Every business runs on a set of assumptions nobody has recently checked. Some are small — a belief that a particular vendor is worth the premium, that a certain time of year always slows down, that a specific channel is the best source of new customers. Some are larger, shaping how significant parts of the budget get allocated year after year. Most of these assumptions started as reasonable observations. Few of them have ever been tested directly against the outcomes they’re supposed to explain.
This isn’t unusual. It’s how most businesses operate, because testing an assumption properly takes more effort than simply repeating it, and because an assumption that’s been repeated confidently for long enough starts to feel like a fact.
Why assumptions accumulate
An assumption forms the first time it’s a reasonable explanation for something observed. A cost goes up, revenue happens to go up around the same time, and a connection gets assumed between the two. Nobody sets out to create an untested belief — it just becomes the accepted explanation because no one had a reason to question it at the time, and once it’s accepted, it tends to stay unquestioned indefinitely.
Over years, a business accumulates dozens of these. Some are true. Some were true once and no longer are, because the business or the market around it has changed. Some were never true at all — a coincidence of timing mistaken for a cause. The trouble is that all three look identical from the inside. An untested assumption that’s wrong doesn’t announce itself as wrong. It simply continues being repeated as if it were established fact.
What a measurement actually adds
A measurement is different from an assumption in one specific way: it can be wrong, and the business finds out. An assumption is safe from being disproven because it’s never actually checked. A measurement puts a real claim on the table — this cost is or is not statistically connected to this outcome — and either the data supports it or it doesn’t.
This is a slightly uncomfortable shift for most organisations, because assumptions tend to protect whoever proposed them, while measurements don’t protect anyone. But it’s also the only way a business finds out which of its long-held beliefs are actually true, which are outdated, and which were never accurate in the first place.
How this applies beyond costs
Measurement in this sense isn’t limited to spend. It applies to any belief a business operates on without having tested it — assumptions about which customer segment is most valuable, which internal process is actually the bottleneck, which change genuinely improved a result versus which change simply happened around the same time something else improved. The specific method changes depending on what’s being tested, but the underlying discipline is the same: state the claim clearly, check it against real data over a meaningful period, and accept whatever the result says, even when it contradicts what everyone already believed.
Cost is often the easiest place to start, because the data — what was spent, and what happened to the outcome it was meant to affect — usually already exists in a business’s own financial records. This is the basis of Cost Impact Analysis: taking a specific cost-to-outcome assumption and testing it directly, rather than accepting the explanation that’s always been given.
The habit worth building
No business can test every assumption it operates on at once, and that isn’t the goal. The goal is building the habit of treating a long-held belief as a hypothesis rather than a fact, and being willing to check it when the decision attached to it is significant enough to justify the effort. Over time, this shifts a business from being run on a set of inherited explanations to being run on a set of things that have actually been verified — a small distinction in principle, and a large one in practice.
If you want to see where this kind of testing would apply first in your own business, a short readiness assessment at danieldiosi.com is a useful place to start.
