|© 2016 Black Swan Telecom Journal||•||protecting and growing a robust communications business||• a service of|
|Email a colleague|
People who live in countries where ATM machines and widespread credit cards abound will be surprised to learn that more than half the world’s adult population is “unbanked” — or underserved when it comes to financial services.
That’s the conclusion of a McKinsey study that provides a country by country assessment of the issue. While a significant portion of the population is unbanked in developed countries, the biggest number of the unbanked, by far, live in the regions of Asia, Africa, and Latin America.
Mobile money is crucial to enabling those regions to expand, increase their GDP, and boost their financial strength. And widespread acceptance of mobile money transfers and payments has already proven key in growing small-to mid-sized businesses in developing countries.
One company who has a major presence in the developing world is telecom fraud and revenue assurance vendor, Neural Technologies. So we invited Deputy CEO and Chief Commercial Officer, Luke Taylor, to join us to discuss the fraud and credit issues he’s seeing in unbanked regions. He also gives us a picture of revenue protection solution challenges and opportunities across the board.
|Dan Baker: Luke, I understand Neural Tech is having its 25th birthday. Congratulations on that. And I’m curious how the fraud threat has changed across those two and a half decades?|
Luke Taylor: Thanks, Dan. We feel lucky we landed in telecom because the industry’s pace and evolution have been dramatic and non-stop. Fraud management may not be the greatest growth market, but there will always be a need for it. Someone will always be there to take advantage of the system to rip people off and steal money.
It’s interesting to note that some of the fraud threats we saw 25 years ago are still with us today. For example, PBX fraud was a major problem in the analog network days, but now it’s come back because the IP-PBX is a very powerful call router if a fraudster can hijack it.
There’s a trade off here. As new technologies penetrate our everyday lives, we leverage greater intelligence at the cost of greater fraud risk. For instance, machine-to-machine conversations with your automobile will fix your location and provide useful data for the driver — but they also open up opportunities for abuse.
|One of your biggest success stories is in Kenya where you are serving the fraud protection needs of the M-Pesa mobile money service.|
Yes, M-Pesa is a service provided by our client SafariCom. It’s a very interesting money transfer service that allows workers in the cities to transfer money back home at weekends to their families in the countryside.
It’s big. More money is actually transferred through M-Pesa than through one of the national banks in Kenya. The service works through SMS — almost a requirement because people in the countryside often don‘t have reliable electric power. 3G networks require you to charge your phone pretty often, so this is one case where a more basic phone has its advantages.
We also work in Indonesia with a similar demographic to Kenya. Indonesia is a huge country with a population of 250 million — that’s 80% the size of the U.S. — and many areas are so remote they don‘t have a bank or ATM machine in the community. Instead they have a small food store where families go and connect via a mobile phone.
Our experience in these markets has us eyeing the huge potential for fraud protection services for the unbanked or financially underserved people of the world.
|One of the skill sets you bring to the telecom fraud solution market is your fraud experience in areas like banking. How does that play into your portfolio?|
Dan, as telecoms become more of a services business and value added services like mobile money remittance and payment becomes a bigger percent of the revenue pie, our cross-industry experience becomes more valuable.
Way back when we first entered telecoms, fraud management in telcos was new and immature. Yet you even see that immaturity today in emerging countries where they have little appreciation for credit checks and whether or not a subscriber is likely to generate revenue for them.
By the way, there are lots of good reasons for banks and telcos to get together and share intelligence. In some cases, we’ve tried to act as a middleman to bring our clients together, but it’s tough to do since both sides are wary that the other will steal its customers.
However the mobile phone is a natural for checking the location of the user. So if the cell phone is near the retail location where a credit card purchase is being made, that’s valuable intelligence to prevent fraud.
Here again, there’s a big need for credit check solutions in the unbanked world. You have several credit bureaus in the U.S. But many countries have no credit agencies at all. However, the mobile phone enables the storage of a history of purchases and that activity can be used as a form of credit check. The financial institutions could make good use of that intelligence.
|Let’s switch gears and talk about the developed world where there’s also plenty of fraud occurring. One of the unpleasant aspects of this is that banks have become super cautious. I use a national bank and they expect me to call and advise them of my travel plans if I roam outside my local area.|
Sure, you may be protected, but as a consumer you are inconvenienced.
We are seeing much the same thing in the telco market. In fact, we’ve won business due to a personal experience of the CEO whose spouse was being barred from making phone calls when she exceeded $300 in credit. And it made no sense because the CEO was earning millions. Dynamic credit limits that align to the individual consumer means increased revenue and less dissatisfied consumers.
The key to more accurate fraud assessments is to apply rules in combination.
For instance, if I withdraw $100 from an ATM when normally I only take out $50, that raises a flag. Another flag is raised when the user is visiting a foreign country. Still another alert is when the user is shopping on-line instead of at a retail shop.
So all those overlapping flags are analyzed to ask: is this a likely case of fraud?
It is a fine line between a profile of a very good customer and a fraudster, they look very similar. If you can identify the nuances, the opportunity to minimize revenue loss and ensure customer satisfaction is achieved.
|In a complex, multi-layered algorithm like that, how does your client set the parameters? How much can they actually set themselves?|
They can not only set the limits or thresholds of the rules, they can add their own rules too. They can even deploy neural models — collect data from various sources to build up a neural profile to complement other analysis features..
To be honest, in our early days we didn‘t fully appreciate the virtue of self-configuration. To us, it was just an expedient for getting business. We wanted everyone to do their own configuration so we could go out and sell product.
But it gave us the capability to say, “Look Mr. Customer, you can do anything you want. We can do it for you, but you also have the capability — you are totally self-sufficient.”
T-Mobile in the U.S., for example, is quite happy to be doing their own configuration. Then we have other companies in the Middle East, Africa, etc., who rely on us fully to do everything due to their resource limitations
In general, though, telecom is a dynamic industry so relying on a vendor can slow you down. If you are desperate for a change, we may not have anyone to help you immediately. But if you have the power to make the changes yourself, there’s no time lag and no purchase order to write.
|Luke, thanks for the interesting discussion. On the new solutions front, where are we likely to see Neural Tech launch something?|
Well, it’s no secret that “big data” is a potent buzzword these days. And we’ve been managing big data for a long time, but we are looking for how we can exploit that capability not just for revenue protection, but also in revenue maximization.
Clearly the data we already process can provide valuable intelligence for marketing and tell them how people are using their phones. Marketing also uses several disparate sources of data, so consolidating those data sources would deliver a nice value add.
When we deploy our revenue assurance, credit check, and fraud solutions together today, they are on the same hardware platform and databases. And all departments can potentially mine that data — or you can create a partitioned system that multiple operators or retailers can share where each one’s customer base and data is protected.
In South Africa, for example, MTN uses our system and several of the service providers selling MTN services are on the system as well. Now because MTN doesn‘t own those customers, the system is partitioned.
However, MTN can still look at the aggregated and anonymized data holistically for their own analysis. And this allows them to see, for instance, a fraudster moving from one service provider to another — quite valuable.
So our research and client feedback will certainly answer the question of where Neural Tech goes next with big data.
Copyright 2015 Black Swan Telecom Journal