This article is part two of our three-part blog series based on our position paper about developing data-centric software. In the first article, we explained why many organizations get stuck in their software landscape and how data-centric information management offers a way forward. Now, we turn to a crucial question: what role does software itself play in unlocking or restricting the potential of your data?

The challenge with traditional software

Most organizations recognize the challenge: data is scattered across applications, inconsistent between departments, and hard to unify for decision-making. At the root of this issue are long-standing habits in the way organizations buy and use software. 

Traditionally, organizations have operated within two limiting paradigms:

1. Off-the-shelf software
Standard software packages are designed for broad markets. Organizations must adapt their unique processes to fit the tool’s limitations to use it. Functionality and user interface often take priority, while questions about how data is structured and shared are pushed aside. The result? Rigid formats lock in data, making it difficult to repurpose or exchange.

2. Bespoke data structures
On the other hand, many organizations build their own bespoke data structures in tools like Excel, low-code platforms, or fully custom-built software. While these solutions fit immediate needs, they often lack the governance required for effective data exchange and reuse. 

Neither of these paradigms helps organizations become more data-driven because they generate and store data for only one very specific purpose.

Figure 1: Level of flexibility of data management applications

A data-centric alternative

To move beyond these limitations, organizations need a different approach. Therefore, we propose a strategy to develop standardized, reusable data structures that serve the needs of the business, not just the application. 

By embracing this approach, organzations can: 

1. Break down silos between applications and departments. 
2. Improve interoperability across tools and workflows
3. Unlock the full potential of their data assets.

This does not mean that software is no longer important. On the contrary: we still need applications to run processes, visualize insights, and make work more efficient. But instead of letting applications dictate how data is stored and exchanged, we should design applications around data models that are flexible, standardized, and reusable. That’s the first real step towards becoming a data-centric organization.

Conclusion

Traditional software, whether off-the-shelf systems or bespoke one-off solutions, locks your data in. To unlock its full potential, organizations need to shift towards standardized, reusable data structures that center data, not applications. 

This makes it possible to improve interoperability, future-proof your information, and truly work in a data-driven way. This does not happen overnight, but by starting with small steps you can begin breaking free from application-centric constraints and move toward a data-centric future.