The Bill That Grows Without a Decision: Who Is Committing Your Money?
Somewhere in most organisations is a monthly invoice nobody approved. It is large, it is growing, and every line on it is legitimate. Almost nothing else an organisation buys arrives that way, and the controls were never built for it.
Somewhere in most organisations is a monthly invoice that nobody approved. It is large, it is growing, and every line on it is legitimate.
That combination is unusual. Almost nothing else an organisation buys arrives that way.
1. This Was the Deal, and It Was a Good One
Consumption pricing was not imposed on anybody. Organisations chose it for sound reasons, and the great majority would choose it again tomorrow.
It removed the need to buy capacity for a peak that happens twice a year. It cut lead times from months to minutes — and turned a capital commitment into something that could go down as well as up.
None of that has stopped being true. The invoice is simply the consequence of a decision that was right at the time and remains right now.
2. What Changed Was Who Commits the Money
The part nobody examined was who does the committing, and how often.
An engineer choosing an instance size, adding a region for resilience or leaving a test environment running over a weekend is making a purchasing decision — though very few would describe it that way, and neither would their manager.
They are doing the job properly. The purchase is a side effect of a technical choice — and technical choices are precisely what that person is employed to make.
3. The Approval Model Stayed Where It Was
The financial controls, meanwhile, did not move at all. Thresholds, purchase orders, a signature above one figure, a business case above a larger one.
All of it is built around a purchase event — a moment when somebody asks and somebody else agrees. Consumption spend has no such moment. It has ten thousand small ones, and not one of them crosses a threshold.
So a framework designed to govern commitment now governs almost none of the commitment actually happening.
4. The Bill Is a Report, Not a Decision
What finance receives at the end of the month is not a proposal. It is a record of things that have already occurred.
By the time the number reaches anybody with authority over it, the money has gone. There is nothing left to approve — only something to explain.
Which is why these conversations have their particular flavour. Nobody is defending a decision, because nobody is conscious of having made one.
5. Cutting It Afterwards Is the Wrong Instinct
The usual response is an optimisation exercise. Somebody is asked to reduce the bill by fifteen per cent, and they generally can.
Much of what they find is real. Idle environments, oversized instances, storage nobody has opened in three years — the work is worth doing, and it pays for itself several times over.
But it treats the number rather than the mechanism, and the mechanism is left exactly as it was. Eighteen months later the same exercise gets commissioned again.
6. Not All Growth Is a Problem
This is where the argument usually overreaches. A rising bill is not evidence of anything by itself.
If the organisation is serving more customers, holding more data or running more of its business through the platform, the bill ought to be rising. A flat bill in a growing business would be the more troubling number by far.
The useful question is not whether it grew. It is whether anybody can explain why in terms a non-technical director would recognise.
7. What Actually Helps
The remedy is organisational rather than technical, and it costs considerably less than the optimisation exercise:
- who may commit spend without asking, and up to what figure
- which team’s budget each part of the bill belongs to
- what the expected shape of the bill is, so a rise shows up in days
- what gets switched off, by whom, once something is no longer needed
The third does more than the other three together, because it turns a monthly surprise into a variance somebody notices while it is still small.
8. Grant the Authority That Is Being Used Anyway
The honest conclusion is an uncomfortable one for a finance function. The people making these commitments are the right people to be making them.
They understand the trade-off between cost and resilience far better than a procurement threshold ever will. What they have not been given is the information, the budget line, or the explicit permission — and none of those is difficult to provide.
An organisation can grant that authority deliberately, or carry on having it exercised by default and finding out at the end of the month.
Final Thought
Most organisations govern the technology they buy a great deal more carefully than the technology they use — for no better reason than that one involves a signature and the other does not.
Which means the bill is not really a cost problem at all. It is a question about where authority sits, arriving once a month in the form of a number.