From Congress to the Silicon Valley, data usage has become a hot button issue. In this bonus episode of Innovation Answered, we take a deep dive into the world of data and how Experian’s DataLabs spread innovative data solutions across the company.

Related Resources:


[ANNOUNCEMENT]

Kaitlin Milliken: Hey, we’re hard at work on the next season of our show. But we need your help. We’re doing an episode of the show where we answer all of your questions about building teams, starting an incubator, creating a network of champions — really anything you want to know. So please ask us anything. And don’t worry, you can be anonymous, or use your name and get podcast famous. So record a voice memo on your phone, or write down your questions and send it to kaitlin@innovationleader.com. That’s kaitlin@innovationleader.com. My email will also be in the notes for this episode. Now, on with the show.

[THEME MUSIC]

You’re listening to innovation answered the podcast for corporate innovators. Right now, we’re in between seasons, but we wanted to leave you with the bonus broadcast to hold you over. In this episode, we dive deep into the world of data by focusing on one company: Experian.

Originally focused on credit reporting, Experian has broadened its range of work with all types of data, software, and decision analytics. The company also has a long history of acquisitions. And today, experience operates in 37 countries around the globe.

To spread innovative data solutions across these different locations. Experian runs four DataLabs in San Diego, London, São Paulo and most recently, Singapore, projects in the labs focus on financial services, telecom, and healthcare.

To find out more, we called Eric Haller, the Executive Vice President and Global Head of Experian DataLabs. During our conversation, Eric discussed Big, Data innovations in security, and the talent that drives the DataLabs.

So Experian used to focus mainly on credit reporting, and now the company works with all sorts of data. What caused the shift in your business model?

Eric Haller:  Well, it kind of came in to two ways, the first wave was really more from the perspective of balancing out our investment portfolio. You know, the core of our business was in credit reporting. Credit reporting is very tightly tied to gross domestic product growth rates in national markets. And so we thought, you know, best to balance ourselves so a cyclical environment that we have other businesses that do well, in times when the economies may not be.

So we expanded ourselves into healthcare, and automotive, and government, and things that would balance things out, and give us a broader perspective around the data that we pulled in and the problems that we solve.

But I’d say in the last eight years, there’s been a bit of a change even from that. And that’s really stem from the fact that there’s been more data captured around individuals and businesses than ever before. And we saw the opportunity to look into data sets that we’ve never really had analyzed before, to see if we can get a better perspective on solving the same kinds of problems that we typically solve, which has to do with with risk or fraud, and building a better customer experience through better marketing.

Kaitlin Milliken: So now your team works with all different types of data, and you have DataLabs at Experian, can you explain what those DataLabs do for people who may not be familiar with that part of the company?

Eric Haller:  So the DataLabs are all about fusing this data or this raw data that’s coming in from all these different markets and businesses with artificial intelligence. We’re the group that works on breakthrough experimentation in a very safe and controlled environment. We have the benefit of pulling in all of these data assets into an environment that’s kind of like the Fort Knox of data. With the way we manage our credit bureaus and the high level security and the vigilance that we have in place, we pull that data into that environment and allows the lab to engage in trying to solve things that really push us into new products and new markets.

Kaitlin Milliken: And can you talk about what those new products and markets are? Just some of the things your team is focusing on and any projects that you can share?

Eric Haller:  So in the area of consumer credit, we’ll look at a whole bunch of different things. From when a consumer doesn’t have much of a credit footprint, so they’re new to a market or they’re young, or they just haven’t had that opportunity to establish credit. We’ll look at alternative data sets to try to figure out what other pieces of information or engagements that our consumers engage in. And how can we extract enough information from that, that can show that their credit worthy?

So we’ve developed the product in markets like for Africa, or Southeast Asia, or environments where the amount of credit information isn’t being collected, as strongly as like, say, here in the United States or in the UK. And the actual way that you use your mobile phone — how you’re connected, who you’re connected to what apps you pull in — we’ll pull enough intelligence from that, so that we can build out a credit footprint. So we can measure character and your ability to repay, and banks can use that data to make a lending decision.

A consumer has to be completely, you know, aware and understanding of what we’re doing. So the applications that we develop, gives the control to the consumer. In fact, that’s a trend that we see this happening over time, is the consumer will have more choice and more control over how their individual data is managed, even in the credit process. So in that case, we built a product that allows the consumer to say, Yeah, I understand that, you know, I can’t get access to credit credit today, if you want to pull information from my phone with my permission, then I would love to see that use for that I can I can get credit later.

We do that kind of thing, all the way to the actual customer experience. So here in the States, we know that a lot of our clients like to see less friction in the process of getting credit to an individual. So you’re in a retail store and you go shopping, and you fill up your basket of goods. You walk up to the cash register, and the associate behind the counters asks, “Do you have a credit card with us?” And if you don’t, then they say, “You know what, would you like one today? we can make that happen right here at the register.”

That was super cutting edge like 20 years ago. But today, it feels pretty awkward, you got a line of people behind you everybody’s waiting, they want you to move on with your goods. You’ve already bought…you’re already shopped for your goods. So if you knew you had more credit, you might have bought more things.

So we came up with a process that allows you to obtain instant credit for each from your mobile device, just by texting a word that a retailer might provide on a poster or an ad to a short code. And in less than a minute we can go ahead and pre-approve you for a line of credit in the store. And then you can actually go ahead and use that around the store.

So, we were the first ones to bring that kind of technology to market. We went through a number of different approaches to get instant credit to the point of sale. And in that case, it was the easiest to implement. And it’s getting traction really quickly now have over 30 products in the market with another 14 that are in handoff to our businesses to get out into the market.

Kaitlin Milliken:  I’d love to dive in a little deeper into the idea of DataLabs all over the world. When did your team decide to expand? And what was the purpose of opening DataLabs in these different markets?

Eric Haller:  So the first lab we created about eight years ago here in North America, and the notion was kind of as we discussed, we saw the world changing, we knew that there was more data coming into the market. And we wanted to create an environment that allowed us to take risks — safe risks, but risks — to experiment with those datasets to see if we could leverage them in some way to benefit consumers and businesses.

Our board of directors said, “Can you do the same thing in other markets?” So we took the challenge head on. We established two more labs: one in the UK, in London, and the other one in Brazil, in Sao Paulo, and tried to do the same thing.

We arm each lab with data scientists. We put all of our scientists through rigorous testing. Most of them have doctorates. I’d say about 70 percent right now of our data scientists have doctorates. And the average data scientist that we hire receives about six offers from other companies. The only folks that we lose to, when we do lose for talent, is Google and Amazon and a number of the the upper end of the pyramid for data science in Silicon Valley. But otherwise, we’re capturing our fair share.

And so we couple those best-in-class data scientists with leaders in product development and consulting. So leaders in product development consulting are the ones that frame problems with clients and business partners, and work with our businesses across the globe. And it’s that combination that we see as kind of the best combination to move quickly and get products out in the market.

The labs that we had in Brazil and the UK also did as successful as our lab in North America. And so earlier this year, we announced building a fourth lab in Singapore. So now we’re in Asia-Pac, and we see a lot of benefits associated with having labs around the globe. They don’t work on the same problems.

In the US, we see a lot of work being done in optimizing the customer experience and deep analysis using artificial intelligence tools, and driving things like explainable artificial intelligence and our credit models.

In Lat-Am and Asia-Pac, many of their markets are forced to work with less data sets. So the solutions that they’re focusing on is, “How can we extract more insights with less data or more disparate data sets, and kind of knitting together a quilt of data around a consumer business so that we can do a better job of assessing risk or marketing opportunity?”

And in the UK, right now they’re suffering — or they’re being faced with — an aging working population. So we’re working on “How do we develop employment solutions where we can map an aging population to the right kinds of work?” They’re also dealing with open banking, which is this ability to allow a consumer to port their demand deposit account or checking account from bank to bank seamlessly. So we’re building solutions around that.

Knowing that open banking is likely going to move into other markets around the globe, every business that we have can benefit from the work that’s being done in the UK. Just like the businesses in the developed markets can benefit from the solutions in the developing markets, because it’ll allow us better insights. And how do we how do we attack problems like financial inclusion, for those that are unbanked or underbanked? So the key for us is to make sure that there’s constant communication taking place, and the sharing of intellectual property across these labs, so we can continue to cross pollinate the different markets that we’re serving.

Kaitlin Milliken:  A lot of the projects that your team works on, involve the customer directly, so maybe it’s an app or something that a customer is able to have access to. Not every customer really knows about big data. Some may even be a little wary of it. It can kind of sound scary. How do you test ideas with these end users, especially those who may not be bought into the idea of having their data used?

Eric Haller:  So we serve different constituents or stakeholders in the process. Clearly, we do offer a lot of products directly to consumers. I don’t think it’s as much about talking about the data itself as much about the value proposition — so what they wind up sharing, and what they wind up receiving in return. That’s more of the consumer side.

On the business side, sometimes it’s a bit different. So we do a lot of work behind the scenes, with businesses. And I’ll give you an example. Like, if you’re a large bank, and you’re underwriting or you’re writing credit lines for small or medium sized businesses. Those small businesses, they struggle to get credit, because it’s hard to establish lines of credit early in their in their process, but a lot of them are on Yelp. And a lot of them are on Facebook, where consumers actually wind up having an opinion on whether the business is very good. They might report that they’ve checked in to the business. They might write comments down. And we found a way to leverage that data to help provide insights to those small businesses that are just getting out of the gate or just growing. So in that case, you know, big data is actually helping small businesses by allowing them an opportunity to receive credit based on how consumers are viewing their businesses.

Big data itself has been around for a long time. But I think what’s happening more now than ever, is rather than having the conversation centered around Hadoop or big data infrastructure, it’s more about what that social impact is to the consumer small business. And I think that’s where it’s getting really exciting, because we can see those benefits actually, you know, taking shape in the consumers’ hands.

Katie Milliken:  And when you’re coming up with these ideas and projects, do you bring customers in and small business owners in to test them at all?

Eric Haller:  On the consumer side, we do a lot of market research. And whether that’s behind those two…or interviewing. And with businesses, especially larger businesses, it’s pretty much a direct sales engagement. We’re working with large banks. In that case, the meetings are more, you know, face-to-face. And we’re talking, we’re working through problems side by side.

Kaitlin Milliken:  So data and security are always really hot topics whenever it comes to this sphere. Can you talk about any security innovations that your team is working on?

Eric Haller: One of the key areas is around cyber security. The solutions that are in the market today are mostly centered around network intrusions and what happens if somebody were to intrude within a network and identifying maybe some damage.

We are focused more on the consumer side and the impact as an employee within a business. One of the areas that experience leads in, is in detecting data or information that is being traded over the dark web. So this is a world where consumers’ data that’s been compromised is being bought and sold without anybody’s knowledge. And we have a way of extracting this information and alerting people that this is taking place. In fact, we offer a service for consumers that allow us to remain vigilant on their behalf to do that.

We’re taking this information, and we’re combining it with other data that we have across the Experian enterprise. And we’ve we’ve determined that we can actually predict — based on an employee base around a business — that enough data has been compromised around these passwords and emails, that it makes them more susceptible to cyber attack. And so it’s a model that we’ve just recently built, and we’re actually out in the market, working with those to test it in the wild to see if the results that we’re getting in our labs play out when this is actually being applied in the market.

Kaitlin Milliken:  What metrics do you look for? And when do you decide if an idea is a success, or if it should be considered a failure?

Eric Haller:  The biggest challenge for us is just determining which ideas to work on. We have a massive pipeline of opportunities that come into our labs, but we literally get over 1000 different projects per year across our four labs that we have to evaluate, because we can’t work on 1000 projects. In fact, you know, really, it’s a very small fraction of that, that actually we can put those really smart data scientists against and consultants and product leaders. So that is the biggest gate.

And so we have a very quick way of working through those opportunities. And we look at two major criterion. The first one is the magnitude of impact. How big is this opportunity? If we focus on it, will it move the needle for our customers? Will it be impactful enough for them that they’ll get enough value out of it? And will be will it be impactful enough for Experian? Will it actually generate enough revenue from this opportunity to make it worthwhile?

In our labs, we need to work on the very biggest and strongest opportunities. That’s our function: to create new markets and new products for the business. So they have to meet a certain threshold in terms of the magnitude of impact.

The other thing is about our ability to execute. We want to make sure that the opportunities that we take on are in the realm of possibility. That at the end of the day, we have something we can we can point to and say we were able to we were able to solve that problem.

We don’t mind taking leaps. We are committed to being early adopters of technology. In our labs, we’re the first ones when Stanford or Cal Berkeley or University of Toronto put out some open source on new technique around artificial intelligence, we’re pulling it in and we’re looking for applications. But what we are concerned is taking too many leaps and trying to boil the ocean, when we know it’s really not feasible.

But if we can make those two thresholds, we, like anybody else, will score those opportunities and, and rate them and then focus on the very best.

Now you asked about how we measure success. And in the end, you know, our labs aren’t measured in revenue. But ultimately, these products need to generate revenue for Experian. In fact, in the last five years, revenue associated with lab products have grown 120% year over year for the last five years. That’s an important goal. I don’t know if we can keep that up for for another five years. But the point is that our products are being used, they’re being sold, and that’s our ultimate line of a line of success.

Before it gets to a business, to get out to market to be sold, we test our products in the market. So we go through the classic beta tests and proof of concepts. And then we look at, “Did we solve the elements of our hypothesis that we had set out to do?” And if we do, is it something that is viable to scale? And if we can scale it, we move it to our businesses, and then they go out into the market.

Kaitlin Milliken:  We always like to end with sort of a forecasting question: what will the world end usage of data look like as we move into the future?

Eric Haller:  I don’t know if you follow the World Economic Forum–it’s something that we stay close to here–and they’re very focused on data these days. They announced 10 areas of focus that they believe are the future around data. And I can tell you that our labs are very active and focusing on many of those areas. So financial inclusion, and how data is going to be used to make sure that the billions of people around the globe don’t have that opportunity to have a banking relationship don’t have an opportunity to extract credit, that those are addressed by extracting data in new ways — like the one I described with your mobile phone — in new ways to allow them to come into the market.

Now there are areas of digital identity. You know, in the US, we take for granted the fact that we can we can seamlessly go on our mobile device, go through apps and access many services. Or we can go on the web, we go through our television, and even now we can go on Alexa or Google Home and start purchasing things on those devices. That’s all made possible through the ability of our identity to be digitized and be validated or authenticated, and be associated with all those products and services.

But there are billions of people in the world that don’t have a digital identity. So being able to figure out how to make sure that everybody has that same ability to access the same goods and services and be a part of their local economy.

Another area that — we’ll say the “the dark side of data” — that’s going to be addressed over time is around sex trafficking and money laundering. That’s another area that the World Economic is focused on. And that’s an area that our labs are very aggressively focused on, making sure that we can use data in a very good way to identify those kinds of bad things, and help the authorities address those and get them out of the world.

And things like mapping jobs to people, a lot of us take for granted because we can go on LinkedIn or other websites to get a job. But in other parts of the world–and employee rates can be very high–there are a lot of people that have skills that they can offer job markets, but they’re not mapped to those opportunities. They don’t have that access. Employers aren’t aware of those skills. And so we’re working to start building out that network of how to map people to the right jobs.

And so our labs are working hand in hand with a number of agencies, global agencies, to try to address those kinds of problems.

[ACKNOWLEDGEMENTS]

Kaitlin Milliken:  We hope you enjoyed this bonus episode of our show. Special thanks to Eric Haller for sharing his insights. Be sure to subscribe to Innovation Answered wherever you listen to podcasts. You can catch up on past episodes of this show and get other great bonus content on our website, innovationleader.com/podcasts. Our third season starts in September. But until then, stay tuned for more updates. Thanks for listening and see you next time.