53% of banks have already defined (or are in the process of defining) their strategy for artificial intelligence in synergy with Data Strategy; 71% of lending institutions are aware of the importance of initiating AI training plans, while 82% plan to engage heterogeneous teams to “ground” such projects. With what skills? The most sought-after skills are those of data scientists and data engineers (76 percent and 71 percent of cases), but business areas responsible for managing risk are also in the foreground (65 percent privacy, 35 percent cybersecurity, 35 percent risk & compliance).
This is highlighted in the latest report by ABI Lab’s Information Governance Observatory, of which Irion is a partner, presented this week at a workshop in Milan. To describe the path taken by banks toward a stable governance of data, the survey highlights some metrics related to staff engagement: 55% of institutions monitor Data Governance training, 64% measure the participation of their employees in a Data Community and 72% the level of engagement of internal resources on Data Management activities; 60% have gone a step further, including the level of implementation of these roles and processes among the KPIs monitored by management.
Sustainable quality and management control
“There is a sustainability issue in data governance today: we need it to be more automatic and efficient, especially with regard to Data Quality. Our Data Owners are decentralized: proximity to the business is key. We have done a lot of work on mapping data usage; each area negotiates resources annually to launch priority initiatives,” says Luca Giordano, head of data governance at Intesa Sanpaolo. There is a lot to be done in this area, so there is a need to measure what brings value.
“Today there is one more productive factor: data. The others (HR, technologies and processes) have long been under the eye of management control,“ Mario Vellella, domain advisory leader at Irion, explained at the workshop.” The same should be done for data: “They deserve a section in every business plan and monitoring, even quarterly, of their contribution to the achievement of the business strategy, to answer the question of the that top management constantly asks us: ‘Is data bringing the value intended by the strategy?”
In these areas, the European Union has moved forward with a number of initiatives such as the Data Governance Act and community data spaces. “These regulations take for granted that public and private entities should share data:but what if I don’t know the value?” This is a problem also highlighted by Gartner in its interviews with CDOs: when it goes well, they manage to show what Data Quality is worth, the rest is difficult.
A metamodel to support Data Valuation
A key aspect in the process leading to enterprise data valuation – Vellella continues – is the ability to use a metamodel to link the representation of the data to its intended use. This makes it possible to understand whether the data is being used on a particular process, project or product. “I will have to identify items that are already familiar to management control and represented in the company’s balance sheet: it will be easier to disclose them. It is also a proactive way to use metadata, which is often laboriously collected, through interviews and connectors.”
