We need to modernize our 12-year-old legacy system — phased approach or complete rewrite?

Three years ago we opted for a ground-up rebuild for a logistics platform of similar size. They quoted us 6 months; it took a full 16, and the bill ended up nearly triple. First week we went live, discount logic was broken across the board and we misbilled hundreds of clients. Doing a rip-and-replace on a functioning core system is a massive gamble.

The standard blueprint for a phased rollout: 1) Map out legacy database relationships and isolate independent domains, 2) Put a reverse proxy / router in front to direct traffic, 3) Carve out non-critical, read-heavy services first, 4) Save mission-critical flows like ordering and billing for the absolute final phase.

In my consulting career, 4 out of 5 enterprise rewrites started from scratch overshot their target delivery dates by at least 6 months. By contrast, gradual refactoring kept budget drift manageable and none of them suffered critical operational downtime. Your internal dev's proposal is anchored in a much more realistic reality.

A phased migration also has a hidden cost: for 18 months, your team will have to juggle two different architectures, two different data models, and the synchronization between them. If your team is small, this dual-maintenance burden can burn out your developers and grind everything to a halt anyway.

Take the first step right away to test your developer's plan. Before signing a 110,000-dollar contract with the agency give your developer 2 months to decouple the simplest module. If they can set up the new service and get data sync running smoothly, shelve the agency's offer for good.

Any agency claiming they can completely rewrite a complex 12-year-old ERP in 8 months hasn't yet seen the undocumented custom customer discounts and accounting rules buried inside; on go-live day, you'll watch that 8 months easily turn into 20.

data migration really is the biggest headache... you cant migrate corrupt or missing records from the legacy database to a newly built system without cleaning them up first. with a phased rollout you get to validate live data step by step, with a big bang migration you get zero second chances.

Whichever approach you choose, I recommend including data integrity guarantees and a rollback procedure for every stage in the contract terms. It is essential to remember that operational continuity always takes precedence over software architecture.

I think differently. If acceptance criteria aren't written, when the work is done is open to debate.

The real issue isn't the number, but what it's based on. If you have questions, write them; I'll answer as best I can.

I'm writing this so you don't make the same mistake. When you try to change everything at once, nothing settles.

If it's your first time, start small; scaling comes later. If you post the result here, it will help others too.

I felt relieved reading this answer, so it's not just me. If you get three different answers on a topic the question was asked wrong.

When making a decision, first look at what data you have on hand. If you have questions write them; I'll answer as best I can.

I have a question, don't want to go off-topic though. If code ownership isn't in the contract, you have no bargaining power when leaving.

Trying to do this alone is the most expensive way. I'm also curious if anyone does it differently.

Let me summarize the topic since several different answers were given... I mean mistakes made on the legacy system modernization side are usually reversible but expensive.

We've heard this a lot but it never happened like that for us. Access credentials should be opened in the company's name, not personal accounts.

This is my opinion, I'm not claiming it's absolute truth.

I think differently. Taking notes for two weeks yields better results than a six-month estimate.

When making decisions, write down the worst-case scenario too, not just the best. Proven by experience.

How did you solve this? When you try to change everything at once, nothing settles.

That's all, sorry if I went on too long.

The discussion got scattered, let me summarize. Hasty decisions become decisions you have to fix six months later.

Just leaving this note, it might be useful.

Saved.