From the public record · Birmingham · 2018–2026

Why the shape of a change matters as much as the system

A council finance and HR system budgeted at £19m in 2018 was reported in January 2026 to be heading for £144.4m. For smaller organisations the transferable lesson is about the size and shape of change.

Bar chart: a 2018 budget of £19 million against a projected cost of £144.4 million reported in January 2026. Timeline: 2018 start, April 2022 incomplete go-live, over £5 million on manual workarounds, re-implementation, completion forecast 2027/28.

What happened

Birmingham City Council began replacing its finance and HR systems in 2018 with a budget of £19 million. The new system went live incomplete in April 2022. Reported problems included a banking reconciliation customisation that did not work, difficulty producing auditable accounts, and more than £5 million spent on manual workarounds. As reported in January 2026, the projected cost had reached £144.4 million, with a re-implementation under way and completion forecast for 2027/28.

The pattern

When a large system is heavily customised and switched on at once, risk is concentrated in a single moment. If one critical function fails, the whole organisation falls back on manual work while it is fixed.

The same shape appears at every scale. A small charity moving its donor records, or a firm replacing its job tracking, can switch everything over on one weekend and discover on Monday that one essential report no longer works.

What a proportionate fix looks like

Comparison: a single big switch-on where one failure stops everything, against four small pieces, each useful on its own and each with a written manual fallback.
The same change, shaped differently.

Fairlead does not implement enterprise resource planning systems, and this article is not an argument that we could have done better. The lesson that transfers to smaller organisations is about shape: change in small pieces that are each useful on their own, prefer configuration to customisation, and make sure every automated step has a written manual fallback.

In practice that means a sequence such as: connect the forms first, then the reporting, then the finance export, with each piece live and checked before the next begins. If one stalls, the others still deliver.

Questions worth asking

  • Can this change be split into pieces that are useful on their own?
  • Which parts need custom development, and could configuration do the job instead?
  • For each automated step, what is the manual fallback, and is it written down?
  • What would we do on day one if the most important function did not work?

Fairlead Systems was not involved in the events described here. The account above is drawn only from the sources listed, which were last checked on 1 October 2026. If you spot an error, write to us and we will correct it.

Insights

More from the public record

Recognise this pattern?

Tell us what the process looks like today. We will say whether it needs fixing, and how small the fix can be.