"It works and costs nothing": what your old application really costs to keep
In the budget, the old application costs nothing, yet it costs money every month. The hidden costs, when modernising pays off, and when to leave it alone.
Vlado Pandžić · Founder · Senior .NET architect
Published · 4 min read
In a company budget, the old business application usually costs nothing. No licence, no project, no line item. It has run for years and nobody touches it.
But having no line item doesn’t mean it costs nothing. The costs are just spread across other places: developers’ hours, outages, people working around it. Nobody adds them up, so nobody sees them.
Where the costs hide
- Slower changes. Every new feature takes twice as long as it should, because nobody dares touch the code, and every change breaks something. Why that happens is covered in the article on technical debt.
- Outages and firefighting. The application stops, half the company waits, and one person hunts for the problem.
- Manual workarounds. What the application can’t do, people do by hand: exporting to Excel, retyping, checking.
- Old servers, licences and remote access. Hardware that can’t be switched off, extended security updates, Citrix for working from home. The deadlines coming up are covered in the article on Windows Server 2016.
- Dependence on one person. Only one person knows how it all works. While they are there, the cost is invisible. When they leave, it shows all at once, as in the article on taking over an application.
- Security. Old .NET Framework, old packages, passwords in configuration files. The cost is zero, until something happens.
A worked example
An example, not a promise: a company with one old business application used daily by about fifteen people.
| Item | How it is calculated in the example | Per year |
|---|---|---|
| Slower changes | 30 extra development days at €400 | €12,000 |
| Outages and firefighting | 6 outages × 4 hours × 15 people × €25 | €9,000 |
| Manual workarounds | 5 hours a week × 46 weeks × €25 | €5,750 |
| Old server, licences, remote access | hardware, updates, Citrix | €3,000 |
| Total | almost €30,000 | |
| Dependence on one person | invisible until that person leaves | risk |
| Security incident | invisible until it happens | risk |
In this example, the application “that costs nothing” costs almost €30,000 a year, and that is before two items that can be more expensive than all the others combined.
When modernising pays off, and when it doesn’t
It is only fair to say: some old applications should be left alone.
| Decision | When it makes sense |
|---|---|
| Leave it as it is | It is stable, rarely changes, few people use it and it isn’t exposed to the internet |
| Modernise step by step | It is core to the business, changes often and many people use it |
| Replace it with an off-the-shelf product | The process is standard (CRM, payroll, accounting) and your application adds nothing special |
The calculation is the same as for any investment. If modernising halves the hidden costs in the example, that is about €15,000 a year. A modernisation costing €30,000 pays back in two years, and after that the saving remains, with less risk.
Modernising doesn’t mean rewriting everything. The safest route is almost always step by step: first moving to current .NET, then piece by piece. What that looks like for desktop applications is covered in the article on moving WinForms and WPF applications to the web.
How to get your own numbers
- ChangesEstimate versus actual time
- OutagesHow long and how many people
- Manual workHours a week around the app
- PeopleWho knows how it works
For one month, write down four things:
- Changes. For each change, how long it was estimated to take and how long it actually took.
- Outages. Every time the application stops: for how long, and how many people couldn’t work.
- Manual work. How many hours a week people do something by hand because the application can’t.
- People. Who knows how the application works. If the answer is one person, that is your biggest line item.
After a month you have your own table, and the decision is no longer a feeling but numbers.
How we work
This is exactly what we do: we review the old .NET application, put together a calculation of what it costs you today and recommend one of the three decisions. If modernising pays off, we do it step by step, piece by piece, without downtime, with Blazor and current .NET. And if it doesn’t pay off, we tell you that too. The first step is a free 30-minute call.
This article is general information only, not legal, tax, financial or other professional advice. Scenarios, examples and calculations are illustrative. Terms of use and disclaimer.