Interview of Dominique Orban de Xivry, Rever Data Engineers’ founder, by Le Cercle de Wallonie.

Dominique Orban de Xivry :
Big Data is the black gold of the 21st and 22nd centuries. It is a possibility for enterprises to finally be able to better target their customers and to distinguish them by offering more punchy service propositions.

CWNetWork :
In each of its daily actions, a company brews an impressive amount of data : socio-demographics, consumption habits, preferences. Dominique Orban de Xivry explains why this information is a goldmine for business .

Big Data is a form of interrogation or use of all the data available today, whether on the Internet, either internally or externally, and in all their varieties, it can be text and sound, images or digital data, etc.

Today Big Data is mainly used for prevention, analysis to understand the phenomena that we could not identify other than on large volumes. Typically if you take 1% in 1,000 data, it is only 10 data and it is not very significant, however if you have 100 million records and you have 10 million data that are in the same situation, you can then legitimately ask good questions. So that’s basically it, the first phase is called understanding.

The second phase of the phenomenon arrives later, once I understood, I can prevent and therefore use this kind of metadata phenomena to do prevention, particularly in industrial maintenance or in the marketing field because if I know for example that 10 million consumers bought such perfume, they are in that situation and they are candidates to buy this perfume. From the moment when companies have mastered this information or have access to that information, they can do what is called differentiated marketing, firstly in order to better target their customers and also to make differentiated proposals, so they can be more efficient and impactful over the said target.

Typically, to take a simple example, today if you sign in Facebook saying “I love spicy food”, you are part of a network of people who like spicy food, let’s say 10,000 people worldwide, if you re-cut it with the people who live in Belgium, you have a sample of 1,000 people and within that, if you are a spice merchant, you may well say, “I want to make a campaign on pepper” and within that population you will be able to identify people who love pepper, and thus target them, let’s say there are 2 people in Namur and 3 in Verviers and from there, you know who you are addressing to and you will be able to make a specific promotion for pepper. So that’s what it is, it is gold bar for companies to better know what the needs of their customers are and to be led to make proposisitons, etc.

These are the type of techniques that are already used by Amazon and other major retailers that offer you what they think or what they have deduced that you like, because you visited their sites.

There are a number of issues that have arisen through this, first is the most obvious : confidentiality of data or rather the privacy and protection of privacy, that is to say, just because I said I liked pepper, it doesn’t mean that it can become public and a selling or buying factor, I want this information to remain privileged and not be disclosed to “marketers” who, to me, are not concerned. When it’s about pepper, it’s not too serious, but when it comes to health problems for instance, it can be much more serious and therefore there are a number of precautions to take.

The European Union takes care of that since the intend today is to put some regulations in place to avoid such excesses. But this is the most negative aspect, it’s the legal aspect, there are of course other more diffuse and complex aspects which are actually how to stop these machines because they could go to the end of some logic and lead to a number of disasters. Today we do not see this coming yet, so let’s remain optimistic, the men who control the machines may be able one day to continue to control them.

Today there is a Belgian company that uses Big Data because they realized that its customers are leaving and that indeed the loss of customers was often linked to networks or to some phenomenon of people who were connected through a network, or because they lived in the same neighborhood or because they share affinities, and when there is one in this network who leaves a supplier, often by contamination we see that all the people in the network leave quickly after. And so today this supplier checks every three months and sees who are the people who leave and then they’re able to warn their sales people who try to prevent these departures by offering the customers a better proposal.

Another example that is quite interesting, they are companies that scan through all that is published on the Internet, about the behavior of their competitors, to try to understand with whom and why their accounts are cheaper than theirs. In particular they exploit a number of informations that are public and are published by journalists in order to know the companies that are provided by said competitors and to know how these companies manage to make things work.