The fact that data is a corporate asset seems to be widely accepted. For those who are still in doubt, the capitalization of the digital giants, who also happen to be Data champions, is particularly interesting to analyze. At the beginning of 2020, the GAFAMs weighed 5000 billion dollars, i.e. 2.4 times the weight of the CAC 40 (the 40 largest French market capitalizations). The gap has continued to widen these last months, with the GAFAMs who are now worth 7 trillion dollars.
Despite the growing consensus on the importance of Data, the effective transformation of practices is far from being trivial. To succeed, companies need to take action on at least three dimensions: technology, people and processes.
Technology is fundamental: it is indeed the technological evolutions of the last few decades with the Internet and e-commerce, connected objects, Artificial Intelligence, social networks and to a lesser extent (for the moment) the Blockchain that are at the origin of the changes. Companies have clearly understood this, and are increasingly equipping themselves with appropriate technologies such as Data management, Analytics and AI solutions, Big Data infrastructures and Cloud platforms.
But despite numerous investments in technology, it is estimated that less than 5% of companies today manage to really benefit from their data assets to create a competitive advantage. This conclusion is reminiscent of other observations on the success rate of major transformation programs in enterprises, which is estimated at less than 30%.
One of the key lessons learned from the studies carried out on the subject is that the transformation of practices on an enterprise scale cannot succeed without an appropriate investment in team skills (knowledge, know-how and interpersonal skills), communication and the implementation of appropriate processes. The evolution must take place on an individual level as well as in multidisciplinary collaboration.
To achieve this, we believe it is essential for data leaders to work on the following points.
Sharing vision and strategy
A clear vision that helps to envision a desirable target state is, in general, a first source of motivation for teams. It helps to establish a common direction, the absence of which can lead to confusion and inconsistency. The strategy, which specifies the directions chosen to achieve the vision, provides a frame of reference for resource allocation and multiple trade-offs. It enables the sharing of issues and creates the preconditions for team commitment and alignment by giving meaning to action.
Identify the variety of roles in the context and ensure the enablement of the stakeholders
Anyone who handles data on a daily basis knows this. The transversality of the subject requires the involvement of a variety of roles.
First of all, there are the leaders and managers who give direction, control, validate and arbitrate. In the Data sector, we find roles such as :
- Chief Data Officer and/or Chief Analytics Officer at an executive level,
- Data Lead, DMOs, Data Governance manager, Data managers, Data Quality managers, Data Analytics officer at different management levels.
In the business sectors, the correspondents of the Data roles are the Data Owners, generally high-level managers, and the Business Data Stewards who are their intermediaries. Of course, there are also managerial roles in the IT department, with application managers and architecture managers.
Then there is a diversity of operational roles that can be linked to the different areas of Data Management:
- Governance: under the responsibility of the Data leader, the people who traditionally have the roles of Data officers / Data managers / Data Stewards in charge of executing both offensive and defensive Data strategies (valuation and monetization, protection and regulatory compliance, quality), implementing policies and principles, managing Data knowledge (metadata, classifications) as well as Data acculturation and change management;
- Analytics and Valuation: BI professionals (descriptive), BI analysts or data analysts, data miners specializing in data mining, data visualization engineers, Data Science and AI experts for predictive (and in some cases prescriptive) that are Data Scientists and ML engineers;
- Lifecycle and Quality Management: specialists in referential data (MDM and reference data) and data quality professionals generally called Data Stewards / Data managers / Data officers;
- Definition and modeling: data architects, metadata specialists, data modelers who may be in the Data or IT department;
- Data protection: administrators, engineers and security architects who are in the IT teams under the direction of CISO and the control of the DPO;
- Infrastructure Management: the administrators, engineers and architects in charge of the Data Lakes, DataWarehouse and data integration infrastructures within the IT department;
Faced with multiple roles and the diversity of terms used to refer to them, several standardization initiatives are emerging.
Although a limited number of the roles mentioned above are, in general, formalized by a job description validated with the Human Resources Department, professionalization around data is well underway. It is therefore useful for companies to map the Data roles that exist in the organization (formal and informal), to carry out skills assessments and to initiate appropriate development plans (training, coaching, support) to make the transformation a success.
Promote multidisciplinary collaboration
Viewing Data as an enterprise asset involves managing Data as an economic resource that holds or delivers value; such as a product or service. To manage the lifecycle of "Data products" or Data use cases, it is therefore essential to establish effective collaboration between several stakeholders.
Not surprisingly, customer satisfaction - the internal or external users who have a need - can serve as a framework to organize collaborations. What customer issues can be addressed? What are the pain points that can be resolved? What progress can be made? The pervasive nature of data combined with technology innovations creates a variety of possibilities : refine customer segmentation, improve the targeting of marketing campaigns, identify churn, detect correlations associated with a business event, ... All business lines are potential customers.
To succeed, Data and IT stakeholders must be able to collaborate in an efficient and repeatable manner throughout the process : starting from the identification of needs to their formalization, through the design and implementation of a solution, to the provision and delivery of support. Agile approaches (and pizza teams) are of great interest for successful multi-disciplinary collaboration around use cases and "Data products".
Maintain effective and regular communication to disseminate acculturation Data
Communication is an essential factor in the success of a Data transformation program. Effective communication must remind the targeted objectives and adapt the content according to the needs of the different stakeholders.
Some good practices are :
- Describing the vision of success: this involves visualizing the target state to enable stakeholders to understand the impact of the change on their scope and project themselves into the future under construction.
- Clarity and relevance: clarity of purpose and its adequacy in terms of detail (neither too much nor too little) will be key elements of conviction. Too much jargon or too much emphasis on technical aspects can push some of the audience away.
- Share stories and examples: Analogies and stories are effective ways to describe and help people remember the goals of Data initiatives.
- Listen to feedback: The reaction of the audience is an indicator of the effectiveness of the communication. If one tactic doesn't work, it is important to try a different angle.
- Diversify the means of communication: exchange formats produce different effects. It is therefore important to vary the means and formats by which messages are transmitted: web page, blog, e-mail, social networks, individual meetings, presentations in small or large groups, lunch conferences, workshops, etc.
- Repeat over and over again: Most messages need to be repeated to be heard and understood by all stakeholders. However, this must be done in a way that repetition is useful for getting the message across and does not become counter productive.
--- Shelemat DANIEL, Octobre 2020