Technical hints: here's just some words about the operating principle of ZonabitPremi, according to confidentiality with which you must deal with this matter. Those who wish more information can write us, better adding a telephone number.

Cards loaded incorrectly can be found looking for < cousin cards >: there are families of cards that typically are loaded together, because a person keeps them in his pocket and uses them one after the other when he gets the chance. First of all ZonabitPremi analyzes your data and finds the families of < cousin cards >.

But this is not enough. There are three types of users of promotional cards that you can find usign ZonabitPREMI and the method of < cousin cards >:

1) Frequent and habitual customers who usually meet at the same time on the same commercial shop. For example at points of sale of fuels, this phenomenon occurs typically in the time slot between 7 and 8 am, and involves professional drivers or people with high mileage for work.

2) Frequent customers with high payments that require loading points on cards out of regulation.

Policy towards these users shold be diversified: in the case (1) customers should be completely relieved from the control and suspicion, in case the (2) is correct that the collection of points is controlled discretely to tolerate partially the phenomenon, while in the case (3) any collection of points should be forbidden.

Simple application of the method of the < cousin cards > can not tell apart behaviours (1), (2) or (3). So the question is: what to do to discover only real frauds?

The answer is simple: employees load their cards when nobody notices them, while real customers of type (1) and (2) do not, because they have no reason to do it. Therefore, you need to evidence the behavior of those who load their own cards when there is little or no traffic in commercial shops, it is when employees have time to work for themselves.

Yes, the answer is simple, but the solution is not at all: how to find those who act in this way? Zonabit knows how to find them. ZonabitPremi uses some parameters in processing which measure the frequency of loading by employees when they are alone. So ZonabitPremi can establish a general hierarchy of the propensity to fraud
Obviously, this assumes the ability to read and process in real time a huge amount of transactions. Hence arises the need for efficient Big Data processing algorithms, but also for the ability to recognize suspicious cards through machine learning algorithms.
Now we can not tell you more: write us, and we'll discuss with you confidentially as the matter requires

FRAUD PREVENTION IN PROMOTIONAL CAMPAIGNS

ZonabitPREMI: the easy solution for a complex problem