What to ask before replacing a legacy system
Replace too early and you throw away something that works. Too late and you are running on something fragile. The questions to answer before you decide, and the middle path most businesses miss.
Every established business eventually reaches the same uncomfortable question. There is a system at the centre of the operation, it has been there for years, it is showing its age, and nobody is quite sure whether to nurse it along or tear it out. It is a genuinely hard call, and the cost of getting it wrong runs in both directions.
Replacing too early throws away a system that still works and knowledge built up over years. Replacing too late means running the business on something fragile that fewer and fewer people understand. Before you decide, there are questions worth answering honestly.
What is it actually costing you to keep?
Old does not mean bad. A legacy system that does its job, that people know, and that rarely breaks may be the cheapest option you have. The costs that justify replacement are specific: it goes down and stops work, it cannot connect to anything else, only one ageing supplier or one staff member can maintain it, or it blocks something the business now needs to do.
Write down the real cost of keeping it, in downtime, in workarounds, in risk. If that page is thin, the honest answer might be to leave it alone.
Is the risk in the software, or in the knowledge?
Often the frightening part of a legacy system is not the technology, it is that only one person understands it, and they are close to retirement or already gone. That is a real risk, but it is a different problem, and sometimes it can be addressed by documenting the system rather than replacing it.
A system one person understands is a risk. That does not always make it the wrong system.
Separate the two questions. Is the software failing, or is the knowledge about it disappearing? They have different and very differently priced answers.
Replace all at once, or in stages?
The instinct is a clean cut: switch off the old, switch on the new, one weekend. For anything the business truly depends on, that is the riskiest possible approach, because if the new system has a problem you have no fallback and no working business on Monday.
A staged approach moves one part at a time, with the old system still running underneath until each piece is proven. It is slower and it feels less decisive, but it means a problem is contained to one area rather than taking down everything at once.
The middle path most businesses miss
The choice is rarely limp along or rip it out. Frequently the right move is to modernise around the old system: put a better interface on it, connect it to your newer tools, or lift out the one part that is genuinely holding you back while leaving the rest in place. This keeps the value that still works and spends money only where the pain actually is.
Decide with numbers, not nerves. Cost the keeping, cost the risk, and prefer the staged, contained path over the dramatic one. The businesses that regret these projects are almost always the ones that replaced everything at once because the old thing simply felt old.
Not sure whether your process passes?
Tell us what it is and we will walk through the questions with you. No charge, no obligation.
Before you automate anything
Four questions worth answering about a process before you spend a dollar automating it. Most of the automation projects we are asked to rescue skipped at least two of them.
Spreadsheets are not the problem
For a lot of what SMEs do, a good spreadsheet is the right tool. This is how to tell when yours has quietly become a liability.