Everyone's telling you to rewrite everything. They're almost certainly wrong.

There's a particular kind of consultant who loves a big-bang rewrite.

New platform. Clean slate. Modern architecture.

It sounds great in a presentation.

It sounds less great when you're 18 months in, over budget, and your users are asking for the old screens back.

The graveyard of failed rewrites is well-documented. Not because the technology was wrong. Because the business logic - the real logic, the stuff that took 20 years to accumulate - wasn't properly understood before anyone touched a line of new code.

The 30-year-old COBOL your business runs isn't a liability.

It's a record of 30 years of decisions. Pricing rules. Exception handling. Edge cases that only exist because something went wrong in 2003 and someone fixed it quietly and moved on.

That logic has value.

The question isn't 'how do we replace this?' It's 'how do we understand it well enough to move it safely?'

Sometimes that means a phased migration. Sometimes wrapping it in an API while you plan. Sometimes leaving the core alone and modernising around it.

What it almost never means is a 12-month rewrite on a fixed price.

We're not in the business of selling expensive plans that look impressive on slide 4.

We're in the business of figuring out what's actually there - and what actually needs to happen.