Data Mesh, acculturation, proof and Granny at the service of Data transformation

Big Data & AI Exhibition, Paris – French companies, united within the French Tech Corporate Community, have been working together for several years now on the subject of transformation through data. Issues around skills, valorization and generative AI are addressed collectively.

Maturity is therefore progressing within large groups, to the point where members of the organization are now collaborating on use cases in terms of data sharing. However, “the levels of maturity vary greatly”, notes Guiraude Lame, Chief Data Officer of Natixis and pilot of the FTCC.

Truly strategic Data in 5% of companies

Stéphane Bout, head of the Digital skills center at McKinsey France, confirms this observation. Level 1 of maturity (performance management or decision-making) concerns “the vast majority of companies, if not all”.

55% of organizations have started deploying AI in at least one function. One in two can thus claim level 2 of maturity. On the other hand, for McKinsey, less than 5% of them can claim to rise to the 3rd level.

“We are still focused today on the use of AI to transform the core business,” notes the consultant. Governance, skills, platforms… several of the fundamentals are taken into account.

“On the other hand, the identification of use cases still remains very empirical and bottom-up, which can sometimes lead to a fragmentation of investments in AI. Only a few companies have managed to adopt a truly strategic approach with investments focused on a limited number of initiatives,” says Stéphane Bout.

Masterclass for Suez top management

How then can we move from fragmentation to strategic? The participants of a round table organized as part of the Big Data & AI Paris show made several recommendations. As a preamble, Claire Mathieu, CDO of Suez, recalls that these transformations take time.

And the complexity of large groups is part of the explanation. “The first lever is acculturation, targeting members of the Executive Committee, managers, directors. We did this through a training program structured around 4 masterclasses for all members of top management,” she testifies.

To get the messages across, the Data Office invited external speakers. That was a few years ago and the results are there. “Acculturation was a trigger to initiate the beginning of transformation.”

This component does not only target the top. The round table experts agree to emphasize the benefits of a dual approach: top-down and bottom-up. To transform, Mick Levy, director of Business Innovation for B&D, calls on “two great wise men”, an economics professor and his grandmother.

Data Mesh to adapt to each profession

These two experts were called upon in 2022 to help a CDO convince internally. The economist recalls that the staircase starts at the top, that is to raise awareness among the Comex, “by speaking their language, by showing the value and by illustrating with numerous use cases”.

And grandma? She reminds us that “if you leave without knowing where you come from, then when you arrive, it’s not there”. Translation: a starting point in the form of a maturity audit is preferable. The meeting of the two wisdoms leading to adapted actions thanks to an “electroshock”.

Mick Levy and Claire Mathieu also promote an approach based on Data Mesh. Formalized by Zhamak Dehghani, it is based on 4 pillars:

  • The division into Data Domains
  • Data as a product
  • Self-service data infrastructure as a platform
  • Federated governance.

Data Mesh is not magic though. It is appropriate, firstly, “to adapt to the level of maturity of the business teams.” The value of this approach consists of giving autonomy to professions, promoting an appetite for data and thus encouraging the creation of (shared) Data Products to prevent silos.

Source link -97