How much does it cost to modernise a legacy system?

Every legacy conversation eventually resolves into three figures: a one-off cost to get the system under control, a monthly cost to keep it looked after, and a programme cost to actually modernise it. They are separable on purpose. Plenty of businesses stop after the first one, or the first two, and that is a perfectly good outcome when the system fundamentally does its job.

This is the phase that makes everything else possible, and it is the same regardless of what you eventually decide to do with the system. The code goes into source control, recovered from the live server if the repository is missing or being withheld. Every credential the system depends on gets found and inventoried. The backups get verified by restoring one, which is the only kind of backup verification that counts. Error monitoring goes in, a repeatable way to deploy goes in, and an audit works out what the system actually contains, so every later decision is priced from evidence rather than guesswork.

The band is wide because the starting states vary so much. A system where you hold the code and the passwords sits at the £4,000 to £8,000 end: the base work is four to seven engineering days, plus a little for each problem the audit has to unpick. A system where the developer has vanished, the code lives only on the server, nothing is documented and the credentials are scattered across the inboxes of people who left years ago runs £10,000 to £16,000, because each of those adds one to three days of recovery work. We have written up exactly what happens during this phase if you want the detail behind the line items.

Two to four weeks, typically. At the end of it you have a system that is safe to change, and, just as valuable, a priced and prioritised list of what changing it would involve.

Once a system is under control, support is what stops it sliding back. Ours starts at £250 a month: monitoring that tells us about failures before your staff do, security patching as fixes are released, and someone on the end of the phone who already knows the system when something needs attention. The figure rises with how critical the system is. Where losing a day of data would be catastrophic rather than painful, the checks run more often and the number sits higher, typically £350 to £450.

What a proper arrangement covers, and the signs yours is wearing out, is a post of its own: legacy application support, what good looks like.

We estimate modernisation by the areas a system covers, because that is how the work actually divides. Jobs, quoting, stock, invoicing, reporting, a customer login: each distinct area runs roughly four to nine engineering days to bring up to date, scaled by two things. Age, because a fifteen-year-old codebase fights back harder than a five-year-old one, adds up to 40 per cent. Condition does the same: no tests, no documentation and years of accumulated workarounds all slow the work down, and a system with most of those runs 20 to 45 per cent over the base bands.

Put together, a four-area system, ten to fifteen years old and in the rough-but-running state most of them arrive in, lands between £20,000 and £50,000. The important part is not the total but the shape: the work runs in slices over months, each slice scoped, priced and approved before it starts, with the system live throughout. You control the pace by controlling the budget, and you can stop at any point holding a better system than you started with. Compare that with a rebuild, where stopping halfway leaves you with nothing but invoices.

Rebuilding the same four-area system from scratch starts with fifteen to twenty-five days of foundations, the server, the database, accounts and logins, an admin back office, before any of your four areas exists again. Each area then runs six to twelve days to rebuild properly. That puts the same system at £30,000 to £60,000, and the money is the smaller half of the argument. A rebuild throws away every rule, exception and hard-won fix that years of trading have baked into the old system, and the business rediscovers them one incident at a time, in production.

There are systems where replacement is still right, and we set out the decision properly in fix, modernise or replace. If you are leaning that way, price it with the software cost calculator rather than taking the comparison band above, because a rebuild deserves a proper estimate of its own.

In order. First, what you still hold: code in your own source control against code that has to be recovered from a server or negotiated back from a departed developer is the difference of several thousand pounds before anything else starts. Second, the stack gap: a system on versions that stopped receiving security patches needs that dealt with first, and the longer the gap the more careful the climb. Third, the number of distinct areas, which scales the programme almost linearly. Fourth, integrations, because every system the old application talks to is a conversation that has to keep working while everything around it changes. And a mobile app alongside the web system adds around 15 per cent to whatever it sits on top of.

The one that surprises people is data. Years of real-world use produce duplicate customers, half-finished records and fields repurposed for things they were never meant to hold. Cleaning that up is rarely quoted by anyone and frequently the messiest single piece of the work, which is why our audits look at the data as hard as they look at the code.