Short Answer
Age is not the test. A fifteen-year-old system that does its job, can still be changed safely, and runs on a supportable platform is an asset, and ripping it out because it looks dated is how businesses spend six figures standing still. Replace when one of three things is true: the platform underneath it is genuinely dying, the system actively blocks things the business needs to do, or the cost and risk of changing it has grown past the cost of replacing it. Between “keep” and “replace” there is also a middle path, modernisation, that is right more often than either extreme.
The Wrong Reasons to Replace
It looks old. Users are used to it, and a familiar ugly screen is often faster in trained hands than a beautiful new one. Aesthetics alone never justify the cost and disruption of replacement.
A vendor says so. Vendors of new systems are structurally enthusiastic about the death of old ones. Treat “end of life” claims as a fact to verify rather than a starting gun. Sometimes the platform really is losing support. Sometimes only the expensive support contract is changing shape.
A new hire prefers something else. Every technical person has tools they would rather work in. That preference is information, not a business case.
The Real Reasons to Replace
The foundation is failing. The language version, framework, or operating system underneath the software is out of security support, and the people who can work on it are retiring faster than they are being replaced. This is the one clock you cannot argue with, and pretending otherwise turns into an emergency migration at the worst possible moment.
It blocks the business. Customers expect an online view the system cannot offer. A new service line needs data the system cannot model. Integration with a key partner is impossible because there is no API and no safe way to add one. When the system dictates what the business is allowed to want, it has moved from asset to constraint.
Change has become dangerous. Every small amendment breaks something unrelated, testing is manual and fearful, and the one person who understands the internals is a single point of failure. Fragility compounds quietly: each risky change makes people avoid changes, which makes the eventual necessary change larger and riskier still.
The Middle Path Most Businesses Miss
Full replacement is a bet-the-quarter project. Modernisation takes smaller steps that buy most of the value: wrap the old system in an API so new tools can talk to it, rebuild only the module that hurts, move the data to a supportable store while keeping familiar screens, or strangle the old system gradually by building its replacement around it a piece at a time. These routes keep the business running on proven logic while the risky parts get retired in order of urgency, and they turn one terrifying cutover into a series of boring ones. Boring cutovers are the good kind.
Costing the Decision Honestly
Keeping a legacy system has a price too. Add up the support contract, the manual workarounds staff perform daily, the projects not attempted because the system could not support them, and the key-person risk. Businesses meticulously price the replacement and leave the cost of standing still at zero, which rigs the comparison. Priced honestly, the status quo is often the most expensive option on the table, and occasionally it is still the right one for the next two years. The point is to decide with real numbers on both sides.
How We Approach This
Legacy modernisation is exactly this work: assessing what a system still does well, what is genuinely at risk, and sequencing the smallest safe steps, whether that ends in a wrap, a partial rebuild, or full custom replacement with the data migrated intact. Our first recommendation is sometimes “keep it, change nothing yet, fix these two risks”, because that is occasionally the truthful answer. If you want an honest assessment of a system everyone has opinions about, get in touch. Related decision: if the replacement conversation keeps circling a specific product versus building, custom software vs off the shelf frames that half of it.
