Creating value with data and standardised KPIs

The full interview with Jörgen Köster from Dignisia where we discuss some of the complexities of data and performance comparison in collections, however also some of the opportunities too, especially if we get this right.

Find more, interviews and video clips-> Here.

Interview Transcript

0:02
So hi everyone, I’m here with Jorgen Koster today who’s the co founder of Dignisia in the credit intelligence, portfolio, portfolio intelligence and sort of debt collections performance kind of space. So you’re gonna thanks very much for for joining me today, it’s great to have you here.

0:22
It’s a pleasure, thank you. Um,

0:23
so I suppose, you know, with your expertise in the in the debt, purchase space and intellect portfolio analytics, so because to hear a little bit about some of the things you’ve been seeing recently across across your clients, or just just generally within the market, I’m given the fact you spent so much time looking at in some detail.

0:42
Now, we can see an increase in, in Portfolio space, if we look at the one off spot, seems to be an interesting selling off back books. There is absolutely an increase in that. We also see some worries about for referral contracts regarding price, we see some price negotiations, etc happening. And, of course, given the state of the economy, etc. There are worries about future prices. I

1:10
mean, how did that change, I suppose post pandemic when the pandemic must have did people tend to sort of stop? And now that sort of, you know, there’s there’s been so like reactivity, but then I suppose there’s concerns now on, like, energy, energy costs, isn’t there?

1:26
Yeah, yeah. Everyone was really concerned about when the credit, strike. And the dip, there wasn’t as severe, we could actually consumers maintain the same payment parents, we could see that collection agencies could have problem keeping operations up. But if you look at the sort of behaviour of the doctors, there was not much difference. And I think everyone fears that now it can be a bit different. There are another kind of fundamentals now, because COVID, Procmon, when the economy was really good, but now it’s, it’s a bit more serious with inflation.

2:12
I know, you look across multiple markets in in Europe, I suppose multiple markets internationally. Do you see those sort of when you’re doing this was for the portfolio analysis to understand the performance? Do you find that that changes a lot by market? So talking a bit about the the energy prices? I mean, that’s affecting Europe quite a bit. But just slightly different in different markets, Italy is a little bit different from say, say, Germany, or from the UK or from France, as example, just with government support? I mean, are you do you sort of see see variations in portfolio performance by market significant, significantly, or is it? And what’s I suppose what kind of drives that?

2:49
I think a major driver is of course, legislation. Like what happens if cases or statute Mark, if you have claiming them? Or if you can continue to do that, for example, and also different ways to use the legal system? So the collection profile is very different for markets?

3:11
Yeah, so I suppose, and you find some markets that are that are easier or more accepting for things like that sample debt purchase than than others? I mean, I mean, what would you say are the ones that are probably the more open versus the ones maybe that maybe it maybe a little bit less, so a little bit sort of more bureaucratic?

3:28
Yeah, and, for example, the Nordic countries where we are very active. It’s easy to get public data about that process. But otherwise, I’ll say it’s pretty generic that you need good portfolio information really need good Oracle’s for the portfolio, historical cash flow, total historical cash flow, or some cases, etc, that really gives the foundation for good forecasting.

3:59
And the MRU just like talking about the EMI and obviously, you spend, I mean, sort of seeing the system, and there’s huge amounts of data on EMI. I mean, do you find that that? Do you have to change that by the different markets? Or is that is that pretty much the same? And so So, for example, valuing a portfolio or looking at activity outside of compliance piece you’ve mentioned? I mean, is it is it? Is the techniques same sort of quite transferable or do you have to do a lot of sort of different between the different different markets?

4:25
It’s quite the same, I would say. It’s the palette of different actions that you can take the difference between different countries like there are things in Germany that you can do that are not available or legal in Sweden and Sweden, that you cannot do in Denmark and the other way around. So, the sort of different kinds of actions that you need to monitor differs by by country a lot, but the general principle is to say, usually, you try to reach out to the After to get the settlements, you have legal options, you can obtain verdicts, you can go to the bailiff if that verdict in some way. And then there are other ways as well, like in some context, you can go to the debtor in some countries that’s not available, you can report after two credit reporting agencies. And that’s not possible in other countries. So

5:22
one of the challenges that there always is, and I’ve seen this, too, is like, is how do you compare some of that data? And particularly, I mean, I mentioned that’s China for you, and like, how do you get the information in to then be able to compare it to make it comparable? So you can do that? I mean, is that do you find that that’s quite a challenge in terms of, I mean, it’s not driven by bank setup as much as or financial services setup, I should say, as much as anything cause, you know, and is it as is it becoming more standard to

5:50
I think that was one of our main objectives when we started the museum, because it’s very hard when you get data in different formats and different definitions. So we really started out with our standardised format, and try to introduce that to the collection industry to get reporting in, in a sort of comprehensive format that is, can be used across different office different agencies. So if that works, then you can do an apples to apples comparison between, for example, two different agencies in the same country or compare across country borders?

6:29
And what are the main what are the main challenges you’ve come across? Is it is it definitional? Or is it missing data? I suppose with the other ones, where they just don’t record it? I mean, what the what were you seeing the sort of the main kind of gaps and having this comparable kind of performance?

See also  Conversational Analytics: Finding value from listening

6:45
I would say it in absolutely in missing data, because it’s kind of reporting, put a lot of stress on on the sort of information systems on the DC side. So of course, if you’re not, if you do not record when you try to call the doctor, if he doesn’t answer, then it will be really hard to report that back to the client. So if you kind of relieves a lot of weaknesses if you don’t record data, in this way, and it can also trigger improvement processes to do that, because, of course, if it’s hard to report this back to the client, it’s probably also something that you are blind to.

7:25
I mean, should you collect you can use your template almost to say like, these are the things you should be reporting on, you know, to get no value add information, and almost like do that sort of gap analysis to circumstance.

7:39
Yeah, we have a comprehensive equal format, with all the kinds of different actions that possibly it could be a sort of tracked, and it’s quite detailed. And then depending on the technical capabilities, the DC or the collection unit, if it’s internal map against download. And ideally, you can then get a really good insight into all the actions that have taken all the reactions on the deck to

8:12
make us feel like there’s been a lot of work done on for example, credit bureaus and making credit bureaus compare comparable even across different markets. It does feel like there’s, like this the next frontier, which is like how do you make portfolios, but then even sort of like operational actions and sort of compliance and make it visible across providers, but then also across markets as well be kind of interesting, and what are the nuances and having a standard set of measures? And do you think we’ll get

8:37
there? I think so because there’s really great value in this kind of data. It’s also usable for credit scoring, because then you can include behavioural parameters, like if the depth is different, if the debtor, for example, doesn’t pay, and he doesn’t respond to anything, compared to the who may call in the meeting to say that he cannot pay. Pay Later, he makes a storm has passed, and he really tries to solve it. And if you have the right data, then you can, you can use this awesome credit, and understand the client behaviour.

9:15
So I’m gonna put you on the spot now, which is so if they were, if there were two or three measures that you think are the most important indicators, performance indicators, you’d look at? What do you think, what do you think they were? What should people look at?

9:29
It’s hard. Yeah, it’s hard to get past. Of course, you have to sort of the Automate results from a collection action. But I will also look at different sustainability cues to monitor the basic experience because probably, especially in terms of this, you wanted to have these collectors as clients again, so it’s important and how do they experience the collection? The process, and that can be from being too passive, that they really aren’t aware that they are, that are done on it. Or it can be too aggressive. If you, for example, call too intensely in the evening. That can work. Also the different ways to use the legal system, the colleague and on very small amounts, for example, without communicate before that, and that can really upset people. And then maybe you have unnecessary shorn of that client. However,

10:34
I know that the Mr. Workshop goes like this, doesn’t it, you start at the top, and it’s very quickly you get into like, analysing lots of different things.

10:42
Yeah, exactly.

10:44
So just looking at it from a customer point of view, or from a client point of view, your client point of view, I mean, what are the what are the big risks or measures that they’re worried about? Or they’re wanting tomorrow, monitor these days? In I mean, obviously, you get to help them do that, but what are they thinking about particular I suppose from either portfolio performance or from, you know, I suppose Debt Debt portfolio performance.

11:09
I will say one thing is collection performance and other things, how segments behave on debt collection, there are changes in that there is a currently, I think there’s a high desire to, to be able to follow up with some risky apps on collection. And even to follow it, if you have sold the claims in fourth floor, you’re really that willing to accept the data loss and more looking for ways to get aggregated data to be able to use that as well and modelling, which is also constantly helping our clients with. And also, of course, pricing, if you are important for contracts, that is a major issue now. But that also then again, goes back to collection performance that you are worried

11:59
about interesting. Some other conversations have been having has been around so almost like arrears levels and whether arrears levels have really been going up yet or not. And we’ve got these, these two, two events have happened about the pandemic, but a lot of people seem to have saved quite a bit because they weren’t going out and we weren’t spending money. Now we’ve got increased energy costs, we’ve got increased cost of living going on. And, you know, people have been spending, but they also they’ve got these increased, and it’s like, what are the early indicators that when people will go into arrears, when people will start to fall further into collections or fall off payment plans, versus not an AMI? I mean, are people looking for those, those almost like, small indicators to work out? Like are things going south or can’t be able to preempt whether things are gonna go south, you know, effects and things like pricing?

12:48
Yeah, I think most are monitoring that pretty closely. Now. I would say, usually, we don’t notice that much. Right now. There is a lag effect here, of course, presumably, more can be seen later in the day. But we also presume that, of course, new tech was when that so they, they are not the typical doctors, but they have problems now with the current for some electricity bills, that puts them in a hot pot. And

13:29
all of this takes all of this takes time to flow through to particular back in depth portfolios. The other thing that’s sort of coming up is how can you preempt you know, and really sort of understand the fact that whilst you might be able to afford a payment plan today, you might not be able to afford it in six months time, because we know pretty much that, you know, I mean, energy is probably going to go up further or interest rates are going to go up and we sort of see some of the the inflate inflation type figures that are starting to happen in different different markets, aren’t we? The price of things like exchange rates versus the dollar and those sorts of things.

See also  Digital interaction - does it lead to higher engagement rates?

14:07
Yeah, and you will have different government actions that can also affect somebody dissipation of help with electricity bills. Possibly how

14:21
so if you’re gonna go back to I suppose your client in terms of recommendations of data to complete or data to help monitor the portfolio? What What should What’s up your recommendation to go back when as you’re going through this definitional phase in terms of getting the data and uploading it? What what do you think some of the key indicators that that that you’ll go back and pretty much ask them to try and put in place?

14:48
We usually try to be as broad as possible. Our view is that secured. You should collect all the data points that are possible, and to keep it really great Angular on sort of individual case level, because it’s hard to say now, what you will fall up on later on any, you might be looking more into depth sanitization, or you are, you need to look further into how disputed claims are handling the hazard in the legal system, for example. So I think the best approach is to really have a large appetite for granular high data, even if you don’t understand the immediate use now. Because then you have all the possibilities, to understand different aspects when they occur, because it’s hard to predict what will be important. And there can be legal changes that can that you wanted to explore how will this affect the collection performance, and if you have that data, and then you can model around that, but it’s always hard to come in afterwards and request the data backwards, it’s always much easier to get it with high quality view record, like this. And I suppose

16:07
with the cost of storage these days, it’s a lot easier than it used to be. Maybe we don’t have to be quite as efficient as we used to be.

16:17
No, I think this is really, really valuable data to have.

16:23
So the recommended recommendation is gather as much as you can now. Because once it’s gone, it’s gone. Right? So we are so we will be able to go back and use it.

16:34
Yeah, we’ll see all the time that it’s hard to, to get back and get sort of correct historical snapshots. Because things happen, your collection agency can change systems or other other factors that can happen. So it’s not sure that you can always go back and get snapshots of the data, it’s much more secure, if you sort of collected in real time,

16:58
and how do you find the access to data and the richness of data affects things like valuations and pricing. So if so, if I’m if I’m a if I’m a financial institution, and I’ve got incredibly rich data, I mean, what’s what what’s, what’s the, what’s the impact on pricing, versus one that maybe just you know, just has much more basic data,

17:17
it gives them much higher confidence in the portfolio. It’s much easier to forecast if you have really good data on it, and it looks trustworthy, because there is always a forecast risk, it’s it’s easy to do mistakes with data. And so, of course, the more complete the data set is the more sort of analytics friendly it is, the easier it will be to make good for naught, that will probably then result in a higher price rather, because it will be more attractive. And of course, if you have action data, then you can also show how the case is have been handled. And I can also then remove some some question marks, like Have they all received very, very good offers that they have rejected, for example, and, of course, it it increases the attractiveness of the portfolio. And also shows that the database and Gordon water if you have this neatly collected, it’s higher chance that the portfolio data, of course may be good, most of

18:33
what you just said was actually quite in was it was also interesting was actually quite interesting, or only, it’s not just the data, but also then making sure you have the data in a good structure as well. So if you have lots of data, and it’s let’s say, less structured, it’s harder to analyse and understand it. But then if you’re having a structured format, and you’re like, you’re like a standard structured format, then it becomes easy for people to interpret. And then that that’s, that’s, that’s an added bonus in terms of accurate pricing and valuations as well, which they don’t have to de risk for.

18:59
Yeah, it all boils down to sort of structured data, I think that’s the basis for the analysis for the forecasting and for benchmarking against each other. Yeah. So to have structured and standardised is really, really key.

19:21
And we’re quite keen on sort of, in the industry been quite keen on getting into almost like sort of behavioural type data, particularly things like, you know, really granular data around things like, you know, I suppose, yeah, interaction data, so things like things like website interaction, digital interaction data, and that, that sort of like it’s the amount of data you get becomes almost like exponential when you get into some of these things. I mean, do you think do you think that’s going to start to feed through to the debt purchase process or the valuation process as well? Or is it is it is that a little way off?

19:55
I think that the first immediate use is of course, as a client, you If you’re your bank, for example, it’s really interesting to see this interaction data, how the debtor works on it. If you have a, if you have a pretty fixed pay collection process, for example, then it’s really interesting to see if the debtor reacts immediately when he receives an SMS, for example, from the collection agency. And it’s also interesting to see how much do the deputies use the web? And how is communication going because you can take learnings back to your own collection stage, you can learn that maybe there are segments or age groups that you can approach in a different way. And then you would have mentioned before it goes on to the collection stage

See also  Toothpic

20:45
is that feedback loop as well as also important particular from a client point of view? is like how do you how do you capture the data? Is it structured properly? You do the analysis? And then can you feed it back to them make changes going back and sort of like repeating things like the loop?

21:01
Yeah, exactly. And that goes all the way back to credit modelling and pre collection, really understanding the different segments. And put it all value to when you sell that, of course, because contact data can also be in quality indicator. This is a portfolio with adapters that you can actually reach in again, they are not all responsive. Of course, it’s a more attractive folio.

21:27
I think on a broader picture, it’s almost like the sort of sort of stuck see this, almost like integration between the different businesses, as we sort of become reliant on each other, the data becomes more standard. And we become much more integrated with each other, which is it’s been a long journey. But it’s it feels like this is almost like the next step of it continuing to certain extent, which is, you know, it’s more integration between different businesses rather than sort of like standalone businesses.

21:54
Yeah, I think so too. And I think there has been a consultation process in debt collection industry as well, I think with the right kind of data exchange, will also be easier for new players to come to come into again, and then then it can be more easy to benchmark a new local agency or a specialised agency, and really understand if they are doing something better than that. Cannot do either.

22:29
So I think that also opens up the market a bit, because it’s, and I suppose, you know, pricing us talk a bit about pricing, and pricing has been sort of quite, quite strong in recent years, I mean, sort of capital has been sort of like looking for bigger returns, the debt Mark has been one of the places that he’s looking, you know, and I think over the US hearing over the COVID pandemic, there was quite low supply, and a lot of a lot of people were doing making debt sales. I mean, what what’s your what’s your what’s your forecast? Or how you’re going to see sort of like, pricing? Or how do you think pricing is going to change sort of going forward? And people sort of, especially with more data over under paying for for debts? Do you think at the moment,

23:12
we think pricing will probably drop? There are several factors that sort of point in that direction. Most of all, I will say it’s the underlying cost of capital, that will require a higher return to make investments now. So that’s going to bring prices down a bit, of course. collectability so that’s absolutely. And but not only that, but of course, also a higher insecurity in how we production curves look, in the future, how will this affect quantities of the MPLS for some projects, there are more MPLS out there on the market that will also push prices down. Because there are more Potential portfolios that you can pick and choose from. We think this trend can be higher, if you compare sort of poor quality portfolios with bad data compared to two data rich portfolios, which are more attractive to everyone. If there will be more such towards the more secure, that we can forecast easier. That’s something we think.

24:32
Absolutely. And so So how do you sort of see suppose the next sort of five years sort of panning out in terms of you know, the journey that you guys have been honest, because continuing but I think what I suppose with the background of the economy, what what do you sort of see is the next development in the next five years.

24:49
I strongly think that a concept like like we are building with the museum, to have this it’s have data gathering and, and, and build on this feedback loop back to people. And back to credit scoring. To understand the pricing, I think that’s this structural change that I think we continue. I think that adds so much value. And so changes the kink is the sort of power balance, because suddenly sellers are more can be more in control of understanding how our portfolios are priced to kind of stand, what effects the pricing. And it’s also easier to benchmark collection agencies against each other. And hope that that’s a trend that can be increasing, especially in bad times when

25:49
there is more focus on these issues, common issues as well, often between them between different players, it’s quite interesting to like, what can we learn from each other and sort of help help them market to circumstance? I mean, it’s a fascinating topic. And I do think, you know, I mean, I know there’s a huge amount of investment just in terms of like, producing something, producing some of this and getting the data, right. But I mean, the big takeaway for me is really around for a second, you’ve got to try and get your data, right, and record as much data and then get it structured. So you can then do some of this performance analysis, and there’s definite benefits for it. So

26:24
yeah, that’s an unnecessary ground. So it’s going to

26:27
take some time. I know it’s a it’s, it’s always more complicated than you think on the on a webinar webinar. I quick chat on video so so it’s, it takes time.

26:42
Oh, yeah, no, it’s it can sound easy to just gather the data, but there’s so much that can happen with it. Okay, cases are from different systems, they can be combined. And there are lots of issues that can happen. So you, we invest a lot of time in sort of preparing the data and washing it and making sure that it is relevant completeness,

27:07
I suppose as it is usually a big issue, I’d say.

27:10
Yeah, exactly.

27:12
Well, you’re gonna thanks very much for making the time today. I mean, it’s a it’s a it’s a fascinating topic, and it’s good to have a conversation about EMI and performance and those kind of things because because I love it. Firstly, it’s very It’s so I really appreciate making the time it’s, it’s great to chat with you. Thanks.

27:27
Yes a pleasure Thank you


RO-AR newsletter

Receive notifications of new content notifications: Subscribe here - unsubscribe anytime