awareness that that servicing is an exciting opportunity space.
24:36
Speaker 1: But you know, much as I hate to say it,
24:39
Speaker 1: when I was a little boy, I definitely did not
24:42
Speaker 1: grow up saying, oh, what I want to do in
24:44
Speaker 1: my life is run a large servicing organization.
24:46
Speaker 2: That that just isn't It's not it's not super.
24:48
Speaker 1: Common, you know, And and I think if you you know,
24:52
Speaker 1: if you think about the space and the relevance for
24:57
Speaker 1: machine learning, well, the servicing data, it's it's fundamentally, there's
25:02
Speaker 1: fundamentally a lot of it that's really good, you know,
25:04
Speaker 1: that's that's one sort of core premise for being a
25:07
Speaker 1: good machine learning problem. But the servicing data is it's
25:10
Speaker 1: really messy and it's really behavioral. That is, you're dealing
25:14
Speaker 1: with really complex interactions. Somebody can pay this month, not
25:18
Speaker 1: pay next month. You're dealing with multi channel interactions. Hardly
25:22
Speaker 1: anybody wants to be the lender that is doing a
25:25
Speaker 1: holdout where you're like, well, I wonder how impactful calling is,
25:30
Speaker 1: and so I'm not going to call fifty percent of
25:32
Speaker 1: my barbers and I'm gonna see which one's default. That's
25:34
Speaker 1: that's really not what anybody wants to be doing. And
25:38
Speaker 1: so trying to sort of inch your way toward understanding
25:42
Speaker 1: causality across payments, calls, hardship plans, you know, third party collections,
25:48
Speaker 1: first party collections. You know, the sort of evolution of
25:51
Speaker 1: that data space over time is really really quite complex,
25:56
Speaker 1: and in many ways, you know, it feels to me
25:59
Speaker 1: like we are inching toward an understanding, and every time
26:03
Speaker 1: we take a little bit of a step forward, well,
26:06
Speaker 1: it broadens our horizon to the future. We have a
26:08
Speaker 1: little bit of a sense, but there's still so much
26:11
Speaker 1: that we don't know that we can't quite see too yet.
26:14
Speaker 2: Now.
26:14
Speaker 1: I think like if we look at machine learning either,
26:17
Speaker 1: and that's sort of that's a servicing view. If we
26:19
Speaker 1: look at machine learning, traditional mL models, well they're mostly
26:23
Speaker 1: built to predict relatively static outcomes.
26:26
Speaker 2: You know.
26:26
Speaker 1: They want to predict something at a point in time.
26:29
Speaker 1: And the challenge for us is like, well, we're trying
26:32
Speaker 1: to build models that then adapt in relatively real time
26:35
Speaker 1: and predict a wide range of hypotheticals. That is, like
26:39
Speaker 1: which intervention today has the greatest chance of improving recovery
26:44
Speaker 1: or improving customer satisfaction tomorrow. And this is an enormous
26:48
Speaker 1: modeling challenge, but it's also, I think, really a natural
26:52
Speaker 1: fit for what Upstart is good at, and a natural
26:55
Speaker 1: extension of the skill set that we've been building.
26:58
Speaker 2: We spend a decade really.
26:59
Speaker 1: Trying to get these models to be good at understanding
27:03
Speaker 1: complex behavioral data over long term outcomes, you know, and
27:07
Speaker 1: we think that incorporating some of these operational feedback loops
27:11
Speaker 1: is a is a really really exciting next challenge.
27:14
Speaker 3: Yeah, and I think it definitely goes into you know,
27:17
Speaker 3: kind of what we started on as we were talking
27:19
Speaker 3: about underwriting and origination and some of the things that
27:24
Speaker 3: impact folks on that end, like the macroeconomic cycle, things
27:27
Speaker 3: like the you know, a recent government shutdown of course,
27:30
Speaker 3: and certain people being furloughed and not receiving their money.
27:33
Speaker 3: You may have borrowers who had a desire to pay
27:37
Speaker 3: but not an ability to pay, which is very different
27:39
Speaker 3: than somebody who doesn't have a willingness to pay, or
27:43
Speaker 3: it doesn't have either a willingness or or an ability
27:46
Speaker 3: to pay.
27:46
Speaker 1: At the end of the day, servicing is oh, I
27:49
Speaker 1: should say, loan performance is really a function of your underwriting,
27:53
Speaker 1: your macro and your servicing. And of course, you know,
27:57
Speaker 1: we want to play a central role in upstarts overall
28:01
Speaker 1: long performance. But you're absolutely right to highlight that there
28:05
Speaker 1: are such complex and confounding factors that understanding which part
28:10
Speaker 1: of that impact came from us is a really ongoing challenge.
28:15
Speaker 2: All right.
28:15
Speaker 3: So we've talked a lot about what you've been doing,
28:17
Speaker 3: you know, and I mean, I think it's fair to
28:20
Speaker 3: say that most people don't as you said, is it
28:22
Speaker 3: you know, six year old and aren't thinking I'm going
28:24
Speaker 3: to run a large servicing organization when I grow up,
28:26
Speaker 3: and I would imagine that six years ago you probably
28:29
Speaker 3: were not thinking that either before you you became the
28:33
Speaker 3: general manager of servicing here at Upstart. But certainly looking ahead,
28:37
Speaker 3: I know one of the big topics on the horizon
28:40
Speaker 3: for us as servicing as a service. So maybe talk
28:43
Speaker 3: about that, like what does that mean, and what does
28:46
Speaker 3: that look like, and what would I would kind of
28:48
Speaker 3: excite you about it, and what do you feel like
28:52
Speaker 3: you really need to focus on to be very really
28:55
Speaker 3: effective there.
28:56
Speaker 1: Yeah, maybe one of the first things is just to
28:59
Speaker 1: say I really think of good servicing as good stewardship
29:04
Speaker 1: of our barbers. That is, like, oh, I believe that
29:08
Speaker 1: by offering customers better experiences, you make them more loyal
29:13
Speaker 1: to you. I believe that by offering lower friction experiences,
29:16
Speaker 1: you make them more likely to pay you back, you know.
29:20
Speaker 1: And I believe that by offering empathetic hardship programs that
29:24
Speaker 1: match customers real life financial circumstances, you can really help
29:28
Speaker 1: them get back on their feet and keep a loan
29:32
Speaker 1: current that otherwise might have rolled to charge off. And
29:35
Speaker 1: so we've talked a lot about the analytical challenges of servicing,
29:39
Speaker 1: But at the end of the day, the customer experience
29:42
Speaker 1: is really front and center. The way that you create
29:47
Speaker 1: better loan performance outcomes is, in my opinion, by offering
29:52
Speaker 1: really next level and excellent customer experiences.
29:55
Speaker 2: And when I think about the magnitude of.
29:58
Speaker 1: This challenge, you know, almost within I don't know, within
30:02
Speaker 1: probably two months of joining the servicing team, I was
30:06
Speaker 1: pretty clear this is a year's long, you know, maybe
30:11
Speaker 1: decades long kind of problem for us to solve. This
30:16
Speaker 1: is not a fast problem. And my goodness, if we
30:19
Speaker 1: build the infrastructure to solve this problem, we shouldn't keep
30:22
Speaker 1: it to ourselves, you know. And that's really where servicing
30:24
Speaker 1: as a service comes in for me.
30:26
Speaker 2: You know.
30:26
Speaker 1: I believe that the infrastructure, the kinds of analytical approaches
30:31
Speaker 1: that we're building, you know, and the kinds of experiences
30:34
Speaker 1: that we're trying to to marry them with are going
30:37
Speaker 1: to be unique and gonna be really best in class.
30:40
Speaker 1: And I want to make sure that we're able to
30:42
Speaker 1: apply those excellent experiences to as many loans as possible,
30:47
Speaker 1: whether or not they were underwritten by Upstart or not.
30:51
Speaker 3: Yeah, I think that's a really exciting, you know, exciting
30:55
Speaker 3: idea as we think about you know, kind of optimizing
30:58
Speaker 3: what's working here and then as you said, share it
31:00
Speaker 3: and like we you know, we there's a lot of
31:02
Speaker 3: things I think we are have very strong competencies in
31:05
Speaker 3: and and this being one of them, and so kind
31:07
Speaker 3: of building that for for what's next, and and potentially
31:10
Speaker 3: servicing you know, loans that didn't even originate on our
31:12
Speaker 3: platform as an exciting place for us to head uh,
31:16
Speaker 3: and that we can really bring all of that kind
31:18
Speaker 3: of machine learning uh and servicing know how and applying
31:22
Speaker 3: it to to in a kind of a long term way,
31:25
Speaker 3: which is I think interesting that this is a long
31:27
Speaker 3: term innovation build. So I think what what are you
31:32
Speaker 3: most excited about? Like what's next on your agenda? That's
31:35
Speaker 3: probably to you the most exciting thing, uh, that that
31:39
Speaker 3: you'll get the chance to build.
31:41
Speaker 1: Yeah, our north star here is really to be able
31:43
Speaker 1: to service not just UPSTARTE originated loans, but partner loans
31:47
Speaker 1: as well. And we have a handful of things that
31:50
Speaker 1: that we still need to work on our core platform
31:54
Speaker 1: capabilities today. You know, of course our bread and butter
31:58
Speaker 1: is unsecured personal loans. We have lots of work to
32:01
Speaker 1: do to maybe UH build out some of those UH
32:06
Speaker 1: revolving credit capabilities UH for loan types like heylocks or
32:12
Speaker 1: credit cards, you know, both of which make up a
32:14
Speaker 1: very large part of the third party servicing UH market today.
32:19
Speaker 2: You know.
32:20
Speaker 1: But but to be honest, UH, you know, platform capabilities
32:23
Speaker 1: are something that we have to do, you know. And UH,
32:27
Speaker 1: And then that's the part that really gets me excited
32:29
Speaker 1: that I'm that I'm super thrilled about is really trying
32:32
Speaker 1: to drive more of this unique value through more personalization.
32:37
Speaker 1: And if there were any one place that I would
32:39
Speaker 1: probably want to put that, you know to me, we
32:42
Speaker 1: already talked about it a little bit. But the opportunity
32:45
Speaker 1: to create more flexible, more personalized hardship plans UH with
32:49
Speaker 1: deep awareness of whether we are increasing the loan value
32:55
Speaker 1: or not UH is UH is something that I just
32:58
Speaker 1: get continuously excited about.
32:59
Speaker 2: It.
33:00
Speaker 1: It's an immense challenge because the search space is so wide.
33:04
Speaker 1: You can change the length of the loan, you can
33:07
Speaker 1: change the interest rate of the loan, you can change
33:08
Speaker 1: the principle of the loan. You can do it when
33:10
Speaker 1: someone is in early or mid or late delinquency. You
33:13
Speaker 1: can offer it via email or text or call. There
33:16
Speaker 1: are so so many options in when and how to
33:20
Speaker 1: do this best, you know, but I think of this
33:23
Speaker 1: as really the most exciting thing on our horizon.
33:26
Speaker 3: Yeah, I love I love that framing, like really thinking
33:29
Speaker 3: about the deep personalization of it, and I think, you know, certainly,
33:34
Speaker 3: I mean Jesse, thanks for joining us, giving so look
33:36
Speaker 3: and how Upstart is really turning Servicing into a true
33:40
Speaker 3: performance engine for the company. And I too, am excited
33:44
Speaker 3: about the idea of us being able to make this
33:47
Speaker 3: available to something that others can make use of as
33:51
Speaker 3: we continue to grow.
33:52
Speaker 2: Thanks Lene, Servicing.
33:53
Speaker 1: You know, it's not the thing that we dreamed of
33:55
Speaker 1: when we were little kids, but it's a pretty darn
33:58
Speaker 1: exciting place to be today.
34:00
Speaker 3: Thanks again to our listeners for tuning in to another
34:03
Speaker 3: episode of Leaders and Lending. Don't forget to subscribe and
34:06
Speaker 3: share the episode and we will see you next time.