What does legacy modernization cost?
Modernizing a legacy system usually costs SEK 1–15 million, and the strategy determines the price: phased replacement under the strangler pattern spreads cost and risk over time, while a big bang rewrite requires the full investment before any value is delivered. Waiting also has a price – security risks, retiring expertise, and integrations that grow harder.
Many Swedish organizations run on systems built fifteen or twenty years ago. They work – that’s why they’re still around – but maintenance grows more expensive, expertise thinner, and integrations more troublesome every year. The question is rarely whether the system should be modernized, but when and how. And it’s the choice of strategy that determines what it costs.
The strategy determines the cost
The SEK 1–15 million range is wide because it spans three completely different paths:
| Strategy | Typical cost | Risk profile |
|---|---|---|
| Encapsulation: an API layer around the system | SEK 1,000,000–3,000,000 | Low risk – buys time but doesn't solve the underlying problem |
| Phased replacement (the strangler pattern) | SEK 2,000,000–8,000,000 spread across phases | Limited risk per step, value delivered continuously |
| Full rewrite (big bang) | SEK 5,000,000–15,000,000 | High risk – the full investment before any value |
These figures apply to business-critical systems of typical mid-size; a small support system can land lower and a group-wide platform higher. What matters is the relationship between the paths – and that the cost isn’t a lump sum but a consequence of your strategy choice.
The strangler pattern: phased replacement as the default choice
The strangler pattern means a new system grows around the old one: one well-defined part at a time – the integrations, the customer portal, a single module – is rebuilt and takes over traffic, until the old system can finally be shut down.
The advantages are tangible in both budget and risk. The investment is spread across several budget years in phases that can each be evaluated. The business keeps running the whole time, and if a phase goes wrong, it’s that phase that gets rolled back – not the whole project. Hidden business logic gets handled piece by piece, where it can be verified against the old system’s actual behavior.
The cost of this pattern is that the whole effort takes longer in calendar time, and the old and new parts need to coexist through clear interfaces along the way. That’s a price worth paying in most cases.
Big bang: when everything changes at once
A full rewrite launched at a single point in time is appealing in its clarity – but the risk profile is the opposite. The entire investment sits before the first delivery, and while it’s being built, the old system keeps changing, so the target moves. The cutover itself is a risk point: one weekend when the whole organization switches systems at once.
Big bang can still be justified when the platform is completely dead, when the system is small and well-contained, or when a regulatory requirement sets a hard deadline. In that case, budget generous margins in both time and money – rewrites are systematically underestimated, mainly because old code carries more business logic than anyone remembers.
The cost of waiting
Pushing the decision forward is also a choice with a price tag:
- Security. Platforms that have left support get no security updates – every year raises the probability of an incident that costs more than the modernization would have.
- Expertise. The developers who know the technology grow fewer and more expensive, and when the people who built the system retire, knowledge disappears that’s written down nowhere.
- Integrations. New business tools, partners, and AI solutions expect modern APIs. Every custom-built bridge to the old system is a cost a modern system wouldn’t have had.
How to get started
Don’t start with a quote for the entire journey – start with an assessment. Which parts of the system are used the most, where are the risks, what needs to be replaced first, and what can keep running? Out of that comes a phased plan where each step has its own budget and its own value. Want help putting one together? Get in touch and we’ll look at your system together.
Frequently asked questions
What is meant by a legacy system?
An older system that still carries an important part of the business but was built on technology that's become hard to maintain, integrate, or hire for. Age itself isn't the problem – a ten-year-old system with good architecture and technology still in active use isn't legacy.
What is the strangler pattern?
A modernization strategy where new functionality is built around the old system and takes over part by part, until the old one can finally be shut down. Each step is well-defined and reversible, the system stays running the whole time, and the investment is spread across several budget years.
Why are big bang rewrites so risky?
Because all the value sits at the end: the business pays for months or years without getting anything, while hidden business logic and new requirements keep piling up on the old system in parallel. The cutover weekend itself is also a risk point, where the whole organization switches systems at once.
Can we modernize without replacing the system?
Partly. An API layer around the system makes it possible to build modern apps and integrations on top, which buys time and solves urgent needs. The underlying problems – old technology and shrinking expertise – remain, though, so encapsulation works best as the first phase of a bigger plan.
How do we know if it's time to modernize?
The warning signs are concrete: security updates stop, key people approach retirement, integrations with new tools pile up as one hack on another, and small changes take unreasonably long. When several signals line up at once, the cost of waiting is usually higher than the cost of starting.