KEY TAKEAWAYS

  • Running a service on hunches means you only learn about problems once they are large enough to complain about.
  • You do not need hundreds of metrics; you need a few that genuinely reflect quality, speed and cost.
  • Measurement turns vague worries (“we feel stretched”) into specific, fixable facts (“Tuesdays are where it breaks”).
  • The point of a number is the decision it changes; a measure nobody acts on is just decoration.
  • Start small, capture consistently, and let the data redirect effort to where it pays.


Every service is being measured by someone. The only question is whether it is you, with deliberate numbers you can act on, or your customers, with their patience and their willingness to stay. When a team has no clear view of how the work is actually performing, it is not running without measurement; it is running on guesswork dressed up as experience. That works until it doesn’t. The honest position is this: if you are not measuring a service, you are guessing about it, and the most expensive problems are precisely the ones guesswork cannot see. The remedy is modest and entirely within reach. Pick a few measures that reflect the work, capture them the same way every time, and use them to decide where effort goes.


Why instinct quietly fails

Experience is a genuine asset, and a good operator can sense when something is off. The trouble is that instinct only registers problems once they are large enough to feel. A backlog has to become visible before it nags at you; a quality slip has to generate complaints before it lands. Everything below that threshold stays invisible, and a great deal of waste lives below that threshold.

Instinct also struggles with comparison. It can tell you that today felt busy, but not whether today was busier than the same day last month, or whether “busy” is the new normal. Without a number to anchor it, every judgement is relative to the last thing you happened to notice, which is a shaky foundation for decisions about people, money and priorities.


What counts as guessing

Guessing is not only the dramatic “we have no idea.” It is the quieter habit of running on proxies that feel like data but aren’t. “The team seems stretched” is a feeling. “We’re probably hitting our turnaround targets” is a hope. “Customers seem happy enough” is the absence of bad news, which is not the same as good news.

These proxies are seductive because they require no effort to produce and rarely get challenged. But each one is a place where reality and belief can drift apart for months without anyone noticing, until the drift is large enough to become a crisis. The crisis then looks sudden. It almost never is.


The cost of finding out late

The real penalty for not measuring is timing. Problems do not announce themselves on the day they begin; they announce themselves on the day they become unbearable, by which point they are larger and more expensive to fix. A small, early signal that turnaround times are creeping up is cheap to act on. The same problem discovered through a wave of complaints is not.

Late discovery also removes your options. Caught early, a rising workload can be smoothed with a small adjustment. Caught late, it forces blunt, costly responses: emergency overtime, hurried hiring, apologies to customers who have already decided to look elsewhere. Measurement does not prevent problems. It buys you the time to deal with them while they are still small.


You need a few good numbers, not a wall of them

The fear that puts people off measurement is the image of a sprawling dashboard nobody understands. That fear is well founded, but the answer is not to avoid measuring; it is to measure less, and better. Most services can be understood through a small set of questions. How much work is coming in? How long is it taking to deal with? How often do we have to redo it? How do the people we serve rate the result? Are we keeping pace with demand, or falling behind?

Answer those honestly and you have most of what you need. A handful of well-chosen measures that everyone understands will always beat a hundred that nobody trusts. The goal is not completeness; it is a true enough picture to make better calls than you would by instinct alone.


Make it specific, then make it fair

The power of a number is that it turns a vague worry into a precise, fixable fact. “We feel stretched” cannot be acted on. “Wednesdays carry forty per cent more work than any other day, and that is where errors cluster” can be. Once a problem has a location, a size and a pattern, it stops being a mood and becomes a task.

Two cautions keep the numbers honest. Measure things consistently, or you will end up comparing figures that were counted differently and drawing false conclusions. And measure outcomes that matter to the people you serve, not just outputs that are easy to count. Volume processed is comforting; whether the customer got the right result, quickly, is what actually matters. Choose measures that reward the behaviour you want, because people will optimise for whatever you put on the board.


The number is only worth the decision it changes

A metric exists to change a decision. If a figure cannot alter what you do next, it is decoration, and decoration is a cost: it takes effort to produce and attention to read, while teaching everyone that numbers are for show. Before adopting any measure, ask the blunt question - if this moves, what will we do differently? If the answer is “nothing,” drop it.

This test keeps reporting lean and meaningful. It also changes the culture around data. When numbers reliably lead to action, people start trusting them, surfacing them early and using them to argue for sensible changes. When numbers are just reported and filed, they become a ritual, and rituals are easy to fudge. A small set of measures that genuinely drives decisions is worth more than any quantity that simply gets circulated.


Start before you feel ready

The most common mistake is waiting for the perfect measurement system before measuring anything at all. You will never feel fully ready, and the cost of that delay is more months run on guesswork. Begin with what you can capture now, even if it is rough. A simple count, recorded consistently, will teach you more than an elaborate plan that never launches.

Once the first few numbers are flowing, the system improves itself. You will quickly see which measures are useful and which you ignore, and you can refine from there. The aim is not a finished masterpiece; it is to replace the worst of the guessing with evidence, then keep getting clearer. The day you can see your service in numbers is the day you stop being surprised by it - and start, quietly and steadily, getting ahead of it.