The most expensive software in your business is often the one you think is free. It was paid for years ago, it still runs, and nobody sends you a bill for it. That is exactly why the true cost of legacy systems stays invisible until something breaks at the worst possible moment.
I am not here to sell you a rewrite. Most legacy systems should not be rewritten, and many replacement projects cost far more than the problem they were meant to solve. What I want to do is help you measure the real cost of what you are running now, because you cannot make a sensible decision about modernizing until you know what standing still actually costs you.
The cost of legacy systems does not show up as a line item. It hides in three places: the hours your staff waste working around it, the risk of it failing when you least expect it, and the one person who is the only one who understands it.
The Workaround Tax
Every legacy system accumulates workarounds. The report that has to be exported to a spreadsheet and reformatted by hand every morning. The two systems that do not talk, so someone retypes the same data into both. The feature that broke years ago, so people keep a parallel process on paper.
Individually these feel like minor annoyances. Together they are a tax you pay every single day. Here is a rough way to price it:
Workaround cost = hours per week spent on manual workarounds x number of staff x hourly salary x 52
Try it honestly. If three staff each spend five hours a week retyping data and fixing exports, and their loaded cost is around 60,000 rupiah an hour, that is 3 x 5 x 60,000 x 52, which comes to roughly 47 million rupiah a year. For one workaround. Most businesses have several.
That number is usually a shock, because nobody ever added it up. The work is spread thin across many people and many days, so it never appears as a cost. But it is real money, spent every year, on a system that was supposed to be free.
The Single Point of Failure
The second hidden cost is risk, and it comes in two flavors.
The first is the system itself. Old software runs on old foundations: an operating system no longer supported, a database version with known security holes, a server sitting under someone's desk. It works until it does not, and when it fails, it often fails completely, at a time you did not choose. The cost is not the repair. It is the days your business cannot operate while you scramble.
The second is the person. Almost every legacy system depends on one individual who knows how it really works. They know which button not to press, how to fix the monthly glitch, where the undocumented logic lives. When that person retires, resigns, or simply takes a long holiday, the knowledge leaves with them.
Ask yourself a blunt question: if the one person who understands your core system left tomorrow, how long until you were in serious trouble? If the answer is measured in days, you are carrying a risk far larger than any software license.
Slow Onboarding Is a Cost Too
There is a quieter cost that compounds over time. Legacy systems are hard to learn. New staff take weeks or months to become productive because the system is unintuitive, the process is full of exceptions, and the only training is watching someone else.
Every new hire pays this cost. In a growing business, it becomes a permanent drag: you are always onboarding someone into a system that fights them. Modern tools are not automatically better, but a well-built replacement can turn a two-month learning curve into a two-week one, and that difference repeats with every person you hire.
How to Decide Whether to Modernize
Once you have the numbers, the decision gets clearer. Add up the annual workaround tax, put a realistic figure on the failure risk, and factor in onboarding drag. Compare that total to the cost of fixing or replacing the system. A few honest principles help here:
- Do not rewrite what works quietly. If a system is stable, cheap to run, and understood by more than one person, leave it alone.
- Target the expensive workaround, not the whole system. Often you can eliminate 80 percent of the cost by fixing one integration or one broken report, without touching the rest.
- Price the risk before it prices you. If a single point of failure could halt your business, that alone can justify action regardless of the workaround math.
Modernization is an investment decision, not a technology fashion choice. It only makes sense when the measured cost of staying put clearly exceeds the cost of changing. The mistake I see most often is the opposite of over-spending: businesses tolerate a growing hidden cost for years because it never arrived as a bill. Understanding what genuinely drives the price of a replacement helps you judge whether it is worth it, which I broke down in What Actually Drives the Cost of Custom Software.
Practical Takeaway
Before your next planning cycle, spend an afternoon measuring what your legacy systems actually cost you:
- List every manual workaround your team performs and estimate the hours, then apply the workaround tax formula.
- Identify every single point of failure, both the systems and the people, and be honest about the exposure.
- Add the onboarding drag for anyone you have hired in the last year.
Put a rupiah figure on the total. Only then compare it to the cost of fixing or replacing. You may find the system is cheaper to keep than you feared, or you may find you have been quietly paying for a rewrite every year without getting one. Either way, you will finally be deciding with numbers instead of instinct.