Data and AI, from technological promise to tangible business impact!

I use the term “commitment” and not “culture” because, for me, we will need to get away from the false pretenses and fantasies of “magical AI” and other “data-centricity” concepts that have led people to think that capitalizing on data and AI can be done by fiat and can emerge alongside existing business processes without really being rooted in anything! A sales manager who wants to become “sales data driven” must understand that this first requires commitment and some investment (e.g., accepting that salespeople “waste” 10% of their time entering data into the CRM software) before paving the way to future benefits (i.e., tools to aid decision-making and sales initiatives).

But other processes (sales, supply chain, production, customer relationships, finance, HR, compliance, etc.) are often more complex because they require the coordination of digital (with more or less open solutions) and human elements. In such cases, that way of working is much more complex, if not impossible, and it becomes necessary to rethink the end-to-end business process. Here, there is no choice: the business line must engage in an in-depth transformation.

Very often, the only way to really get things moving is by forcing them through, backed up by close support—to put it mildly—from data experts. After several years of acclimation and other incentives in the form of project funding, the right balance must now be found between encouraging people to make the shift and compelling them to do so. Compelling adoption is no more or less than imposing a serious and in-depth induction into “Data/AI” before launching any business project. Logically, this implies putting everything to a stop if the adoption is deemed insufficient. Encouraging means providing data offices and data factories with real support resources for the business lines (data product owner, data-evangelist, etc.), resources that will help them to think differently about their processes by making better use of data!

This will prevent Data/AI teams from falling behind business projects or proposing solutions that may sound incredible, but which rarely fit into a business project that is already underway.

As always, some will try to make the problem more straightforward with overly simplified concepts. In 2023, these will include “Data-mesh”, “Data-Fabric”, “Customer-Data-Platform”, “MLOps”, “Low-code / No-code” and so on.

Far be it from me to denigrate such concepts, all of which contain some nugget of truth, and sometimes genuine technological advances. But they often make us forget that AI and data are first and foremost a question of business commitment, or—in more practical terms—of organizing and integrating data models into business processes. They do not get anywhere if the business teams are not truly committed to the data and AI objectives. To take just one example, data-mesh solutions are currently interesting and, in some cases, fruitful. But their technical implementation will result in nothing if the production, quality, data documentation, and part of the handling has not been transferred to the informed responsibility of the business lines, with all the ensuing implications for the teams and their skills!

While some usage cases will continue to apply to the majority of companies (cookieless targeted marketing, omnichannel customer relationships, fraud, operational excellence, etc.), 2023 should, in my opinion, give priority to investments that will favor those of AI and data while taking into account and affirming certain differentiating factors. There are many examples, depending on the industry. Consider a company that has made the physical customer experience a core part of its value proposition. Logically, this will put the focus on data/AI usage cases that will further enhance the quality of the physical experience (prediction of peak customer traffic times, making an appointment with the customer’s preferred salesperson/advisor, personalized training of salespeople/advisors, etc.). In the same way, a company that has chosen to become mission-driven will give greater weight to usage cases that fit in with that mission.

Unsurprisingly—and true to that logic—the acceleration will be even stronger for AI for Green/for Good projects serving the company’s CSR approach, from the basic but complex reliability of ESG indicators to more ambitious usage cases of truly green finance, genuinely scrutinized supply chains, inclusive banking/insurance, less energy-intensive production processes, and so on.