Big data is not just a technological tool or a means available to companies. Extending the field of business intelligence, big data has become a prerequisite for the ISD within the context of the digital transformation of a company seeking to innovate. And in this context, urbanisation of data will change, perhaps even drastically, the governance of the ISD.
“The biggest challenge of big data is to transform the ISD!”
Is big data a “driver” of the ISD? Yes, as we don’t need to stick to the simple functional acceptance of the term. Most certainly; it produces a new mass of collected data, as well as the emergence of new cross-functional uses. But it’s even more than that: big data, which extends the field of business intelligence (BI), makes data a new company asset which helps to create value.
It’s much more than a technology, a tool or a means available to a company. Having become one of the main drivers for innovation , big data deals with the biggest challenges of the ISD and guides its transformation. Data urbanisation therefore drastically changes the governance of the ISD. With a string of challenges concerning organisation, architecture, HR…
Besides the technological aspects which are beginning to be grasped, implementing big data involves eliminating the historical compartmentalisation of data and transforming it into a valuable asset for business processes. It also involves transforming the management approach adopted by information systems by developing data-specific skills, business units and governance.
Urbanising data: a fundamental change
Data urbanisation is at the heart of the transformation of the ISD by big data. It is essential to understand it by taking into account new data and the value assigned to them: we need to shift from an urbanism of applications to an urbanism of data. This change is quite profound.
This urbanisation requires measures to limit data being incessantly repeated in the IS and requires tools such as a Data Management Platform (DMP) to facilitate the management of data exchange flows. But urbanising is primarily about governing all of the mapped data available. They must be linked to the processes and activities, as well as their “owners” (which are often multiple in number), namely the business units. It will also be necessary to implement good reference systems (customers, products or services, organisation, etc.).
Main challenge: organisation…
And governing…is organising. Firstly in HR, with the increased proficiency in technologies. The sources of information are rich and considerable training is available on these topics (MOOC, e-learning, training centres, support for publishers). Moreover, ISDs can be supported by service providers who are able to offer their expertise. The resources are available – it’s about organisation and investment: create an action plan and support its implementation.
We are also seeing new types of application appear with new business units which are necessary to ensure the correct use of data (to avoid interpretation errors). Data scientists (of which there is a shortage of between 140,000 and 190,000 in the United States at the time of publication, according to McKinsey) and data stewards are the profiles in mind. But bear in mind that these are roles that fall more under the category of business units than the ISD, whilst having an impact on it: the ISD must therefore adapt itself by finding a “mirror” organisation to work with these new business units.
Moreover, the variety of data collected and the increase in collection points at the heart or on the fringe of existing processes blurs the boundaries between the professions. As Becoming a valued asset of the company, data will need an “arbitrator” and a manager ensuring that all stakeholders are respected (producers or consumers, internal or external): the chief data officer (CDO). This is the approach adopted by numerous organisations and, since 2014, by the French State with the nomination of Henri Verdier as General Data Administrator (AGD) for the State.
This organisational challenge also involves a cross-disciplinary view of the data. The trap to avoid, for example, would be the implementation of a cross-disciplinary data lake to produce data silos for each core business under the pretext of security: this would go against the fundamental principle of data sharing, only capable of really delivering the expected value.
In most cases, it will be necessary to shake up the standard structure of the ISD in application silos, to introduce a cross-functional data-driven organisation and user application services for data.
… without forgetting the architecture
Finally, one of the key values of big data lies in being able to combine data. The issues related to decompartmentalisation of data are fundamental: confidentiality, governance and implementation of a common reference system as each division has its own vision and therefore its own definition of the company’s objects. The challenges of the data architecture also concern governance, convergence of reference systems, and definition of the points of convergence between the various data sources.
Ultimately, in this new context, the challenge of transforming the ISD emerges as a division of responsibilities concerning the data on the one hand, and concerning the integration of IT services on the other hand, with an increasingly clear distinction between CIO (Chief Information Officer) and CDO (Chief Data Officer).