Data: the 21st century’s new black gold? It is, according to today’s successful Internet giants. But what challenges does Big Data pose to commercial players in traditional sectors? How can a company maximise the value of its data? Laurent Herr, Data Director for Orange Applications for Business, and Mick Lévy, Business Innovation Director at Business & Decision, decipher this technological and economic revolution.
What are the main challenges today around data?
Mick Lévy: In recent years there has been a general awareness of the business value of data. It’s now seen as a lever for development. For a long time, the world’s largest company produced oil, and now they exploit data! The GAFA, natively data-centric, are using this power to “uberise” certain activities and benefit from extremely rapid development. All companies can take inspiration from this model and arm themselves with the right resources.
Laurent Herr: The potential of data affects all companies, in all sectors and also all functions. It influences sales – Amazon achieves 30% of its revenue through its recommendation engine which is based on data analysis – as much as it optimised costs, thanks to predictive maintenance or autonomous robots in distribution warehouses. Data also changes business models, such as in the case of driverless and connected cars.
Companies need to consider their data as a real asset! And, like any asset, it must be valued, and allocated human, organisational and technological resources to gain maximum value.
How does data influence decision-making?
M.L.: Nowadays we know how to process data instantly, so decision-making is accelerated. With artificial intelligence (AI), some decisions can even be taken completely automatically from analysing the context of the decision to acting on it. For strategic decisions involving top management, Big Data provides objective insights by cross-linking internal and external information for the company. It’s very powerful because it enables a deeper understanding of the company’s circumstances and in the right context.
L.H.: Take Google for example. All of their processes are driven by data. Engineers modify the software daily using A/B Testing methods. Depending on user feedback, the solution is deployed: more widely if users are satisfied with the process, or modified if they are not. This new type of process is enabled by mass data.
How do you organise a smooth flow of data within a company?
M.L.: It’s a major challenge for organisations! If you want to maximise the full value of data, you have to make it more fluid, more accessible and more exploitable: creating a maximum number of collection points, of self-service professional tools, of systems that enable data to be exploited and centralised smoothly: Big Data, Data Lake, Data Hub, etc.
And you have to balance this freedom of data with strong control over the way it’s used. Data governance is necessary. Data scientists, Data analysts, Data stewards and Chief Data Officers (new CDOs) sit at the heart of these challenges.
L.H.: Until now companies have been usually organised around processes (sales, logistics, HR etc), which creates data silos. A new type of model is emerging in which the data bearers unify the data and organise processes around it. It will radically change the way a business operates.
M.L.: Yes, the challenge is to break the silos that ultimately constrain organisations. We are fighting against what has now been coined the “data silo tragedy”..
Optimal use of data involves moving from a ‘process centric’ to a ‘data centric’ approach, in which processes are organised around data and not the other way around.
What role will artificial intelligence play when it comes to exploiting customer data?
L.H.: On the IT side, AI will lead us to review how information systems are built. It’s about enabling them to learn algorithms by giving them a large amount of data and requesting a certain deliverable. The rules based on the type of “if then else” coding are no longer valid; machines learn behaviours based on statistics. The more data you can input, the more the algorithm learns. These new systems are very powerful. In just a few days of learning, Google’s “AlphaGo” software was able to beat the world-class Go players.
M.L.: We’re at the beginning of a revolution that will affect absolutely all professions. With very practical and easily activated applications today: chatbots, automated customer support services based on data science… For example, we’ve developed predictive models for our customers to reduce absenteeism and accidentology, or predict the sales turnover in a retail chain’s 1,200 points of sale. Even if AI is just at the start, it’s already seeing fast ROI. .
The public are still quite concerned around the use of their data. How do you reconcile data and ethics in a business context?
L.H.: Generalising data exploitation requires individual learning because it’s necessary to carefully control uses and encourage best practices. You need to be transparent with customers about how their data is stored and give them the ability to modify or delete it. As an operator, our customer promise is orientated towards cautious data use: we want to lead the field in this area. For our Flux Vision offers for example, our algorithms are designed to guarantee complete customer anonymity.
M.L.: This collective awareness has brought about the General Data Protection Regulation (GDPR), which comes into effect on 25 May 2018. This founding regulation is historic! It sets out two essential principles: the company is legally responsible for the data it collects, stores or processes, and it must take into account data protection requirements right from the initial product design (privacy-by-design).
L.H.: GDPR shouldn’t halt development projects. We can extract the value out of data while restricting the practices related to its exploitation. The key concept around data management and value creation is trust. GDPR provides a framework to inspire and strengthen it.
What synergies coming from the Orange and Business & Decision merger will benefit the way you manage your customers’ data?
L.H.: This merger will enable us to accelerate our data-driven development with cross-functional teams, benefiting from 25 years’ experience from Business & Decision. Their “service” approach is also very complementary to our “product” approach.
M.L.: We have a common data vision. The synergies resulting from this merger will enable us to combine our services so we’re present at all stages of the data journey: collection, transport, security, storage, analysis and sharing. We’ll be able to support our customers throughout the entire value chain.