In thirty years of enterprise delivery, we've watched modernisation programmes fail. Almost none of them failed on technology.
The stack is rarely the problem. SQL Server, Azure, a sensible architecture — the engineering is well understood and the patterns are proven. What sinks these programmes is everything around the code. Three failure modes show up again and again.
No single owner
When a programme is "owned" by a steering committee, it's owned by no one. Decisions stall, scope drifts, and the timeline slips a fortnight at a time until the business case no longer holds.
Financial misalignment
The technology team is measured on delivery; finance is measured on cost. When those two are never reconciled, the programme gets cut at exactly the wrong moment — usually just before the benefits land.
No definition of done
"Modernise the platform" is a direction, not a target. Without a hard definition of the end state, teams rebuild indefinitely and the old system never actually gets switched off.
The pattern
None of these are technical problems. They're governance, ownership, and financial-alignment problems — which is precisely why a strong technical team alone often isn't enough to get a transformation over the line.
The technology is the easy part. It usually was.