At many companies, there are some data systems and applications in use that were state-of-the-art when implemented but now seem dated, hard to use, and lacking some key functionality. That once-sexy system is now squarely in the “legacy software” bucket. While many assume legacy systems need to be completely overhauled or replaced, this doesn’t have to be the case—there are other compelling options available.
OUTDATED CLOTHES OR OUTDATED KITCHEN?
There are cases where old, outdated legacy components need to be fully replaced. Like with clothing, once something is out of style, you really don’t have much choice but to buy something new. Altering existing clothing to fit a new style often isn’t worth the cost or effort. Of course, you can hide the old clothes away for a few decades and hope that, like with bell bottoms, the old look will make a comeback. But that is a long shot at best, and you’ll still need new clothes in the interim!
In other cases, like with a kitchen, there is often a reasonable middle ground between a full rip-and-replace and leaving the dated look in place. Many people will leave the plumbing, cabinet structures, and appliance placement as-is, but replace cabinet doors and hardware, update countertops, and replace appliances. While the fundamental structure of the kitchen remains unchanged, it can look almost as good as a full remodel at a fraction of the cost and effort. This is because the outward-facing components of the kitchen have been modernized, and most people won’t realize or care that the infrastructure underneath what they see and interact with hasn’t changed.
This begs the question: When dealing with legacy data systems, do they need to be tossed and replaced like outdated clothing, or can they be updated like an outdated kitchen? While many assume that a full replacement is required, this often isn’t the case—and making that assumption can be costly.
THERE ARE OPTIONS OUTSIDE OF REPLACING LEGACY SOFTWARE
Gartner posted an article outlining several options to modernize legacy systems. In the piece, they provide seven increasingly difficult approaches to modernizing a software system or application, all of which can be applied to data-focused systems. At the difficult end of the spectrum, of course, is a full replacement. However, at the other end of the spectrum in the easiest-to-implement spot is what Gartner calls encapsulating.
Encapsulating involves leaving a system much like it is, but putting some wrappers around it to enable it to interact as though it isn’t a legacy system. By shielding a legacy system behind a series of modern services and APIs, users will perceive the system to have been modernized since the user-facing components are updated, even though the underlying system hasn’t actually changed.
Similar to the kitchen analogy, encapsulating an outdated data application or system can offer significant benefits to users without requiring expensive infrastructure replacements right away. Eventually, the core system itself may need to be replaced. The encapsulation process postpones the full replacement and allows a phased, incremental approach to be pursued instead.
HOW COMPANIES CAN RISE TO THE CHALLENGE
If your organization wants to go with the encapsulation approach, what is the best way to do it? One of the newest entries in the market is a concept called a data operating system (full disclosure: The Modern Data Company offers a product called DataOS).
A data operating system has the fundamental goal of offering a consolidated perspective of all data-related resources and activities within an organization and ensuring that reliable and compliant data is supplied to the business, AI/ML, analytics, and IT teams. Similar to a conventional operating system, a data operating system is flexible and can adopt various data architectures, such as data mesh, data fabric, or data lakehouse, among others. These architectures can be easily implemented and modified to keep pace with evolving business requirements.
Regardless of the approach taken, companies should be able to leave legacy data systems behind the scenes while simultaneously providing users with a modernized view and updated functionality. Security and governance should also be simplified by adding a single, consistent layer across an organization’s infrastructure. Being able to centrally manage security, access rights, and data processing improves and modernizes corporate systems without the expense and risk of fully replacing everything from scratch.
Before your organization undertakes the costly and time-consuming effort to completely replace many of its systems, first assess the feasibility of the encapsulation approach or other ways to keep legacy systems in place behind the scenes. It could well help save money and time while lowering risk. Why not at least explore the option?
Emanuel Younanzadeh is VP Marketing at The Modern Data Company.