For Subscribers

Get the complete article now!

TWST: Shall we start off with an introduction to Retail Decisions (London: RTD.L)?

Mr. Clump: Let me tell you a little about what Retail Decisions does. The Company works in the card payment space. Our specialism is all about leading the fight against payment card fraud. We make sure that payment card transactions are valid on behalf of the banking community, retailers, including ecommerce merchants, oil companies, telcos. We have a range of services, from the very simple checking of card data all the way through to very sophisticated services, which include artificial intelligence, which is about as complex as one can get. Additionally, in Australia we actually have our own payment card business whereby we issue our own cards; they are specialized purchasing cards for the fleet industry. And this actually gives us tremendous insight into the entire payments process. We were founded on January 12th, 2000 as a result of a de-merger from our parent organization. The reason for doing that was to allow us to concentrate on this payment space because our previous parent was actually a diverse portfolio of activities, not all of which were involved with payments. One of the key reasons for wanting to concentrate on the payment space at that point was because we had a skill set that could be transferred to what's known as the card-not-present market. By the card-not-present market, I mean the retailing environment where the retailer never actually sees the card holder or the payment card initiating the transaction. This is a specialist retailing environment and our skills were absolutely appropriate for it.

TWST: Talk a little more about your technology, if you would. You mentioned sophisticated services and the complexity of artificial intelligence. Can you elaborate and give us a sense of what you are capable of doing in the fraud prevention space?

Mr. Clump: We have got some very clever neural-technology, which was developed by academics from Brown University. This neural-technology, this artificial intelligence, thinks like the human brain; it thinks based on experience. So what we have to do is actually feed through the engine, a history of good transactions and a handful of bad transactions and then we map those transactions on a multidimensional array. This multidimensional array is constructed on the basis of which dimensions we think actually give rise to fraud. So, it might very well be the size of the transaction, the value of the transaction, goods purchased, time of day, etc,. And we end up with an array which is not two or three or four dimensional, but perhaps 20 to 40 dimensional. We plot all the good transactions on it and we plot the bad transactions. Then, the next time we see a transaction coming through that too is plotted on this matrix and its proximity to a previously known fraudulent transaction gives it a score. If that score is sufficiently high the transaction is put into an out sort for further interrogation and investigation.

TWST: How would you describe the current condition of your markets and what’s you general outlook for the next 6 to 12 months?

Mr. Clump: I think the best way of describing it is that currently we are cautiously confident about our markets. Having said that, we wouldn’t want to be too aggressive because at this time last year we were quite positive about our markets. But I have to say, towards the end of the second quarter, prospects had evaporated; people had stopped spending. So, I would hate that to happen again. But we have a view that customers who are not spending money on fraud prevention and risk management are simply creating a pent-up demand and we certainly hope that is the case.

TWST: What are you hearing from customers at this juncture? What are their needs and how do you meet them?

Mr. Clump: Well, over the last 3 years we have been quietly assembling a range of services to offer our clients, because customers don’t always have the same fraud payment issues. Therefore, it is important for us to have a range of services so we can offer what the customer needs for that particular circumstance. In the fraud space, obviously we are offering them a reduction in their fraud levels, but we start off from the perspective of a retailer, I would say, meaning we try and maximize their revenues and good transactions whilst reducing fraud. It can be very easy to eliminate fraud totally by having draconian rules in place, but that’s not the name of the game. The name of the game is actually to maximize revenues and good transactions for our merchants. The other thing we do is to actually automate those processes for our customers. So I think there are a lot of the successful e-merchants around who are keeping their fraud relatively low, but they have large departments doing exactly that. We automate the process so they can reduce the cost of those departments. It’s typically not a cost that they should be carrying or can afford to carry. We also minimize what's known as the insult rate, i.e. turning away a lot of good customers. We rarely do that. Whereas if it is a merchant who is putting in place his own rules, his own schemes, he probably creates rules that aren't necessarily granular enough. He may, for example, ban transactions from a particular geography altogether and we don’t think that’s really the best way to proceed.