A business can spend weeks writing a customer service manual: how to greet, what to ask, how to close a sale, how to handle a complaint. That document is the necessary foundation: it defines what's expected, aligns teams, and gives every location the same starting point. But the standard, on its own, describes an expectation — it doesn't confirm how it's actually being executed, day to day, at every point of contact.
That confirmation depends on something else.
The origin of an idea that became a cliché
There's a phrase that circulates constantly in management circles: "what cannot be measured cannot be improved." It's commonly attributed to Lord Kelvin, the Scottish physicist and mathematician William Thomson, one of the most influential scientists of the 19th century, known for establishing the temperature scale that bears his name and for major contributions to thermodynamics and electricity.
Worth clarifying: that exact phrase doesn't appear in his writings. What Kelvin actually said was a related but distinct idea: that you only truly know something when you can express it in numbers, and that without that measurement, your knowledge remains vague and unsatisfactory. Over time, that idea got simplified and popularized in business language into the phrase repeated today — worth noting that similar variants also circulate attributed to other management figures, like Peter Drucker, which reinforces that this is more of a widely diffused idea than a quote with confirmed single authorship.
Setting aside the exact attribution, the underlying idea is useful for thinking about service standards. A written protocol describes an expectation. But until that expectation translates into something measurable — how it's actually being executed, in what proportion, with what variation between teams or locations — a business has no way of knowing whether the standard is working or just paper filed away in a drawer.
Why a standard without measurement tends to erode
A protocol that no one checks tends to get interpreted more and more loosely over time. Not out of bad intent from staff, but because without periodic verification, each person ends up adjusting the standard to their own judgment, and those small adjustments accumulate until real execution drifts pretty far from what the original document described.
This is different from deliberate non-compliance. It's more of a gradual drift, nearly invisible day to day, that only becomes evident when someone compares current execution against the original standard — something that rarely happens without a sustained measurement mechanism.
What measurement means in this context
Measuring compliance with a service standard isn't simply asking staff whether they're following the protocol. It requires a source of information independent from self-assessment: audits, mystery shopper evaluations, review of operational indicators, or some combination of these, applied on a defined cadence and with consistent criteria across locations.
Without that measurement, a business only has the perception that the standard is being followed — and that perception, as tends to happen with internal oversight, is usually more optimistic than reality.
From measurement to improvement
Measuring alone isn't enough either. Kelvin's idea, in its popularized version, tends to stop halfway: measuring is necessary, but it's not the same as improving. A business can measure standards compliance religiously and still not improve anything, if those results never turn into decisions — adjusting training on a specific point, fixing a process, following up on a location that deviates repeatedly.
Measurement is what makes the problem visible. Improvement happens afterward, when someone acts on what that measurement shows.
FAQ
Why doesn't a well-written service standard guarantee it will be followed?
Because a standard describes an expectation, but doesn't include a mechanism to verify its execution. Without that verification, interpretation of the standard tends to vary between people and teams over time.
Who actually said "what cannot be measured cannot be improved"?
That exact phrase doesn't have a confirmed author. It's commonly attributed to Lord Kelvin, though what he actually stated was a related idea about the connection between measurement and precise knowledge, not that literal phrase.
What's the difference between measuring standards compliance and improving service?
Measuring shows whether the standard is being followed or not. Improving requires an additional step: turning those results into concrete decisions, with owners and deadlines, that correct what the measurement detected.
How often should compliance with a service standard be measured?
It depends on the size of the operation and how much execution varies across locations, but a sustained cadence over time is what allows a business to tell whether a deviation is a one-off or turning into a pattern.
