Auchan builds its adaptable Data strategy around the Data Mesh
Big Data & AI Paris Exhibition – The Data Mesh was a featured theme during the 2022 edition of the conference. One year later, the subject has not been relegated to the background by generative AI – not totally.
The holistic approach indeed finds its place in organizations such as Suez, Carrefour and Auchan Retail. Samir Amellal, who combines the caps of CIO and Chief Data Officer of the distributor, came to share his conception of the Data Mesh.
Data a factor of adaptability against shocks
The valorization of data constitutes a strategic challenge for Auchan, in particular in the current context marked by high inflation and a low concentration of mass distribution in France.
There are also various “shocks”, including inflation and recent riots. In this environment, “Data is a factor of adaptability”. The mission of the CDO therefore consists in designing an “adaptable strategy” on tech and data in a very Darwinian conception.
Samir Amellal illustrates this through a use case of artificial intelligence put into production to adapt to the inflationary context. Through an algorithm, Auchan retail measures price elasticity and can thus predict consumer behavior based on this parameter.
AI is a lever for action, but it is not the only one. For the CIO, the primary ambition is that of the democratization of data in the company. Auchan “generates a huge amount of data”, the most obvious source of which is the sales receipt.
HR, logistics, marketing … are all producers and consumers of data. On marketing, for example, the retailer recently announced an initiative in retail media to promote its data to brands.
Ambition democratization of data for Auchan
”This issue of the democratization of Data is truly a subject of change for us,” emphasizes Samir Amellal. “All professions are destined to transform”, and AI is a massive vector of transformation. The pricing profession has experienced this.
“The boss of pricing has been doing his job for 30 years (…) The first reflex is to consider that AI will intervene in its business process.”But to produce value, AI must be designed with this expert, who thus participates in reinventing his business process.
But this ‘change’ does not only concern professions. IT and Data teams are also impacted, for example through the use of generative AI to produce code and empower business employees.
The Data Mesh is therefore essential for Samir Amellal as a natural evolution of the Data Office and the way to democratize the uses of data for the purposes of reinvention. And if the approach is interesting in different sectors, it is because it is considered as a way to overcome the bottlenecks arising from a centralization of data experts.
BI self-service and a Mesh portal
It is impossible to democratize widely without at the same time decentralizing skills and responsibilities. The central Data teams therefore change their role, to manage federated governance, standardize practices and support adoption.
The first Data Mesh step of Auchan Retail involves the progressive deployment of BI self-service – based on the Tableau solution. The low code no code raises the barriers to entry, but beware of governance, warns the CDO to limit the creation of technical debt.
Tableau is currently being deployed. The business departments gain autonomy, frees up resources within the Data Office, and at the same time take responsibility for the Data Products designed.
This conception of the Data Mesh constitutes “the starting point for the dissemination of the Data practice in the company”. And the work promises to be considerable, if only to change the culture and convince the professions that the exploitation of data is an integral part of their function.
“The stage promises to be perilous and we are approaching it with great care”” says Samir Amellal, who adds that the less mature professions, HR for example, will benefit from close support. The Data Office also remains central for the most complex projects.
On the decision-making side, the departments now have a platform, a data and AI portal called Mesh. The product is intended to be a Netflix of the data available to employees, on which they will also be able to propose new reports – as a condition of documenting and applying the governance rules.