Many organizations have the objective of becoming data-driven, i.e., basing their strategic decisions not on hunches or trends, but on accurate, reliable data and analysis. This implies a process of storing, documenting, and making available this data to make the most of it. If these companies equip themselves with modern tools to democratize access to data, they are faced with a multitude of difficulties that can slow down the process. This article is based on our experience with the Actian Data Intelligence Platform users from organizations of various sizes and sectors to describe 7 obstacles frequently encountered on the road to data democratization.

Tools are not Sufficient

Among the users of the Actian Data Intelligence Platform solutions, the democratization of data and the desire to switch to a data-driven decision-making model are, of course, major priorities. Moreover, access to the data of these organizations is partially democratized since they are all equipped with dedicated tools like data lakes and data labs. Naturally, the deployment of a Data Catalog in these companies is also an illustration of this, with the use of a unique platform capable of centralizing an entire data ecosystem that is shared with all employees.

These tools are essential building blocks for any data-driven approach, but they do not, on their own, make access to data more democratic. If we take a data catalog, for example, the tool becomes especially effective when it is used by the largest number of people in the organization. It is the multiplication of use cases and the documentation of data assets by as many employees as possible that allows the value of the company’s information to be unlocked. Everyone at their own level can then benefit from the work of their colleagues, a virtuous circle in short. To encourage this, a cultural change is necessary.

Shifting Corporate Culture

There is sometimes a lack of awareness of the value of the data available in the organization and a lack of commitment to the process of documenting and sharing data. The challenge lies in the use of the tools mentioned above, with data often remaining in silos between the different departments and teams. This mindset is even more difficult to change at the business level, whereas IT teams are culturally more aware and inclined to document and share data.

Governance units were created to promote this awareness, but the lack of legitimacy within the organization complicates their work of raising awareness of the central role of data for the company. In data mesh literature, it is recommended to federate/decentralize data governance. Business teams must be integrated into this process, at the risk of creating a language gap: governance teams must work with data owners, data engineers, data analysts, etc. The democratization of data access must involve both data producers and consumers.

The notion of changing the company’s mindset is a necessity to complement the tools in place to democratize data. Research published by Gartner shows that historically, organizations have evolved into a defensive culture of “never share, except” for good reasons to share it. The research institute insists on the need to switch to an “must share, except” philosophy. Tools (data lakes, data labs, data catalogs, etc.) are not enough to democratize data if they are not supported by this cultural shift.

Documenting After the Fact

Many projects are primarily driven by costs and time, and in these cases, data governance and data quality are typically not priority topics from the start. There is a tendency to document after the fact, making sharing and documenting data more difficult. Data quality, and even more so its documentation, is all too often the last task to be executed.

The Lack of Time

The lack of documentation is a bias that is heightened in organizations whose products and value are created through the exploitation of data – where the obstacle to democratization is more related to the lack of time for documenting than a lack of data culture as mentioned above. If we go back to the example of a data catalog and focus on the data scientist profession, we can see that this type of population has more or less the desire to document its activity but does not take the time to do so, since the completeness of the data catalog is not a priority.

Furthermore, documenting and making data available is not always part of the employees’ mission. There is therefore also an HR dimension to data democratization. The documentation mission can be added to the scope of the employees’ responsibilities to promote democratization and accountability.

The Volume of Data

A form of fear sometimes arises when contributors are asked to share their own business data within a large common container (a data lake or data catalog). This is the fear of finding oneself drowning in an ocean of data added by other entities of the organization, and of not being able to find one’s way around.

The data catalog is a valuable tool to alleviate this fear among data producers. Indeed, the tool offers them the possibility to not only easily explore their own data, but also to use data produced by others for their own use cases.

Data Security

The security aspect regularly comes up as a pretext for not sharing data within the company. However, there are effective systems for managing user permissions, such as the one integrated into the the Actian Data Intelligence Platform data catalog, for example, which, coupled with a culture of sharing and accountability, can make it possible to overcome this barrier.

Data Ownership

As far as the notion of ownership is concerned, we too often observe ownership of datasets at a local level. Yet data is a corporate asset, a common heritage, and only regulatory aspects should justify local ownership. In other cases, this ownership quickly becomes an obstacle to documentation: the corporate culture must favor making data available to the greatest number, under the responsibility of an entity or individuals.

If you would like to discuss the obstacles to the democratization of data described in this article, or if you would like a presentation of the Actian Data Intelligence Platform solutions for data-driven companies contact us.