# Where did the saving in the business case go? > A saving written at its best case, and owned by nobody, shrinks and is forgotten. How a £3m business case fades, and what to fix before approval. Source: https://barkingcat.ai/thinking/where-did-the-saving-go/ ## Thinking · November 2026 · Barking Cat series In the cases I have seen, it was never as big as the case said, and what there was faded once nobody was looking after it. A business case for a saving is written at its best case. The saving then depends on the business doing something, sometimes repeatedly, that nobody has been asked to own: renegotiating a rate, updating a configuration, checking a line every month. When that work slips, the saving shrinks, and when nobody owns the line it lands on, the shrinking goes unnoticed until the project is simply forgotten. Delivering a project and banking its benefit are two different events. A project plan tracks the first. Unless the approval names the line the saving lands on, the person who owns it, the baseline it is measured against and the work that keeps it alive, nothing tracks the second. ## How a saving disappears The numbers below are an illustration of how a cost saving shrinks and then vanishes from view. Suppose £400m of card payments a year, and a project to route each payment to the customer’s own bank, where the average fee falls from 1.5% to 0.75%. That is £3m a year in the business case. Now apply two things the case may not have priced. Routing works only where the business has an agreement with the customer’s bank. If 70% of the volume can be routed, the saving is £2.1m. Banks offering their best rates to win volume may offer them for short periods, so the rates have to be renegotiated and the routing reconfigured every time they change. If the configuration falls behind the negotiation for half the year, that half saves half as much, and the saving is about £1.6m. The more often the rates change, the more of the saving depends on someone keeping up. The last line does not depend on the first three. Card fees can be booked as a central payment cost, outside the P&L the team that ran the project reads. A £1.6m saving booked there, against fees that also move with sales volume and the mix of cards, looks like ordinary variation. ## Three explanations, and the one that leads The first is that the case was optimistic: written at its best case, because a business case is written to win approval. The second is that the saving depended on work the business could not sustain, such as a renegotiation every week or month and a configuration change after each one. The third is that nobody owned the line the saving landed on, so nobody noticed as it shrank. I think the third leads, because it lets the other two run unchecked. An optimistic case can be corrected and a demanding routine can be resourced, but only if someone is watching the number and is accountable for it. What would prove me wrong is a saving with a named owner, a baseline and a monthly read that still faded unnoticed. The pattern is from my own record: a payment routing project with £3m a year in its business case, which I agreed to and which was delivered. The case was the best case, and the savings never got close to it. The rates were renegotiated with the banks weekly or monthly, the configuration updates did not keep pace, and in the end the project was more or less forgotten. ## What to do before the next approval Name the line in the P&L the benefit will land on, and the person who owns it, as a condition of approval. For a cost saving, agree the baseline: for a payment project, fees as a percentage of card volume, by bank, for the months before the change. Where the saving depends on repeated work, such as renegotiating rates or updating a configuration, name who does it, how often, and how quickly a change has to follow a negotiation. A saving that needs more discipline than the business has is a smaller saving, and the case should say so. The cheaper middle route, if the business is not ready to restructure its reporting, is a single page: the case’s number, the baseline, and the monthly actual, sent to the owner every month for a year. ## How you will know The measure is the cost line against its baseline, every month. For a payment project, the guardrail is the approval rate on card payments, because a change that saves fees but declines more customers has not saved anything. Read it after three months and again after twelve. Agree the rule at approval: if the saving is below half the case’s number after twelve months, the business finds out why before it approves the next project on the same logic. ## The decision The decision for the board is whether a business case can be approved without a named line, an owner and a baseline. The question for Monday is one the finance team can answer this week: for the last three cost-saving projects we approved, which line did each saving land on, and who can show it? This piece rests on one project and an illustration. It cannot tell you how much of a typical business case’s saving survives its first year, and I have not found anyone who has measured it. Related: [The Barking Cat problem](https://barkingcat.ai/thinking/the-barking-cat-problem/); [Ecommerce KPI Framework](https://barkingcat.ai/thinking/ecommerce-kpi-framework/). ## About the author Steve Webster is an independent adviser to the boards and investors of UK retail and consumer businesses from £20m to £1bn: an operator’s judgement on ecommerce, technology and AI, before the board commits the capital. Formerly Global Ecommerce Director at Barbour, with 27 years running retail P&Ls up to £240m. ## Contact steve@barkingcat.ai Barking Cat Advisory Limited. Registered in England and Wales, company number 17338003.