Legacy Systems: Should You Replace, Wrap or Integrate?
The old system everyone complains about may still be doing its job. A clear way to decide whether to replace it, wrap it with something better or integrate it with newer tools.

In many established companies there is a system that everyone complains about and nobody dares to switch off. It was built in-house years ago, the developer who wrote it has moved on, and it still runs production planning, policy administration or the student register every single day.
The instinct is to replace it. Sometimes that is right. Often there is a cheaper, safer option that keeps what works and fixes what does not.
Three options, clearly defined
Discussions about legacy systems get confused because people use the same words for different things. Here is what each option means in practice.
- Replace. Retire the old system and move its functions and data to a new one, either a package such as an ERP (enterprise resource planning) system or a custom platform. The old system is eventually switched off.
- Wrap. Keep the old system as the engine underneath, but build a new layer around it. That could be a modern web or mobile interface, an API (application programming interface) that lets other systems read and write data, or both. Users and other systems talk to the wrapper instead of the old screens.
- Integrate. Keep the old system as it is, but connect it to newer systems so data flows between them automatically. For example, orders from a new e-commerce site are pushed into the legacy order system, and stock levels flow back.
There is also a fourth path that combines them: replace the system in stages. You wrap or integrate the old system first, then move one function at a time to the new platform until nothing important is left in the old one.
A scenario: the production system built in 2009
Consider a metal fabrication manufacturer in the Sahab industrial area. Its production tracking system was written in-house on a desktop database in 2009. It handles job cards, material issue and machine time well, and the production team knows it inside out. But it does not talk to the accounting system, it cannot be used from the shop floor tablets the company bought, and only one external developer still understands the code.
Management is torn. A full ERP with a manufacturing module would solve the integration problem, but it would also mean retraining the whole production team and migrating years of job history. The real pain points are narrower: double entry into accounting, no mobile access, and the dependency on one developer.
In this situation, integrating the old system with accounting and building a simple tablet interface on top of it might solve most of the pain for a fraction of the cost and risk. The dependency on one developer remains, so the plan should include documenting the database and setting a date to revisit replacement. The choice is not permanent, and it is honest about what it leaves unsolved.
How to decide between the options
| Question | Points toward replace | Points toward wrap or integrate |
|---|---|---|
| Does the core logic still match how the business works? | No, processes have changed and the system forces workarounds | Yes, the problem is access, interface or connections |
| Is the technology still supportable? | Unsupported database, operating system or language, with security risks | Stable platform that can run safely for several more years |
| Can anyone maintain it? | No one, or one person close to leaving | A small team or a reliable supplier |
| How good is the data inside it? | Poor, with structures that block reporting | Reasonably clean and well understood |
| How much change can the organisation absorb now? | Capacity for a major project with training and process redesign | Limited capacity; other projects or a busy season ahead |
| Is there a suitable replacement? | Yes, a package or platform fits most needs | No, the system handles something specialised that packages do poorly |
Answers rarely all point one way. The table helps you see which concerns weigh most and makes the decision a business discussion rather than a technical preference.
Involve the people who use the system every day in this assessment. Production supervisors in the Sahab example know which screens they rely on, which reports they rebuild in Excel and which features they have never touched. Their answers often show that the real value sits in a small number of functions, which makes any of the three options easier to plan and to cost.
Risks to manage with each choice
Replacing carries the classic project risks: underestimating the hidden logic built into the old system over many years, migrating poor data, and disrupting users. Before replacing, document what the old system really does, including the small rules nobody remembers until they are missing.
Wrapping adds a new layer that must also be maintained. If the wrapper is built quickly without documentation, you end up with two legacy systems instead of one. Keep the wrapper simple, well documented and owned. A good test is whether a second developer could pick it up from the documentation alone.
Integrating creates dependencies between systems. When the integration fails silently, data falls out of step and people stop trusting both systems. Every integration needs monitoring, error alerts and a clear owner, and it should be designed so that failed transfers are retried or flagged, not lost. Automated reconciliation checks between the two systems are a sensible addition.
Making a decision you can explain
Run the decision table above with the system’s main users, IT (information technology) and finance in one meeting, and write down the answer to each question with its evidence. That one page usually shows which option is realistic and what must be true before replacement makes sense.
Whichever route you choose, set a review point. A wrap or integration decision made today should be revisited in two or three years, or sooner if the old platform loses support or the business changes direction.
Legacy system integration and system design are part of our Enterprise Systems service. We assess what your current systems do well and where they hold you back, then help you choose and deliver the option that fits.



