“The system is 15 years old, nothing can be done.” Really? In my career, I’ve seen dozens of systems that were declared dead – and are still running productively today. Better than before.
IT restoration is the art of making the best of what you have. Not because it’s sexy, but because it’s often the smarter decision.
Why Not Just Buy Everything New?
I hear this question often. The answer is usually: money, time, risk.
Money: A new ERP implementation for a mid-sized company quickly costs EUR 500,000 – and up. Hardware refresh including cloud migration: similar ballpark. Upgrading the existing system? Often a fraction of that.
Time: New systems take time. Evaluation, selection, implementation, training, parallel operation. Easily 18-24 months pass. A targeted restoration? 3-6 months, sometimes less.
Risk: Every new system brings new problems. Interfaces that don’t work. Processes that aren’t mapped. Employees who rebel. The known system has quirks – but you know the quirks.
What Does IT Restoration Actually Mean?
IT restoration is not a marketing term for “we’re doing nothing.” It’s a systematic approach:
1. Inventory: What Do We Actually Have?
Surprisingly often, nobody knows exactly what’s running. Servers that were forgotten. Applications that “someone installed once.” Interfaces that nobody documented.
The first step: A complete inventory. Hardware, software, licenses, dependencies.
2. Analysis: What Do We Actually Need?
Not everything that runs is needed. I’ve seen:
- Three different CRM systems in the same company
- Servers that haven’t received requests in years
- Software for which nobody can find the license anymore
What’s not needed can go. That simplifies everything.
3. Stabilization: Securing the Foundation
Before modernizing, the existing must run stably:
- Apply patches (yes, even those from five years ago)
- Set up and test backups
- Implement monitoring
- Create documentation
4. Optimization: Getting More Out of It
Now it gets interesting:
- Performance tuning: Often it’s enough to find the right configuration
- Memory/RAM upgrade: Amazing what a few GB more RAM can do
- SSD instead of HDD: The classic with the biggest effect
- Virtualization: Migrate old physical servers to modern VMs
- Containerization: Package legacy applications in containers
5. Modernization: Step by Step into Today
Not everything at once, but targeted:
- Modernize frontend, keep backend
- Put APIs in front of old systems
- Use cloud services for certain functions
- Gradual migration instead of big bang
Three Real-World Examples
Case 1: The Immortal Inventory System
Starting point: DOS-based inventory system from the 90s. “Must be replaced,” said all the consultants.
Problem: 20 years of business processes had grown around this system. The employees knew every key by heart.
Solution: Terminal emulator on modern PCs, virtualization of the DOS system, API bridge for the webshop. Cost: EUR 25,000. A replacement would have cost EUR 400,000.
Today: System has been running stably for 8 years – we made it Windows 11 ready again last year.
Case 2: The Server from the Basement
Starting point: 12-year-old physical server, last update: unknown. “Nobody touch it” was the motto.
Problem: No backup, no documentation, nobody knew exactly what was running on it.
Solution: Careful analysis (without restarting!), create image, migrate to virtual machine, integrate into modern backup. Only then: updates, hardening, monitoring.
Result: System now runs cleanly virtualized, fully documented, with daily backup.
Case 3: The Frankenstein IT
Starting point: IT landscape grown over 10 years. Three different backup systems, five different antivirus solutions, nobody had an overview.
Solution: Consolidation. One backup system, one endpoint protection, one central monitoring. Not through new purchases, but through cleaning up and choosing the best from what was available.
Result: Fewer licenses, less complexity, better overview. And incidentally: 30% lower IT costs.
When Is Restoration the Wrong Choice?
I’m not dogmatic. Sometimes new really is better:
- End of Life: When the manufacturer no longer provides support and security gaps aren’t closed
- Scaling problems: When the system fundamentally can’t grow
- Skills shortage: When nobody knows COBOL/RPG/AS400 anymore
- Business model change: When the old system can’t map the new process
- Economics: When annual maintenance costs exceed the new price
The Pragmatic Approach
My approach is always: First understand, then decide.
- What does continued operation cost? (realistic, not optimistic)
- What does replacement cost? (realistic, not optimistic)
- What are the risks? (both)
- What does management want? (sometimes “new” is a strategic statement)
Often the answer is: A mix. Restore the core system, modernize peripheral processes, use cloud services for new requirements.
Conclusion: Old Doesn’t Mean Bad
A 15-year-old system that runs stably and meets requirements is not a problem – it’s an asset. It only becomes problematic when:
- Nobody knows how it works anymore
- It’s no longer maintained
- Security suffers
- It slows down the business
All of this can be addressed – often without throwing everything out.
Do you have “old systems” that actually work well – but nobody dares touch them? Let’s take a look together. Often the solution is simpler (and cheaper) than expected.