In an interview just after the Data Transformation Conference, organized jointly by Netexplo and two HEC Paris Centers: Hi! Paris and Innovation & Entrepreneurship, Julien Lévy, Associate Professor and Academic Director of the Innovation & Entrepreneurship Center, shared his vision of Data Transformation, a vast topic and a new step in the digital revolution.
Can you define Data Transformation and clarify which companies are the most concerned by this transformation?
Just like Digital Transformation, “Data Transformation” is a fashionable term in companies right now, but its meaning remains ambiguous. I would say that it basically consists in innovating the way we do business, by leveraging data to boost productivity, develop business and create value for customers. In other words, data transformation is an organizational innovation policy, and the tool of this policy is data and its processing.
In fact, the expression “transform everything into data to transform everything via data” is, in my view, the best definition of the digital revolution and its implementation. This is reflected in our current practices because we are digitalizing everything, everywhere and all the time. Innovations such as connected objects, voice and image recognition with AI, and the power of networks and the cloud have literally led to an explosion in the amount of data produced and readily available.
Transform everything into data to transform everything via data
In my opinion, companies are the first to be concerned by data transformation, because they have or can have access to a mass of information on their activity and market. There is no longer any point in distinguishing between digital and non-digital companies; the reorganization of business around data concerns all companies, regardless of their sector. If we had to make a distinction, it would be between “digital natives” on one hand and “digital learners” on the other. The former “speak” data naturally, while for the others, it is a learning process.
What does a company need to launch its data transformation?
With my colleague, Jean-Rémi Gratadour, Executive Director Program Development of the Innovation & Entrepreneurship Center at HEC Paris, we carried out three surveys over four years on corporate data policies. We were amazed by the acceleration of this transition: what was just a topic for discussion four years ago has in some cases become a strategic priority. I would therefore say that there are three steps to this transformation.
The first is data processing in silos. All information systems imply data creation, and this data is collected and processed each time for a specific use, such as human resources, logistics or management control.
The second step is the strategic aim of becoming a “data-enabled company”, in other words, supported by data. Here, the goal is to improve the productivity of the company or its departments by using data in a different way.
The third step is the aim of becoming a “data-driven company”, which means that data pervades all the company’s activities. Added value is thus created increasingly through data processing, which in turn leverages organizational agility. It must be pointed out, however, that this is not a mandatory step for all companies. It is more of a strategic choice than a requirement. Moreover, claiming to be “data driven” when you are not a native data company is a challenge, it’s a gamble that can have numerous implications. However, in my opinion being “data-enabled” is now essential.
In both cases, whether companies are “enabled” or “driven” by data, they cannot shift from data silos without the strategic backing the board of management.
What are the main challenges facing companies in terms of data transformation?
Besides the company’s vision and strategic ambition, data transformation raises three key issues: technological governance, data governance and project governance.
Technological governance is the foundation; you cannot have data transformation without an agile information system. This takes the form of a data lake (or data hub), API and tools for processing and visualization with standard interfaces, where everything goes through the cloud. In some cases, the technology works alongside the company in a gradual process; in other cases, highly-structured strategic choices are made and imposed with regard to technology, as is always the case for data-driven companies.
Data governance is key: as soon as a company acquires powerful tools, it becomes more aware of its powerlessness, due to a lack of extended and reliable data. The key issues are data quality, which is very complicated to guarantee, and compliance with the regulations.
Lastly, project governance raises a new challenge. In order to work more closely with business and form part of an innovation drive, the profusion of projects and experiments (POC: Proof of Concept) is the best option. At the same time, companies are seeking to prioritize projects that can be industrialized, and that have a cross-functional impact on the company. Reconciling these two requirements is a challenge.
What are the key benefits of data transformation for companies?
Firstly, what does a company have to lose by not embracing data transformation? Data and data processing have such a major impact on a company’s internal and external activity that not committing to this transformation process would be result in lower levels of productivity, performance and agility. At present, companies aiming to be data-driven are looking for productivity gains, as well as enhanced agility and the ability to adapt much faster to ever-changing markets. For the moment, the focus is not yet on product innovation or new offers, but this phase will follow.
Nevertheless, data transformation is not a panacea; it’s is a vision, a policy and a mass of initiatives which raise new challenges and issues. It is, therefore, not a solution in itself, it is both a new challenge and a set of opportunities.
Is data transformation reserved exclusively for data experts?
Not at all, data transformation primarily concerns non-experts. “Innovating how you do business” is not the exclusive domain of data scientists, no matter how good they are. Data projects must be business-driven and led by the different professional sectors in the company.
All the complexity of a data transformation policy lies precisely in ensuring cooperation between data experts, information system experts and businesspeople. In this way, data experts determine what can or cannot be done with a data set, or inversely, they define the type of data required to obtain certain results. Managers are the only ones who can set targets and expected gains, and tell data experts if the result of the data processing is meaningful or not for their business. This is why virtually all data projects use an agile approach, to facilitate communication and make any necessary adjustments at each step of the process.
What role do employees play in data transformation? How can we involve them?
The main obstacle to data transformation in companies is also an opportunity, in other words it must be driven by the different professions and operational teams, that is to say by the business. This is one of the major differences between digital transformation and data transformation. The former can be focused on certain departments or overseen by specific teams, whereas data transformation impacts all activities. Business people are not data experts. This why effective communication between data experts and managers and developing a “data culture” as widely as possible throughout the business activities of the company are so important. Beyond awareness, managers must be trained in data projects. Of course, all this must be supported by a clear directive from management to become a data-enabled or data-driven company, in order to establish the key priorities.
I also firmly believe that business culture and data expertise will grow with the spread of use cases. The more data projects a company launches, the more managers will participate and understand the interest of these projects, and the more this culture will spread, and have a ripple effect. There is therefore a threefold challenge of awareness, training and emulation.
Could you tell us how the Data for Managers program at HEC Paris was created?
Data for Managers was co-developed in collaboration with partner companies that we have been working with on data issues. They told us that “what holds us back in data transformation is not the issue of data science skills, but that, when it comes to data opportunities, managers are like rabbits caught in the headlights.”
The idea was therefore to design a program that would enable managers to understand the issues around data science and data transformation, and beyond that, understand how a data project is led, their role and that of their teams in this project.
This said, and to our great surprise, around 20% of participants are tech and data experts, who we hadn’t targeted. They tell us: “I signed up to learn about data issues on the business side and how to work better with business teams.” This works and is fully coherent with the aim of the program.
Why did you choose to form a partnership with Netexplo?
Netexplo is an observatory for digital innovation and has been a partner of HEC Paris since its creation around ten years ago. Our surveys on corporate data policies were carried out via their business club, in association with HEC. It was Netexplo who asked me to think about a training program which would respond to the needs of businesses. They suggested a three-way co-development process to the Board of HEC Paris and Executive Education: HEC for the content and teaching expertise, a few companies in which the program would be tested, and Netexplo, who have the agility to rapidly build and co-promote the program. For example, when France was in lockdown, we shifted the program to a digital format, which meant adapting, recording, producing and posting the content online within four weeks!
Can you tell us more about the program and initial participant feedback?
Basically, the aim of the program is to provide training in four key skills:
- Understanding the new concepts, uses and legal framework, as well as the data issues of their company and business
- More clearly defining the role they play in a data-driven company and in the data chain of stakeholders
- Increasing awareness of best practices to identify, collect, add value, store and use data
- Initiating or participating in concrete projects within their company
Over the course of a year, the online program was followed by more than 1,000 participants, with a record level of satisfaction (4.5 out of 5). As a result, companies send us new participants. Satisfaction aside, the most rewarding feedback is the extremely high level of positive responses in terms of the relevance of the program to their work and their business.
What would be your last piece of advice for companies beginning their data transformation?
At present, companies that are not data natives and that have a strong desire to become data-driven tend to be very modest in their expectations, all emphasizing the fact that they are at the beginning of the process. Their challenge is to move from the profusion of localized POCs to organizational and industrial data projects, without stifling local initiatives. I don’t believe in just “one best way”, each company will manage this dialectic based on its own culture.
But as Bob Noyce, the inv entor of integrated circuits and one of the founders of Intel once said: “Optimism is an essential ingredient of innovation. How else can the individual welcome change over security and adventure overstaying in safe places?”
Julien Lévy is Associate Professor (Education Track) and the Director of the Innovation & Entrepreneurship Center at HEC Paris.
Julien has completed extensive studies in philosophy and political sociology (Panthéon-Sorbonne University, Paris), is a graduate of Science-Po and HEC, and earned a PhD in management from HEC. He has been a Visiting Scholar at the Amos Tuck School of Business Administration in Dartmouth and a Visiting Researcher at the University of California in Berkeley.