Providing instant gratification, a panel discussion from TransUnion Philippine’s Big Data Summit

Last week I had the pleasure of hosting a panel discussion at a virtual event hosted by some of my friends and ex-colleagues at TransUnion in the Philippines. The theme of the day’s talk was on how lenders can meet the ever-growing demand consumers have for instant gratification, and since that include talk of BNPL and alternative data scores among other topics that have been popular here, I decided to record and edit it into an episode.

“So imagine I'm on a shopping site, I've added stuff to my cart, then I get a message that says if you have a certain card you'll get 10% cashback. But I don't have that card. I'd like to get that 10% cashback, after all, I've made a substantial purchase, right? But then it flashes at you and it says, ‘apply now’. So I click on that, go through the credit approval process in a few minutes, get back the card number, the CVV number, expiry date, and use that to complete the transaction get that 10% cashback. That’s instant gratification. And if I'm going for lunch today, can I get that card approved, the plastic embossed, and in my hand before I call my friends and take them out for lunch. That’s instant gratification”.

I was joined by Michele Tucci of CredoLab, Andre Marques of Provenir, and Anit Antony from TransUnion

If you have any feedback, questions, or if you would like to participate in the show, please feel free to reach out to me via the contact page on this site.

Regards,

Brendan

The full written transcript, with timestamps, is below:

Brendan Le Grange  0:00 

"The risk of giving into temptation is as old as humanity. But there are reasons to think that people today are having to work harder to resist it, particularly when it comes to consumer behaviour. Digital technology has made it easier and faster to buy goods and services in an instant, without the delays of processing that once comprised an inbuilt cooling off period". This might sound like a headline from today's papers, but in fact it was from an article in The Financial Times published seven years ago, almost to the day - at a time when Klarna was around, yes, but only just beginning its global expansion, Affirm was only two years old, and AfterPay only a few months old. Welcome back to How to Lend Money to Strangers, the podcast about consumer lending strategies across the credit lifecycle and around the world.  

Today we're back in the Philippines with a different sort of an episode. This morning, I had the pleasure of chairing a panel discussion for TransUnion as part of their Big Data Summit. And since the topics that we discussed include buy now pay later and alternative credit scores, two topics that are already popular on the show, I thought I'd take the opportunity to record our discussion and turn it into this episode.  

I was joined on the panel by Michele Tucci of CredoLab, CredoLab works with neobanks, challenger banks, BNPL... any digital lender, really, to build alternative data credit scores that help increase access to credit, and Michele is their Chief Product Officer. Michele has held product management, business development, and international consulting roles and has delivered projects in 51 countries around the world. He was joined by Andre Marques, head of Provenir's sales for Asia Pacific, 'loved by startups and trusted by decacorns', Andre and his team help institutions to adapt to a digital first world where everything is instant. Like Michele, he's internationally experienced having previously held roles as a strategy director and worked across Europe, Africa, the Middle East, and the Americas. And we also had Anit Anthony, TransUnion Asia's head of digital products. Anit has spent over 18 years in the technology industry in Asia and North America, and has been involved in several digital transformation projects for retail banking, but also for telco and retail. So with guests of this calibre, I think you can see why I couldn't resist turning this discussion into an episode. Join me in a second. 

Anit, I'm going to go to you first. I started with that article from seven years ago, mainly to underline the fact that the striving for instant gratification, for instant approval, isn't something new this is a much more human desire, as soon as we could go digital, people wanted to go faster and smoother - perhaps now, it's just that we're at that point in time when it's a realistic prospect for for lenders. Before we start talking too deeply about who's doing a great job of delivering instant approval, andthe risks inherent in meeting demand for instant gratification, can you give us some context of what we mean when we talk about instant approval? When we talk about instant gratification? What does that look like in terms of a product scenario?

Anit Antony  3:32 

When I look at this, from our from a consumers standpoint, not just in the Philippines, here are some of the expectations that we have, right, what really is instant gratification? How does that actually play out in the real world? So imagine I'm on a shopping site, and you know, I've added stuff to my cart, I get a message that says if you have this card you'll get 10% cashback. I don't have that card. But I'd like to get that 10% cashback, after all I've made a substantial purchase, right? But you know, it flashes at you and it says, okay, apply now. So if I want to now click on that, I go through the credit approval process in a few minutes, get back the card number, the CVV number, expiry date, I use that to complete the transaction. And then obviously that 10% cashback. So that to me is instant gratification.

And I'm going for lunch today, right? So can I get that card approved, the plastic embossed, given my hand. In the meantime, I call my friends and take them out for lunch using this card. That to me is instant gratification. How do we make things like this possible? I don't think we're too far from that, just months away from that.

And cash, right, cash is more important because I was talking to a money lender in Hong Kong and he was actually saying when people actually go to them for cash, they need that cash today. They need it instantly, right, they're not going to wait for weeks. When they go to banks, the say, oh, they need to submit address proofs and income proofs and it's a week, a couple of weeks, and instead they want to go to, you know, moneylenders giving the cash instantly, even though the interest rates are higher. And the BNPL product, Brendan, you mentioned that, but I'm sure the other panellists can also talk about that in a bit more detail.

Brendan Le Grange  5:06 

Yeah, thanks for that. And speaking of BNPL, that's obviously a field that Provenir know better than most, and we're lucky to have Andre Marques here from Provenir. So Andre, I'm going to hand over to you, and maybe you could start by giving us a quick introduction of Provenir, what you do there, and then talk us through these demands that you see from BNPL as well as any other fintechs and lenders that are that are out there trying to meet that increased demand for speed that we're seeing from consumers.

Andre Marques  5:36 

Brandan, thanks for that and Provenir, for those who don't know, is a company that makes smarter decisions faster by simplifying the risk decisioning process. But what allows us to do that is a no code, cloud native SaaS approach. We have over 2 billion transactions per year, and that allows us to help both fintechs, financial institutions and payment providers across more than 40 countries around the world. This includes companies such as BBVA, GM financial, Hitachi but also be BNPL providers such as Klarna, and Zilch.

To your point of how we help our clients. I guess, in this case, I would highlight four things, Brendan. (1) The first thing is, we're reducing the application processes from days to seconds. That's achieved with an instant, automated risk decisioning workflow. (2) The second is reducing the reliance on IT teams in order to be able to make significant changes in your processes, to launch new products. That's achieved as per our no code technology. (3) The third one I would call out is the simplification of integration to an increasing number of third party data sources. And I'm sure Michele will have a chance to talk a bit about the importance of those as well. But more and more as we go forward, then you're looking at connecting to bureaus, but also KYC, AML, alternative data sources and open banking. And that's just increases how much complexity you'll have to manage on your own. We provide a one stop hub that brings together more than 500 partners, including TransUnion and CredoLab and that minimises not only the effort to integrate and manage those, but also takes what I would say is a headache from the internal organisation. (4) Finally, I would call out scalability. We're a cloud native provider. And we also build commercial models to match what are the needs of clients from the moment they incept until they become decacorn status which, gladly, we can call out for several of our clients.

Brendan Le Grange  7:36 

Thanks, Andre. And if you were to put that into more practical terms, what are some companies that maybe we've heard of that are doing this really well?

Anit Antony  7:48 

Happy to, maybe for a bit of context, what I would say, makes typically the common link between all those examples, is really the focus on meeting customer expectations. You'll see today we're especially driven by Millennials and Gen Z's that there's really an expectation for instant everything. And so just giving two simple data points on this: 80% of consumers today rank speed as a key buying factor, and 76% of customers around the world are more likely to make a purchase if an easy simple payment plan is offered at the point of sales.

Andre Marques  8:24 

And this actually brings me to my first example, buy now pay later. And I can talk here maybe a bit about Klarna, a leader in this space and one of our clients. Klarna was a pioneer in what I would say is providing tailor made BNPL offers, not just BNPL specifically to their customers, this is something they were able to do early in their journey because they adopted advanced decisioning solutions to both their customer onboarding processes but also their merchant onboarding processes. Because of this, they were able to perform risk analysis and decisioning as for data provided by the front end, and they were able to return immediately with a decision that was not just fast, but custom tailored for that specific consumer. The second example I would give is in a different space, is in secured lending. I'm going to tell you a similar story from one of our clients Instabank. So Insta bank is a disruptor I would say, in Norway, where they a few years back basically announced the ability for their customers to get in less than a minute a loan approved and specific offers designed for those customers. And they did so with minimal information requested. To me this is more than a product, it was their differentiation point to the mark. And the reality of that is they have since grown their lending book to more than $150 million. This is a fantastic story of how in this case a challenger but I don't see a reason why that couldn't be we an incumbent player was able to shake up the market. This was possible also because of the agility of systems and, well, I don't want to trump too much on systems and how we are a strong solution provider that can be used, but really the challenge that Instabank had to fix before doing this was to replace a legacy system that relied on 1,000s of lines of code that didn't provide them the flexibility to move forward.

The third and final example, I would give you, I guess, to this Brendan, is credit cards. And I know Anit touched a bit on this as well, one of our global clients that's focused on consumer lending, and I would say is positioned as a digital bank, and their focus was to have a credit card, I would say, as a part of attracting customers into their bank. Their view was simple: customer experience needed to be focused on speed and digital. With this, the result was quite straightforward - we need a digitally seamless, available credit card application process. So in a nutshell, clients expect to be able to apply and immediately get a credit card as per Anit's example. We were actually able to deliver a decision response in less than five seconds, despite more than twenty 3rd party integrations, and more than 5,000 business rules, as well as machine learning models. And I think that's really the challenge you need to be able to consider if you want to get into a world now of instant gratification.

Brendan Le Grange  11:20 

So, in my introduction I read that passage from the Financial Times, as a reminder that the search for instant is fairly well established. But as you were speaking there, you helped to crystallise something in my mind, when you brought up the idea of incumbents versus disruptors, because for a while now, lenders have been able to offer instant in the form of pre-approved loans or shadow offers that were sitting in the background. But they could only do that to existing customers, because they would look at the data of the existing customers, and then decide what they would lend if they were asked. But for the bulk of people sitting outside the organisation they couldn't. Now that gave incumbents an advantage because incumbents had customers. And it was very hard for a disruptor to grow, because they needed to bring customers onboard and only then could they offer those customers the best possible experience. And particularly if I think like a country like the Philippines, logistics were just holding back that growth. So if a customer had a credit card, they were on the credit bureau, sure, you could offer them a digital experience that could be very fast, approaching instant. But the bulk of people weren't there. And to try and service them to try and bring them on board, you had to rely on either very unreliable or very expensive processes to get the word out. But now the industry is starting to get its head around alternative data, thanks in part to to companies like CredoLabs.

So Michele, let me bring you in here. And can you talk a bit about what alternative data and the scores powered by alternative data brings to the table, even when you look at a market that has a strong credit bureau, has very predictive credit bureau scores?

Michele Tucci  13:04 

Thank you, Brendan. And thank you to TU from for having me here. Having credit bureau scores, credit bureau data is important. At CredoLab we are not advocating for changing that - we position ourselves, as well as other alternative data providers and alternative score providers, as complimentary. So traditionally, you had a credit history, and it was good to assess existing customers, to assess the creditworthiness of customers that are known to the credit bureau. But there is, you said just now and Andre as well, what if the customer is new to the bureau? How do you provide an instant gratification and how to bring those customers into the bureau without them having to go to payday lenders, for instance, or to illegal lenders?

So the two sets of data work really well together, they complement each other. Of course, the predictive power of alternative data is higher, where there is no credit bureau data or there is a thin file. But the uplift in terms of marginal gini points is also important on the bureau population. And the reason is that we are assessing two different components of the same customer, the same applicant - the bureau tends to focus on the ability to repay, when we look at CredoLab's data, we focus more on the behavioural aspects, we focus on the willingness to repay. So when you combine the two, affordability and ability to repay with the willingness to repay, you have a stronger assessment of that particular individual.

Also, I'd like to bring in examples of tech companies that claim to have a lot of data, you can think of Grab or Traveloka, for instance, in Indonesia. So massive consumer brands that offer a lot of services to customers, and they are all into buy now pay later now, they are offering unsecured loans to their customers, and yet, they are using CredoLab's technology to compensate the lack of data that even they have for their own customers. So you have transportation, you have wallet, and you have food delivery type of data, how many customers are generating enough data across the three services to be able to be used for credit assessment purposes? I'm not gonna disclose the information, but it's not as high as you think. So, buy now pay later, as well imagine Traveloka welcoming a new app user, they have no travel history, they have no ecommerce purchases that they can rely upon to make an assessment. This is just like a thin file or no file at all with the credit bureau. So what do you do? Are you going to close the door in front of this customer? Or are you going to give him a chance to be part of the ecosystem? And that's where we come in, I believe CredoLab's is a facilitator for financial inclusion. And it's also a facilitator for personalization, for better customer experience.

Brendan Le Grange  16:26 

Yeah, I think that's a very good point as well, because those of us that have grown up in the traditional data world, we would look at it consuming the data be very structured, it's a certain number of fields and they're all fully populated every month. Whereas now in the big data/ alternative data world, you might have a customer who's got three or four sets of one type of data, but none of another, different exposures to different services, and you're pulling all sorts of little fields, and none of them are going to be as strong necessarily as something like delinquency, but you've got so many that you can still build a picture, but you need to think about it a bit differently. It's not... you don't just look at it on big grid like you're used to, you've got to play around.

Andre, Michele brought up the NPL again, now, maybe we can pause briefly on that. You've got a lot of exposure at Provenir to the BNPL Industry. Can you maybe start with a little bit of sort of a dip in the market? What's the market looking like right now? Both, I guess, locally, if there's news in the region and internationally?

Andre Marques  17:25 

Absolutely, yeah. It's an area we're proud to do a lot of work on. And, you know, it's, I guess, unexpected by most how fast it's been growing. There's predictions that it will grow still 15 fold the current volumes by 2025. And I think that clearly calls out the first naysayers that thought the industry was a short term trend. The recent deals, particularly Square buying Afterpay at $29 billion. And both PayPal and Goldman Sachs making inroads at more than $2 billion each are clear indicators that there is value to this. And I think a lot of value comes beyond and above what BNPL is, as a specific service. It relates more to the ability to match customer expectations right now. I think another interesting point that we can see at this stage is how you start seeing specialisation within BNPL providers. So there's things like BNPL offers for vet care, for rent relief, utility bills, way and above what were the original purchace use cases.

Anit Antony  18:37 

Another area I'd like to highlight, I guess, is regulation, right? It's been early days and regulation bodies are still looking into how they're going to do it, but if there's one thing that seems more or less consistent, and everyone can align to, it's that regulation is imminent.

Andre Marques  18:52 

In the long term, I think this will be a clear fact. And the focus will necessarily be on credit safety, which will force retailers, consumers, and I would say providers to work together to make sure customers' interest is at heart. This will necessarily and once again force some changes, and the ones that have not prepared for it will probably struggle.

Another point I would like to highlight you know, because as you said, there's a lot of headlines on this industry, but I'm not sure everyone sees the non success stories. I think that's an interesting point to keep in mind as well. I would say some are no longer with us. Others, you can see struggles and, to your point, I'll give an example from the region: just earlier this week Hoolah, a BNPL provider in the region, headquartered here in Singapore, but with presence in other markets, have announced both the exit of the current CEO but also cutting jobs as they head for restructuring efforts. Looking back I think there's a few interesting points I would like to call out here and I only have an outsider's view to the story, but you will see who will have been visibly quite focused on growing their merchant network, I'd say expanding within the region. And as they've done that, they've been also building their own technical solutions while betting on what I would call is a simplified risk management approach where one fits all. I can't stop wondering if this is a matter of too many fronts diluting their attention.

Brendan Le Grange  20:18 

But lenders will be looking at that those growth numbers, those sale prices you've just mentioned. And when you look at the idea of buy now pay later, logisticaly is seems not that different to what we've been doing in the past. And you're wondering, can we achieve something from this as well? So when financial services seek to follow that path, what are some of the pros and definitely some of the risks that they take on board and need to think about managing?

Andre Marques  20:43 

Oh, absolutely, Brendan, and there are certainly several commonalities here that I don't think are specific to a FinTech focused on BNPL alone. I think a traditional bank would face the same.

The main one I would necessarily highlight is customer experience. We're in a Millennial Gen Z generation world, where Spotif and Netflix set the expectations of experience and instant gratification. I would dare to say that it's important that providers don't dictate the terms to the customers, it's important that we use data to get insights on how to better serve them. This is a case where we need to think of rewriting, a bit, the script of how to meet the needs of the new generations, as well as how they think and feel about credit and access to financial products. Just in 2020 alone, there was an increase of 63% of financial app usage. That said, it's not just about the digital experience, but it's also matching the customer expectations of engagement, the front end and the back end, talking the same language, if you want.

The other thing I'd say is the access to data. And the access to data has exploded. And so this has created a completely different ways that data can be accessed, but also explored. You have every second 7,000 tweets, I think 30,000 Facebook likes, but if you look at a traditional bank platform, it can equally generate a million transactions per second or 100 billion per day. The amount of data that's created, it's fantastic, and to a certain extent, hasn't been fully explored to the advantage of customising the customer experiences. This, to me is really a key point, and financial institutions have this massive opportunity to provide real time engagements that match what customers are looking for.

Brendan Le Grange  22:34 

Yeah, I think that's probably the area that's going to be the biggest stickiness now. I don't want to besmirch the whole banking industry, we know they've made a lot of strides recently to deliver better customer experiences, but painting in broad strokes, we would normally think that fintechs are the people that are great at delivering experiences, that are great at trialling out new data types. They're less constrained by their investors, they're less constrained by regulation and reputation risks. So that's where we would expect experimentation. They're also out to win customers from someone else, often, so they tend to be focused on delivering great experiences. But on the other hand, banks have lots of cheap capital, they have existing customers, they have good reputations. Michele, if I can come back to you here. Why don't we see more partnering - or easy partnering at least - between big established banks who maybe can reach out and work with a FinTech to get the best of both worlds?

Michele Tucci  23:39 

Depends on how politically correct you want me to be, right. I think the biggest hurdle that at least we face as a small company pitching the large clients like a bank, is the level of scrutiny that we receive from compliance, from info security, scrutiny around data governance, data ethics. In the Philippines specifically, we've been approved already five times by BSP (Bangko Sentral ng Pilipinas). Not because we had to, but it was easier for the bank for us to get approval from BSP. So we went through the process. We are now ISO certified, which is an overhead for us, but it helped to go smoother through the onboarding process. So I think what is preventing more banks from partnering with fintechs is internal processes, which are there to protect the bank and they should be there, I'm not arguing that they should be removed, but if there were a sandbox within the bank to test the new solution to make sure that it complies. I think it would be faster for the bank itself to appreciate the kinds of improvements that they can bring in if they partner up.

There are also some instances where banks are willing to build versus buy. So they believe, I don't know why, that it would be faster for them to build the solution rather than buy it from a SaaS business. There are new banks coming into the Philippines, we know of Tonik, we know of Tyme bank, and they are already launching lending products. Why? Because they realised that offering deposits alone is not enough to be profitable. The banks are already profitable, so if they were to allocate a particular product to a, probably a riskier segment or a younger segment, that's where they could be partnering with fintechs, to showcase how well something can be executed in partnership rather than alone. And we have two partners, here we have TransUnion, we have Provenir, another way that banks could be partnering with more fintechs is to just leverage what capabilities TransUnion or Provenir have already built into their offering. So if the bank trust the due diligence that TransUnion has done, then perhaps they would want to work with a FinTech through TransUnion. And that means less paperwork, and go to market faster. I wish I wish things would move faster anyways.

Brendan Le Grange  26:27 

Speaking of TU, let me come back to you Anit. We've spoken today a lot about the benefits of delivering instant. And we've spoken a little bit about the risks, how to mitigate those when it comes to credit, you know, building new types of credit scores with alternative data. But what about the fraud risk? Obviously, as we try and make it as smooth as possible for people to come on board, we also potentially open the door for fraudsters to take advantage.

Anit Antony  26:55 

Yeah. And that's a big question. Right? So I was sort of waiting for you to ask me that question.

So yes, that's what we're trying to achieve is instant gratification. But every 10th transaction in the Philippines is found out to be a fraudulent transaction. So think of all the data breaches that are happening, right? So identities are stolen every single day, where, if you're looking to the Philippines, you have the Cathay Pacific breach of about 100,000 Filipinos data was compromised, Facebook had about 500 million users that was hacked, and out of which 900,000 were from the Philippines, there's the passport breach well, so when you look at all this identities that are getting stolen, then not to mention phishing, that's happening almost on a daily basis, right. And to buy all of these fake identities, and people online, and the dark web is again, easy.

The fraud's getting really, really sophisticated. So you'll start to see fraudsters apply with multiple ID cards with the same facial profile, for example, right. And we've seen a lot of that. We will see applicants who apply multiple times with the same ID, but with different names. And all that mix and match happens. And the lack of security features in some of these ID cards that the banks accepted makes it even harder. And relying on just one fraud provider is like putting all your eggs in one basket. So yeah, there is a big increase in the front end, it is expected to go up because all the banks now have their own digital initiatives. And this is a big risk that is there in front of them.

Brendan Le Grange  28:20 

Yeah, and my first job was in credit card fraud so I always try to remember the bank's side. But as soon as you're a consumer interacting with that, it's incredibly disruptive to your life when when they get it wrong. So you know, if you're making an application, and it gets paused, because they're doing a fraud check, or as I had, I made a payment when I was moving house, I paid the rental deposit, and it got flagged by fraud for a warning, they sent me an SMS, I didn't read for that day so they shut the account for a bit. And I was really annoyed, because I had to phone in the bank and sort out... try and remember old phone numbers and whatever else was in their ID checks. That's the downside. It makes such a bad customer experience as soon as you stop the customer. But it's obviously impossible to hit only the fraud. So how do you defend against fraudsters? What is it that you're looking for that you can catch a fraudster? What's giving them away. But at the same time, that's not just gonna catch an innocent customer here with a dirty ID card or with something?

Anit Antony  29:20 

That's right. That's right. And so when you look at the balance that we need to come up with, right, between instant credit and being able to catch fraud, and the problem there, though, is if it affects the customer experience, if it's going to add a lot of friction in the process, the good ones go away as well. Now if I catch the bad ones, but the good ones are going away, as well. So now how do you strike that balance?

So you need to do very reliable fraud checks. But you will have to do them discreetly, without hindering the consumer experience. And that's where at TransUnion we've been looking at ways and means to do this right, not just in the Philippines or APAC, but also across different regions. Now, how do you do that? So one of the things, yes, we have a face scan, and we have an ID scan, we got to match them - but you're always going to get two false positives and false negatives, right? You need to add layers of defence, right? One of the best ways that that, you know, to me is the device.

You know, at times, it might not even be a human applying, so it could be bots, could be simulators. Now, one thing that gives the fastest way is really the device. Chances are that you might use one, two, or three, or five devices to do all these fraudulent activities, you know, we can actually pick up things like velocity, or how many applications were made from this device? What is the longevity of this device? And then is there any evidence that it's been reported against this device. So there's a lot that the device gives them, a lot of information that we get from the device totally seamless to the user, he doesn't know what's happening. So let's just say we have someone that does the selfie scan, an ID scan, and let's just say that goes undetected, it's fraudulent goes undetected, the device is going to give it away. And not only that device, but all the linked devices to that. So you will actually be able to identify all those different types of fraud that you weren't able to detect with just a face scan, and Id scan, you can still have that, right. But you have to go beyond that.

Now, what else would you check? You will check email, for example, same things right would be checked about the device, you check the same activity with the email, like, what's the velocity? How frequently is this email address being used? What is the longevity was created yesterday? Or was it just created for the sake of this application? You can exactly do the same thing for phone number. And you're looking at geo location. And then I think this one takes the cake for me - we also have a service that you can check whether the credentials have been compromised. So if this identity is being sold in the dark web, there is a way for us to detect that. And you also have to see whether that sort of identity exists in the credit bureau as well, right? If one layer doesn't catch it, another one definitely work. Now we look at all of these right device, email, phone, geolocation, compromised credentials, there is nothing that the user needs to do - of course, you need to scan his face and scan is ID - but that's the only friction point everything else is seamless.

Brendan Le Grange  32:12 

It's no longer the questions you can never remember the answer to it's, it's happening in the background.

And then I guess, before we close, we need to address COVID. I know we're already a little bit over time, but there's no way we can ignore that, really. Andre, when we look at COVID, and the lessons it's taught us, how do you see risk management of the sort of large shifts that external factors can bring - be they the COVID epidemic, or as you brought up earlier, regulation?

Andre Marques  32:40 

Yeah, absolutely. And I think this is actually true for both the regulatory shifts the pandemic, but also the ever evolving customer expectations - what to me, glues, all of them together is the need to be flexible, and to be responsive to those changes. And that's where, you know, you want your business teams to be able to react, and not to be dwelled by limitations on technology or hard coded rules that are difficult to change. And that's probably what drove our customers to success, was how their business teams identify the changes required, but also were able to quickly act on it. That in a nutshell, is how I would summarise that. And maybe one final example I would give, you know, the pandemic that you called out, right, several financial institutions were forced by regulatory need even to change their credit policies. And that's something that in a traditional world with legacy systems is a three to six months project, a change request to be discussed, and the need to count on a vendor to support that. That's what really we're trying to get out of the way, make sure that business teams can directly make those changes. And one of our clients by can UK facing this exact challenge. They were able to do it themselves and do it in just a few weeks.

Brendan Le Grange  33:59 

Hopefully, we won't ever have another thing where one week, it's normal and next week, everything's changed. But we were working with a client who had to reassign laptops from one group of staff to another because not everybody they needed to work from home had laptops, and it is just sort of underlying how these old systems were so structured that suddenly we had to change was really difficult.

For me, Michele, the other big question about the sort of disruptions is obviously shaking up all our lives but more than anything, probably our online lives. So more people online, got different people online. That's obviously created the potential for new data for you but also potentially changing the data you've already got. And that may call into question things like score stability. So, you building scores of alternative data, can you talk to us about score stability, has that been damaged in this rush to online?

Michele Tucci  34:53 

It's a tricky question. So from a data point of view, COVID has been good to us, as insensitive as this may sound. But people have spent so much more time on their mobiles. According to App Annie, in Asia, people spent on average five more hours per day on their mobile phone. That means also that they have generated a lot bigger digital footprint on smartphones. They have downloaded so many more apps, tried different categories of apps. And so yes, you're right, we wondered and our clients wandered with us, how stable they've been able to score in the presence of such a different shifted kind of behaviour.

And much to our surprise as well, the score was extremely stable. We calculated, also in the Philippines, the population stability index of our models, in the worst scenario was a change in psi of only 0.03. Which means that we didn't even have to calibrate the model. How is that possible? So you look at a digital footprint, we access the permissions that we receive, we access, if the mobile is three years old, we access three years worth of data. So when we identify the micro behaviour patterns that are predictive of risk, we go back not just six months or three months setups, like a telco would do, but we go back in time, and we can see how stable these behaviours are. Of course, the risk profile of these customers, incoming population was a lot higher but in terms of modle predictiveness, that date didn't change. So good news for us, good news for our clients.

And I would say if I had to make a wish was that more clients were already ready with digital channels so that they didn't have to shut down branches and also shut down originations. But that's for the next pandemic, hopefully, in 50, 100 years time.

Brendan Le Grange  36:59 

Thank you very much. Yeah, so I think that's a whole discussion on its own. I want to close by giving the floor back to Anit, we've only got a minute or two left. But can you just try and sum up how a lender can think through such a big problem? You know, we've spoken about new data sources, new types of scores, new types of very sophisticated fraud defences, how does that all get pulled together?

Anit Antony  37:25 

What we just did in the end was to combine with our partners and come up with a solution that you don't have to do employment checks anymore, you don't have to do income verification anymore. And the address verfification. This guy's spending a lot of time in the office hours, that location happens to be his office address, and where he spends time in the night, that is most likely going to be his home address, again, I'm talking pre-COVID not post-COVID, right? We have partners, that partner with telcos to give you that sort of information. All these things, right, instant decisioning, super quick fraud checks without hindering customer experience - that meets that 730 rule, right, seven minutes to apply, three minutes to process, with zero human intervention.

So that's the kind of thing that we're trying to offer. Things like the ability for you to ping a company email address - let's say brendan@transunion.com, for example, and you provide that to the bank, the bank should have the ability to ping the email, and that should be a valid email. And that it goes a long way to show that you are indeed employed with that so employment verification call that can be taken away that way, right? Income estimator for some of the data products that we have will help you with the income proofs. Then you're saying that, okay, employment is verified, income is verified with TU's income estimator. So, you know, this actually completes the whole process of lending which can happen. I mean, those use cases that I talked to right at the beginning, can be achieved. One last mile that we need to run is with Visa or MasterCard, our friends there, they will be able to generate that instant card, which is still pending as part of our product, but everything else for instant decisioning is something that we already offer today.

Brendan Le Grange  39:00 

Right. Well, thank you very much. Thank you to everyone from the panel. And thank you for listening. This has been How to Lend Money to Strangers for podcasts about consumer lending strategies around the world and across the credit lifecycle. I've got a really good show coming up next week, so join me next Thursday for that.