With Big Data suddenly on everyone’s lips, it’s easy to forget that certain aspects of life have been steeped in the interpretation of large data sets for decades. Sports fans can instantly rattle off reams of data relating to their favourite teams and players. The insurance industry knows pretty well how to calculate risk. And credit rating agencies have been evaluating people’s creditworthiness for as long as we’ve had the concept of credit.
A pioneer of the modern industry, Fair, Isaac and Co. was founded in San Rafael, California in 1956, then went public in 1987, and introduced its famous FICO Score in 1989. Today FICO counts nine of the top 10 Fortune 500 companies, two-thirds of the top 100 global banks, more than 90% of the U.S.’s largest financial institutions, and all of the big credit card issuers among its clients, along with the biggest names in retail, pharma and government agencies.
In every meaningful sense, FICO has always been in the business of predictive analytics, a term that has achieved buzzword status recently, but has been practiced to scale by only a handful of companies for a significant period of time.
While the rest of us grapple with what we believe to be early days in the Big Data revolution, it’s worth noting that we are all playing catch-up with those few companies that have been doing this for a while, FICO among them.
One of the consequences of the Big Data revolution, and of the sudden availability of cloud-based data processing solutions, has been the potential for leveling the playing field for small-to-medium size companies, who can now access tools previously only available to huge companies capable of opening and running large data processing centers.
Kevin Deveau, who is personally based in Montreal, has worked for FICO Canada for four years as Senior Client Partner, managing the company’s banking, government and retail accounts. Recently, he was appointed Managing Director of FICO Canada. In 2009, when Canadian Tire decided to get serious about implementing its Canadian Tire branded MasterCard, it selected FICO Canada to handle the optimization and decision modeling. That’s one example from among FICO’s more than 120 Canadian clients.
Cantech Letter recently asked Kevin to break down the history for us, and which aspects of working with Big Data in a credit rating agency context are applicable to other verticals.
Kevin Deveau: Yeah, FICO has been around since 1956. When you say “credit rating agency”, what we really do, is we’ve developed since 1956, the FICO credit score. That is a predictive analytic algorithm that determines, based on your risk patterns and your use of credit, what your credit risk is. So we work very closely with the credit bureaus. In Canada, that would be the Equifaxes and the TransUnions, who have a lot of data but they will consume and leverage our FICO score, and that’s what they use to provide to the banks and all that, what they use in different credit decision-making. So 1956, and we’ve been doing predictive analytics way back, starting at that time. Big Data, yeah, you’re right, the last year or two it’s been a big buzzword, just like we’ve had “CRM” as a buzzword in the past, but what we feel is that it’s the availability of tools now and the low cost of processing time that allows people now to really work with it. They’ve always had the data. You can talk to many customers or clients who’ve had so much data, but they’ve never really been able to use it or to manipulate it or to analyze it. But now that the technologies are there, you can really get a return on it. You can analyze the data, you can take action upon it, you can take action in a real-time environment. So we feel we have some very good tools and capabilities in that space. We feel that we’re one of the number-one people around that can really manipulate and provide solutions in the Big Data arena.
They’ve always had the data. But they’ve never really been able to use it or to manipulate it or to analyze it. But now that the technologies are there, you can really get a return on it. You can analyze the data, you can take action upon it, you can take action in a real-time environment.
Yeah, obviously, it’s a question that’s all of a sudden become a very pressing concern for almost everyone now, especially around questions like predictive analytics and, basically, insights. You have a vast amount of data. How do you act upon it? FICO has also made inroads into cloud software, relatively recently, too, hasn’t it?
Correct. As we’ve seen the market and we’ve seen the popularity of Big Data, about two years ago we decided that we needed to have a cloud version, to be able to go after and support customers that were in the smaller-to-midsize, who didn’t have the skill sets. Or even doing work with big organizations, but they have a specific line of business or a specific department that needs to do something, but they can’t wait and rely on their existing IT department, because they’re just too long and too slow. So we’ve built the FICO Analytic Cloud, and within that Analytic Cloud we’ve taken all of our capabilities and tools and provide them in the cloud. So anything with regard to building business rules, to determine how you act upon specific results from a predictive analytic tool. You can build scorecards, decision trees, you can build predictive models. You can also optimize these decisions. And also, we’ve got tools that allow you to go in and look at the data, and connect the data, and then be able to build the models and act upon it. So we have, I don’t know, a couple hundred different patents out there in regard to predictive analytics. Building the FICO Analytic Cloud, we’re bringing a lot of data, but we’ve opened it up to many other companies or toolsets. We’re trying to make it as open as possible, so that the end customer, the client, has everything they need at their disposal to build solid Big Data analytics.
How long has FICO had a presence in Canada? What are some examples of FICO applications?
We’ve been in Canada now for about 20 years. We have offices in Toronto and Montreal. We work across many different verticals. There’s financial services, banking, retail, healthcare and government. Everyone who uses a bank uses us in some shape or form, whether they’re using the FICO scores in their credit decision-making, or we have a number of pre-built applications, specifically for the banking sector around different customer lifecycles. So whether it’s originations, you want to originate a credit card or a mortgage, whether it’s fraud detection. So using your credit card, we have predictive consortium models that the banks are able to use to determine whether you’re using your credit card outside of the regular pattern or usage. So we can basically stop the transaction in real time. And recently as well, we’ve done a lot of work in the retail space, really doing one-to-one marketing. So taking lots of data, if you think of many of the big retailers, and determining what is the optimal offer to provide to the end consumer. We have a number of different predictive tools, as well, even down to the point that we can determine 1) if a consumer wants a new LCD TV, but also when he’ll buy it and what do you think he’s going to be willing to pay? Those are timed-to-event analytic patterns that we use, that are proving to be very, very successful. And we leverage those, not just across retail, but we’re also doing a lot of this in the financial services, where what is the optimal offer when you think of your banking products, increasing wallet share.
Yeah, it’s something we hear over and over again, that the secret spot for retail is to have the right product in front of the right customer at the right time, and this talk of mobile, real-time, one-to-one marketing. It’s becoming kind of a crowded space, with large and smaller players coming up. And it’s driven by a few different sectors, like loyalty programs and stuff like that. You mentioned that you have a patent portfolio, which strikes me as a differentiator.
When I say patents, it’s a number of patents around specific analytic algorithms or ways of doing business, a way of calculating stuff. So when you talk about a crowded space, we feel that we’re in that space, but we’re one of the leaders because we still haven’t found a competitor that can do the true one-to-one offer, one-to-one marketing. We see a lot of our competitors that do more the segment, if you want, type marketing. Or they can say they’re going to sell shaving to men over 55, whereas we can get down to the point that Kevin Deveau is losing his hair, or he has a beard, so he doesn’t need to shave. We don’t need to show him shaving cream advertisements. And also many of our recent acquisitions are around the customer connect and mobility. We’re able now to take all of the data through our different engines, our offer engines, and then once the offers have been determined, we have tools now that we can disburse these offers, whether it’s via mobile on a smartphone, or whether it’s driving the consumer to the retailer’s website, to get coupons or whatever. So we really feel that we’re one of the leaders, and we have the full breadth of solutions, to take you right back from taking your data, doing the offers, and distributing those offers in a timely manner with the device of choice and the method of choice of the end consumer.
Yeah, that’s really kind of the Holy Grail. I just keep hearing it over and over again, about the importance of mobile and real-time, one-to-one marketing. Canadian Tire turned to FICO when they were trying to set up an internal credit card, to power the back end of that. Canadian Tire is an interesting story in the Canadian retail system because we see a lot of legacy retailers going down, like Sears and Future Shop, while Canadian Tire has really kind of differentiated themselves in terms of innovation around that. What do you think led them to choose FICO? What did they see in you?
We do a lot of work with their financial services branch, as you said around the credit card. We have a number of optimization tools, we call them. So if you have a credit card, what is the optimal credit line increase to provide to each customer? It can be a very complicated mathematical problem, based on your purchase patterns, you history, whether you pay or not. You’ve got to key in on what you provide. So leveraging all that data and working with the Canadian Tire financial services team, they leveraged our optimization software to be able to provide the optimal credit line increase per each individual. And that has been very successful. And our model with our customers, and specifically with the great story of Canadian Tire, is that the first project we worked, here’s our software, we worked hand in hand with them 50/50, and our eventual goal is to have them be fully self-sufficient on our software, and then be able to use the software on other decision areas within the credit card space. We talked about credit line increase. Well, there are other decision areas, as we call them. What about initial line assignments, when you open a brand new credit card, what is the optimal initial credit line you should provide the customer? So there’s a lot of mathematical crunching of numbers, and there’s various simulations that need to be done. Canadian Tire has really taken that on full-bore and doing various types of decision areas using our software and are very successful at it.
We have, I don’t know, a couple hundred different patents out there in regard to predictive analytics. Building the FICO Analytic Cloud, we’re bringing a lot of data, but we’ve opened it up to many other companies or toolsets. We’re trying to make it as open as possible.
Yeah, it strikes me, as far as retail is concerned, that the attitude of legacy companies, or even small-to-medium size businesses, who simply just don’t know how to deal with the new reality of what retail is becoming is going to spell the difference between who gets left behind and who survives.
Definitely. And that’s why, with the launch of our FICO Analytic Cloud, these tools are going to be easily and readily available to small and midsize retailers, at a cost point that meets their objectives and meets their bottom line. They don’t have to go buy big, honking machinery or software to install on-premise. That’s all going to be done by FICO.
I think that’s the other issue around Big Data and how to adapt the idea that everyone now needs to be hiring a full-time, in-house data scientist or something. A company with the resources, obviously, of Canadian Tire, they’ve installed a big data processing centre in Winnipeg recently. But for small to medium-size businesses, it’s just not practical. And yet they need some kind of means to be able to solve various specific problems that they have. So it’s, I guess, encouraging to know that they will have access to something that’s appropriate to their scale and to their specific problems.
Yeah, and that’s our feeling, as well. We offer the Analytic Cloud, and it’s offered with a cloud option or on-premise. As you can imagine, some organizations are very particular about letting their data out of their four walls, so they want all their tools and expertise on-premise. And that’s one of our core offerings, as well.
Aside from retail, what are some other fields that FICO has been offering its services for?
In the government space, we do a lot of work with some of the crown corps. One specifically is Business Development Bank of Canada, where they provide commercial lending to Canadian entrepreneurs. So all the aspects around the originations and getting the loan, doing the risk analysis, analyzing the data, all of that is being done with tools provided by FICO. Other areas that we’re working with is around the airline industry. The recent requirements where airline pilots or flight crews can only fly a certain number of hours, and if they do they have to be grounded or take a rest, all that math and complexity and travel, that’s very difficult to predict with an Excel spreadsheet, so we’ve done work with a number of the major airline carriers in the world, to help them optimally plan, so that they’re not losing time by having to wait overnight in a specific airport and all that good stuff.
Yeah, it’s amazing, when you begin to think about it, that there’s not really an aspect of life that isn’t going to be affected, or can’t somehow be either improved slightly or just changed wholesale, not just figuring out how to harness vast amounts of data but how to apply the insights, to be able to extract meaningful, actionable information out of that data. That’s going to be a real skill to develop for a lot of people who aren’t used to it, in the near future.
I totally agree with you. Like, the processing power for hardware now, it’s not that expensive. And the offer for FICO Analytic Cloud, it just lets the small to midsize players, that could never have envisioned this maybe five years ago, now can get in the game, and can really demonstrate themselves, or market themselves, like the big guys in whatever space they’re in. So you don’t necessarily have to be a big, big player or a big retailer to benefit from these new Big Data capabilities.