It is often said that data is the new oil. But although data is everywhere, most of it is often stored away, completely untapped by businesses. So, how to make the most of it? Mick Lévy, Business Innovation Director (Business & Decision), answered to Isabelle Barth during an interview she conducted for Xerfi Canal.
Isabelle Barth: Hello Mick! You wrote a book which, though targeting specialists, is quite accessible to all: “Sortez vos données du frigo – Pour une entreprise performante avec la Data et l’IA” (Get your data out of the fridge – The effective use of data and AI by companies). So I will ask you for some advice! Where is this data and why is it stored in “fridges” by companies?
Mick Levy: Hello! First, I would like to share a very meaningful statistic: only 32% of data is exploited by companies. Here is a raw material that all companies possess but that is severely underused.
Every business gathers data on its customers, processes, suppliers, products, distribution methods… This unique data is only known by them. And the effective exploitation of this data is a must as the return on investment directly generated can be very high. When we talk about the “21st century oil” well-known expression, it comes with certain economic implications.
Isabelle Barth: So, we exploit the data to heat it up so to speak, since if we stick to the fridge analogy, that data has gone cold…How do we do that?
The more we live, the more we produce data. What we have therefore is an infinite resource that virtually anybody can use.Mick Lévy
Mick Levy: First of all, the fridge represents those company information databases in which a lot of mostly inert information is stored.
Isabelle Barth: When we talk about “data”, we can be referring to a customer’s file or even simpler things in reality?
Mick Levy: It is a broad term and yes, it can be a customer file. There is also a growing interest in what is known as unstructured data, text data, even images. For instance, production chain images can generate significant value. Coupled with AI, images can help to detect automatically any manufacturing defects in products. So the quality controls can be optimised and automated for even more efficiency.
The set of data being considered is thus extremely large. It is fascinating because it is a very dynamic resource which can be supplemented with data from outside the company obtained from partners, open data (open data directly accessible on the internet) or even new devices created for the sole purpose of generating new data. It is the case for connected objects with IoT (Internet of Things) sensors generating huge volumes of data (if ever you needed more! 😉).
We will not run out of it anytime soon as, unlike oil, it is not a finite resource, quite the contrary! The more we live, the more we produce data. What we have therefore is an infinite resource that virtually anybody can use.
Isabelle Barth: Let us take a simple example. Imagine a small shop now forced to offer a Click & Collect service, although it has not really managed to put together a customer file over the years. Are there any initiatives that can help this shop extract data to create a list of prospects?
Mick Levy: Yes! This small shop would be well advised to join forces with others, within what are called marketplaces. The marketplace then provides alternative services to all of the shops.
Whenever Data is involved, we tend to think only of big companies. When in reality, even very small ones are concerned. Let me give you a concrete example from a colleague’s experience.
The company concerned sells town hall items, such as medals bestowed during celebrations, civil naming ceremonies, weddings … In short, all the “goodies” given as gifts by communities.
You may be wondering: what does that have to do with Data? Well, a very simple application was set up using Open Data which provided the company with forecasts regarding the number of births and marriages that would occur in each French town. As a result, its production planning was improved, it gained better control over its business and now knows which type of product to offer to what town, etc. Using data that it initially did not possess, it was able to create new services, grow and improve its overall efficiency.
The need to put an end to the silo tragedy
Isabelle Barth: In very large companies, we also have the famous silos! With, as a result, bits and pieces of files scattered here and there and information about one same contact person or product located in various places in the company. This requires a different approach…
Mick Levy: The silo situation is well and truly a business tragedy! The way in which it manifests itself varies according to sector and company history: there can be organisational silos, product silos, sales channel silos … with each their own IS structure. This is why there can be records regarding a customer in 3-4-5 different information systems…if not 20 (I have seen it before!).
To obtain a 360° customer view and to optimise customer experience, it is imperative that you show the customer that you know him and are paying attention to what he says. All these 20 different IS must therefore be reconciled. Which is a real challenge, including in terms of data quality and data cross-referencing to obtain a coherent view of the customer and create the best services.
The organisation, the development, the company history, and the way in which the company builds itself all contribute to the silo tragedy.
Data is an asset!
Isabelle Barth: In concrete terms, what are the first steps to taking data out of the fridge?
Mick Levy: The very 1st thing you have to do is wake up to the reality. Data is an asset! Recent years have seen companies try to optimise everything: means of production, processes, logistics, human resources (that have been under a lot of pressure). And yet many of them have totally overlooked a true pot of gold!
2nd advice: launch the first use cases rapidly. You need to adopt a precise, practical approach and make your first attempts at creating value from data.
And 3rd advice: organise the company in such a way that data is treated as an asset in its own right. You need to set up a data exploitation-oriented structure and define dedicated processes. This is what is known as data governance. Then, you can create an information system by acquiring tools and technological means that will help you process it.
Isabelle Barth: Of course, this implies having asked ourselves beforehand what the desired outcome is. The idea is therefore to fill the car’s fuel tank and to know in which direction you want to move?
Mick Levy: Exactly! You must never lose sight of the uses, or even the return on investment generated by these uses or the creation of new services. It is in fact imperative that you define indicators that will allow you to guide the definition and execution of the company’s data and AI strategy.
Isabelle Barth: As researchers say, data is never given, it is constructed! It seems we have reached the same conclusion 😊 Thank you Mick!
Mick Levy: Thank you!