Data centric architecture

Many organizations claim to be data driven or data centric, using these terms loosely and interchangeably. But the two are not identical and have very different meanings.

Data driven vs data centric

For an organization or a company, being data driven means being driven by data. It is first and foremost a mindset that involves making strategic decisions based on data and information analysis.

Being data centric is much more than that. Companies that are consider data to be assets in their own right, properly catalogued. Data governance establishes an organizational framework and rules to support all data-related projects.

Data centric architectures

Being data centric also means adopting a data-centric architecture, thought of as a robust whole capable of serving the entire enterprise over time. It is secure, reliable and, above all, scalable in terms of volumes and uses.

In recent years, Big Data techniques have brought new styles of architectures – data lake, data hub, data mesh – allowing to answer almost any problem.

Initially developed by the web giants to cope with ever-increasing volumes, these techniques are now increasingly used in business computing, even though volumes remain relatively low. And for good reason, they bring new features. The most interesting one being thescalability of the storage space that can be expanded indefinitely thanks to the cloud. So, no more data silos: you can now have them in a single space and bring them together without limits and without compromising performance.

Benefits of the data centric approach

As a result, a data centric company has an undeniable advantage over its competitors because it can very quickly process all the data at its disposal.

For example, in case of war, pandemic or natural disaster… A manufacturer must be able to reorganize its supply chain very quickly. And it can do this thanks to the data centric approach that keeps all possible and imaginable data at its disposal with a controlled quality.

In contrast, a company that is not, takes days or even weeks to answer questions that can be vital to a quick response.

Becoming data centric is therefore a considerable challenge in the 21st century; but how to achieve it?

Become data centric

The four ingredients of a data centric organization are:

Alignment of the digital strategy with the company’s objectives

A transition to a data centric model requires an alignment of business objectives with the digital strategy.

Before this alignment can be achieved, the strategic needs of the business must be identified so that the technology can meet them. The digital strategy must satisfy the requirements defined by the different business lines of the company.

By making data accessible and usable by all its employees, a data centric company has the advantage of being able to change its digital strategy at any time to survive changing business conditions.

Data architecture and data science

Once the strategies are aligned, a company needs to put them into practice and start leveraging its data.

It must build a data architecture that puts data at the heart of operations, as explained above.

For a few years, the cloud has considerably facilitated such architectures and for good reason: it allows to free oneself from the limits of a traditional on-premises information system which are finite storage space, fixed investments and lack of flexibility.

The cloud allows you to outsource the management of your infrastructure and its evolution to a provider who specializes in this area. The three global leaders Amazon, Microsoft and Google are generally considered to be a guarantee of long-term stability.

This gives data architects all the building blocks they need to focus on designing a scalable data centric architecture.

This usually results in the ability to ingest, store, process and use data. However, without effective coordination with the trades, these technical resources would be largely ineffective. It is up to data governance to meet these requirements.

Data governance

Data governance establishes the human, technical, and organizational requirements for trusted data to uncover new opportunities; but also to lower the cost of data management, while reducing security and regulatory risks.

The three pillars of data governance
The three pillars of data governance

Reporting and data visualization

The purpose – and not the least – is very often reports and data visualization that will allow to pilot the company by taking the best decisions at a given time.

When companies first start looking at data, they often see it as a bunch of indistinguishable numbers. They need to understand the different ways to present data in a visual format. Most decision makers are not usually data scientists or statisticians. This is why it is necessary to use data visualization tools.

Data visualization software plots data to present it in a more understandable format: charts, graphs, maps… Business intelligence managers create dashboards, reports and visuals that can be adapted to different audiences within the organization.


The natural evolution towards thedata centric approach is a consequence of the increasing competition between companies. The very strong technical evolutions of these last years have largely accelerated it. Moreover, the cloud is almost a sine qua non for its success. Nevertheless, it must necessarily include alignment with the company’s strategy, supported by sustainable data governance.