The Digital & Data maturity of Lake Geneva’s enterprises

By way of introduction, it may be useful to present a few key facts about CDOs: Experienced professionals or Data/Digital Heads stand around 600 employees in the Lake Geneva region. This figure has risen by 4% in one year, and is mainly split between the two major cities: Geneva (55%) and Lausanne (45%).

The maturity of organisations when it comes to employing CDOs differs greatly from one company to another.

The sectors where Data & Digital teams are the biggest are :

• the Public Sector: +12% in 1 year
• NGOs: +7%
• Banks: +11%
• Luxury goods: +8%
• Tobacco: +13%

Nevertheless, brand new industries are starting to gain interest in having a Digital/ Data department in their teams and having a strategy around data governance. Industries that are newly recruiting in this area: Oil & Energy, Construction, Retail, Logistics & Supply chain.

Depending on the level of maturity of the organisation, or the reliability of the data being used, digital transformation issues can sometimes be entrusted to newly appointed CDOs. This is not necessarily someone with a background in digital technology.

One of the CDOs interviewed for this study works for a well-known company in the commodities sector: one of the main projects for 2022 was to implement a global digital strategy. Thus, why they choose to appoint as CDO an employee with a strong track record within the group (10 years working together) and to entrust him with this challenge because of his excellent knowledge of the company and information flows.

By its very nature, data is disseminated across all company departments. A CDO cannot therefore work in silos but must necessarily collaborate with business lines to improve their internal processes or optimize their external performances. Whatever the case, this implies a significant transformation in the way business stakeholders work (the tool they use, the AI-based recommendations they need to digest, …). Harnessing this strength, the track record of local leaders shows that they naturally evolve into cross-functional transformation positions.

The hiring process for a Data Scientist lasts an average of 10 months in Geneva, rising to a year for a Data Engineer, and it can take much longer for even more specific professions such as scan-to-bim experts.

To overcome this shortage, and in line with companies with strong local visibility such as Swisspost (which we did not meet during this study), more and more companies are turning to the creation of nearshore centers of expertise. Based on the sample of companies we interviewed (n = 25), around a third told us about their plan to set up an HR strategy. Lisbon appears to be the main choice, but Italy and North Africa were also mentioned.

Emphasized by the emergence of generative AI, the current power of algorithms has marked the end of the quest for technical performance. Leaders are no longer seeking to optimize the relevance of their models,but rather the rate of adoption of their data products by the business.

Take the case of a sales prediction use-case: the incremental business gain linked to optimizing the relevance of a point prediction (let’s say from 95% to 96%) is infinitesimal vs. the business value residing in the use or otherwise of this information by business stakeholders to optimize the allocation of end products in their points of sale, avoiding stock shortages. This adoption rate therefore becomes the sinews of war, and human support remains the best way to maximize it.

Far less sexy than the glamorized AI use cases, difficult to link to tangible business ROI, and complex to deploy with the business because time-consuming and constraining… Deploying data governance across multiple departments is no easy task. This complexity explains why leaders have been somewhat tardy in addressing this work stream, which is nonetheless crucial to transforming data into a differentiating asset for the company. The study revealed that many companies have launched such initiatives, showing their willingness to tackle this topic, but those usually remain highly localized and have difficulty scaling up.