How Europe's largest local authority turned a working system into a catastrophe
Birmingham City Council had a working SAP system for twenty years. Then they decided to modernize. What followed is the most expensive municipal IT failure in history — and every warning sign was visible from the start.
In 2018, Birmingham City Council — the largest local authority in Europe, serving over one million citizens — embarked on what should have been a straightforward ERP modernization. They had run SAP for twenty years. The system worked. It wasn't elegant, and it wasn't modern, but payroll went out, invoices got paid, and the books balanced.
The council approved a £19 million budget to migrate from their aging on-premise SAP system to Oracle Fusion Cloud. The vendor promised a modern, cloud-native platform that would reduce costs, improve efficiency, and position Birmingham for the next decade. The timeline was aggressive but achievable — eighteen months to full go-live.
"The project was doomed before the first line of code was written. Not because the technology was wrong, but because the governance structures ensured that no one independent was watching."
Four years later, the project had consumed £216 million — more than ten times the original budget. Citizens couldn't access basic services. Council workers went unpaid for weeks. Housing benefit payments failed, leaving vulnerable families without support. The Oracle system, rather than improving operations, had created chaos at every level of city government.
The Birmingham failure wasn't a technology failure. Oracle Fusion is a capable platform used successfully by organizations worldwide. The failure was structural — a perfect storm of conflicted advice, absent oversight, and governance designed to rubber-stamp rather than challenge.
Three factors converged to create the catastrophe:
First, there was no Phase Zero. Birmingham skipped the foundational preparation phase entirely. They moved directly from "we need a new system" to vendor selection without documenting current processes, assessing data quality, or building internal readiness. The council didn't know what they needed because they'd never done the work to find out.
Second, the advisory structure was conflicted. The consulting firm advising on vendor selection had existing commercial relationships with the vendors being evaluated. When you're choosing between platforms and your advisor profits from one of the choices, the advice is structurally compromised — regardless of the individuals' integrity.
Third, governance was designed for compliance, not challenge. The project board included vendor representatives alongside council members. Decisions were made in rooms where the people implementing the system were also evaluating their own progress. There was no independent voice asking uncomfortable questions, no outside perspective measuring reality against promises.
At the same time Birmingham was spiraling, a mid-sized American city faced an identical challenge: replace an aging ERP system that had served them for decades. Same scale of complexity. Same vendor landscape. Same political pressures to modernize quickly.
But this city did something Birmingham didn't. They started at zero.
They spent 100 days in Phase Zero — documenting every process, mapping every integration, assessing every data quality issue, and building genuine stakeholder alignment before a single vendor entered the conversation. They engaged an independent advisor with no vendor relationships. They established governance that separated advice from delivery...
Enter your email to unlock the full chapter — including the complete Birmingham forensic analysis and the framework that could have prevented it.
No payment required. Chapter 1 + 5 ready-to-use templates sent instantly.