Measurement Without Targets: Is It Possible?

Measurement Without Targets: Is It Possible?

The moment you set a target, the data starts getting gamed. The sentence sounds harsh, perhaps unfair to people who work in good faith. But it is not a verdict on intentions; it describes a behavioral law that operates whether we want it to or not. The instant a number is pinned to the wall and the team is told 'this is the target', the question quietly shifts from 'what is actually happening?' to 'how do we make the number look the way it is supposed to?'. And between those two questions lies exactly the distance through which reality slips away from our dashboards.

This article poses a question that sounds like heresy in a management culture that worships numeric targets: is measurement without targets even possible? Let us be clear — this is not a defense of chaos, nor a call to abandon ambition, which remains indispensable. But there is a fundamental difference between knowing where you want to arrive and turning every operational process into a pass-or-fail exam against a rigid number. The article makes the case for a concrete alternative called trend-based measurement: reading the direction and variation of performance over time, instead of judging it by pass or fail against a numeric threshold. It is a deeper extension of 'Measures, Not Indicators', pushing toward a further question: not where to place the number, but whether we need it at all.

How Gaming Is Born from the Womb of the Target

To understand why targets distort reality, trace the journey of the number from the moment it is fixed. When a strict target is set — 'close the request within 24 hours', 'keep the error rate below 2%', 'achieve 95% satisfaction' — we send one unmistakable signal: what matters is that the number lands on the right side of the threshold. And every human process responds to the signals it receives, not to the intentions we hold.

Here the distortion begins, in three successive layers. First, distortion of behavior: people rearrange their work around the threshold rather than its real value, advancing what is near the line and deferring what is far. Second, distortion of classification: when reality is hard to improve, reality itself is redefined — the complex case is closed and reopened under another number, or classified outside the counted scope. Third, and most dangerous, distortion of reporting: when even reclassification is not enough, the reported number becomes the product of an implicit negotiation between what happened and what must be said to have happened.

The painful irony is that each layer consumes genuine effort — effort that could have improved the work itself. The employee who spends an hour reclassifying a case to lift it off the overdue list has not improved the service by a single minute; they have only improved the number. The target turns from an incentive into a tax on honesty, paid twice: once in the wasted effort, and again in the decision later built on a number that does not mean what it thinks.

“A numeric target does not measure performance; it measures people's skill at pleasing the number.”

Goodhart's Law: When a Measure Becomes a Target It Ceases to Be a Measure

This is not a passing observation but a law, formulated by the British economist Charles Goodhart in 1975 in the context of monetary policy, before it proved to apply to almost every human measurement system. Its common formulation runs: when a measure becomes a target, it ceases to be a good measure. A good measure works silently at the margin of the process, capturing its reality without interfering. But the moment we attach a reward, a punishment, or a judgment to it, we draw it into the heart of the process as an active party, and it begins to change the very thing it was meant to observe.

The difference we must grasp is between a target and a measure. A measure says: this is what is happening. A target says: this is what should happen, and you will be held to account. The instant we add the second sentence, we corrupt the first. The essence of the problem is not the existence of numbers but their conversion into thresholds of pass and fail. The number itself is innocent; the threshold pollutes it.

Because Goodhart is sometimes misread, a clarification: the law does not say measurement is futile, nor that people are corrupt. It says any measure to which we attach pressure becomes a target for direct improvement instead of the reality it represents — and the harder the pressure, the wider the gap. The counterintuitive corollary: the cleanest data in an organization may be the data no one is held accountable for, because no one has an incentive to embellish it.

Deming: Management by Numeric Targets Is Management by Fear

W. Edwards Deming was no enemy of measurement; he was a man of statistics who spent his life teaching organizations how to read their data. That is why his warning against numeric targets carries special weight: it came not from someone who hated numbers, but from someone who loved them enough to refuse their abuse. In his book Out of the Crisis, Deming placed numerical quotas and management by objectives among the deadly diseases of management, not among the practices open to improvement.

Deming's core argument was that most of the variation we see in performance is produced by the system itself, not by the individuals working within it. When we set a numeric target and hold an individual accountable for it, we burden them with responsibility for something whose reins they do not fully hold — the performance of the system — then act surprised when they resort to a trick. The target does two bad things at once: it loads onto the individual what is the system's responsibility, and it converts data from an instrument of learning into an instrument of indictment.

The deeper result is what Deming called fear. An employee who knows that bad data will be used against them learns one lesson: do not surface bad data. That kills learning at its root — a problem that is not exposed is not solved. Measurement without rigid targets is not leniency toward performance; it is a precondition for people to dare to tell the truth, and there is no learning without truth.

Here the paradox at the heart of management by objectives comes into focus: we set targets because we want better performance, yet rigid targets produce an environment where people fear the data — and an environment that fears its own data cannot improve. We get the exact opposite of what we wanted: a number gleaming on the wall, and a system quietly rotting beneath it.

Shewhart and the Idea That Removes the Need for a Target

If Goodhart explains why targets corrupt, and Deming what they cost us, Walter Shewhart offers the alternative. In the 1930s, at Bell Laboratories, Shewhart developed what came to be known as Statistical Process Control. His central idea is astonishingly simple, yet it demolishes the foundation of management by targets: not every change in a number carries meaning.

Shewhart distinguished two kinds of variation. Common cause variation is inherent in any process; it is always present and calls for no intervention. Special cause variation is a deviation outside the usual pattern, pointing to a specific cause worth investigating. The entire skill lies in telling them apart: when is what I see merely the natural noise of the system, and when is it a genuine signal?

And here lies the undoing of the numeric target. A rigid threshold treats every number below it as a failure and every number above it as a success, regardless of whether the difference is nothing but meaningless natural variation. You may succeed today because luck fell within the natural range, and fail tomorrow for the same reason — the target applauds you in the first case and punishes you in the second, while nothing in the process has changed. It generates false signals: alarming when there is no danger, reassuring when there should be worry.

Trend-based measurement, by contrast, asks the right question: not 'did we cross the line?' but 'has the behavior of the process changed?'. This needs no target to answer; it needs only the process's natural range of variation and an eye that reads the trend. When performance stays within its usual range, the process is stable even if its level does not please us — and the remedy is to redesign the system, not to scold individuals. When performance leaves its range, a special cause deserves immediate investigation. So we read reality as it is, with no threshold lying to us in either direction.

What Is Trend-Based Measurement?

Trend-based measurement replaces the question 'did we hit the number?' with a more honest and useful one: 'where is performance moving, and is its variation natural or exceptional?'. It does not abolish the number; it frees it from judgment and returns it to its original purpose — telling the story of the process over time. Instead of a single point hung on a threshold, we read a moving line that has memory.

This approach rests on four elements any process owner can apply:

To make this concrete, imagine a measure of request-processing time. The traditional approach sets a target — say 48 hours — and paints every request that exceeds it red, producing a dashboard that tells us who 'passed' and 'failed' but nothing about the health of the process. Trend-based measurement plots processing time across the weeks and sees its natural range, revealing what the threshold never does: that the average is stable but the variation is widening (a sign of losing control), or that a jump coincides with a system change. These are real operational questions, and the threshold is blind to every one.

But Don't We Need a Standard? On the Difference Between Trend and Drift

The most legitimate objection to measurement without targets is this: if there is no line we strive to cross, what prevents a gradual slide into poor performance that merely looks 'stable'? Is the threshold not what pulls us upward? The objection deserves a candid answer.

The answer is that trend-based measurement is not the absence of a standard but its replacement. Instead of a fixed threshold we leap over once and forget, our standard becomes the behavior of the process itself over time: is it improving or deteriorating? Is its variation widening or settling? This is a more demanding standard, not a looser one, because it does not reward you for reaching a number and stagnating; it continually asks about your direction. The threshold lets you relax the moment you cross it; the trend never does.

As for the danger of 'bad stability' — a process settling at a low level — trend-based measurement exposes it more precisely than a threshold, not less. When a process is stable within a natural range, but that range sits at a level that does not satisfy the customer, the diagnosis is clear: the problem is not a stray deviation but the design of the whole system, which needs re-engineering. The threshold hides this behind the daily noise of pass and fail; the trend lays it bare.

Here a decisive distinction emerges, between a strategic aim and an operational target. A strategic aim is legitimate — a service time that delights the customer, a competitive level of quality. But it is pursued by redesigning the system to be capable of it, not by hanging a number around the necks of individuals and holding them to it daily. Trend-based measurement keeps the ambition intact, but places it where it belongs: in the design of the process, not in the evaluation of the person who executes it.

What Changes on the Operations Floor?

The deepest effect of moving from targets to trends shows up not on the dashboards but in people's behavior. When a process owner knows that their measure will not be used to hold them to account, but to help them understand their process, data turns from a threat to be warded off into a mirror to be used.

The first thing that changes is honesty. In a target environment, bad data is an enemy to be hidden; in a trend environment, it reveals an opportunity to improve. The employee who does not fear their number records it as it is, and for the first time the organization can see its true reality — and every improvement begins from an honest view.

The second change is in the kind of intervention. A process owner who reads a trend rather than a threshold learns to ask before acting: is this natural variation or a signal? They stop reacting nervously to every number — that 'tampering' Deming warned against, which adds chaos rather than calming it — and intervene only when the signal warrants.

The third change, perhaps the most important organizationally, is in the nature of the request for help. When the trends show that a process has slipped beyond its owner's control — variation that keeps widening, or a decline against which local interventions are useless — the owner can escalate not as a complaint but as a documented case: 'the trend shows a sustained deterioration over six weeks, I tried this and that, and the problem is in the system's design, not its execution'. This is not an admission of incapacity but the highest professionalism — and the difference between an organization that learns and one that pretends.

Where Do Targets Remain Legitimate?

For this argument to be honest, we must acknowledge its limits. Trend-based measurement is not an absolute doctrine that abolishes every number; it is best suited to the repetitive operational process, where gaming occurs and accumulates. There are contexts in which numeric targets remain legitimate, even necessary, and honesty requires naming them.

The common thread is that none of these is a target on which the operational individual is held to account for managing the number day by day. The contractual constraint is honored by the system, the safety limit protected by design, the strategic aim a guide for leadership. The problem was never the existence of a number somewhere; it was always its conversion into a daily whip on the back of the process owner. Remove that whip, and the number returns to being innocent and useful.

How to Begin the Transformation in Practice

The shift from a culture of targets to a culture of trends does not require a sudden revolution, but a deliberate experiment that proves its worth before it spreads. The smartest entry point is a single process whose numbers you suspect are 'too beautiful' — because that is often where the most gaming hides, and where the revelation will be most persuasive.

A candid warning: the cultural transformation is far harder than the technical one. Plotting a trend instead of a threshold takes hours; persuading a manager accustomed to asking 'did we hit the target?' to ask instead 'where is performance heading?' may take months. So begin small, and let the results speak. When managers see for themselves that removing the target exposed a hidden problem, and that the decision built on the trend was the truer one, the old question fades of its own accord.

Closing: From 'Did We Succeed?' to 'What Is Happening?'

We have returned to where we began, but with a different understanding. The moment you set a target, the data starts getting gamed — not because people are bad, but because the threshold tempts us to please the number instead of understanding reality, converting data from a mirror we learn from into a court we fear. Measurement without rigid targets is not a surrender of ambition; it restores measurement's original function: to help us see what is actually happening, not to prove what we want to believe.

Trend-based measurement offers this alternative concretely: we read performance as a story moving over time, distinguish natural from exceptional variation, intervene in proportion to the signal, and leave ambition in its proper place — in the design of the system and in strategic aims, not on the backs of those who execute the work. It is stricter than the threshold, not looser, since it lets no one relax merely by crossing a line.

In the context of Saudi Arabia's transformation under Vision 2030, where data is meant to be the foundation of decisions rather than an ornament on reports, this distinction becomes strategic in the fullest sense. An organization that runs green dashboards which do not match its reality deceives itself before others, and builds its largest decisions on shifting sand. An organization that dares to measure without fabricated targets possesses the most precious thing it could hold: data that tells the truth.

So the question we leave you with is not 'what are your targets?' but a deeper one: if you removed the target from your most 'successful' process tomorrow, would you discover that you had been measuring performance all along, or only your team's skill at pleasing a number pinned to the wall?