The Two Languages of Change: Why Transformations Are Launched on Values and Run on Targets
The measure is a proxy, and every proxy is a loan against reality that comes due the moment someone is paid to move it.
Executive Summary
Every serious transformation is launched twice. It is launched first in the language of values — the town-hall address, the statement of purpose, the promise that this time the organisation will put the customer, the citizen, or the mission at the centre of what it does. Then, a few weeks later, it is launched again, more quietly, in the language of targets: the milestone plan, the benefits tracker, the status report the steering committee will actually spend its forty-five minutes reading. The two launches speak different vocabularies, and over the life of a programme the second vocabulary almost always wins.
This essay is an attempt to understand why — and to resist the easy conclusion. The comfortable story is that targets are the enemy of purpose, that measurement crowds out meaning, and that the remedy is to lead with values and let the numbers look after themselves. That story is half right, which is what makes it dangerous. A transformation with no measurable aims is not more virtuous; it is merely unaccountable, and unaccountable change is where good intentions go to die slowly and expensively. The real difficulty is subtler. The two modes are not equal partners, because one of them is far easier to operate than the other. A target is legible, delegable, auditable, and reassuring. A value is none of these things. And so, absent deliberate counter-pressure, the machinery of delivery selects relentlessly for what it can measure over what it meant — until the programme is reporting green against a scorecard that has quietly stopped describing the thing it was built to change.
The argument runs in four movements. First, that transformation is bilingual by nature, and that its two languages describe different objects. Second, that a set of structural forces — not weakness, not cynicism — keeps tilting the ground toward the target. Third, the strongest case for the target, which is stronger than its critics allow and must be met honestly. And fourth, a proposed settlement: not the triumph of values over targets, but a defensible hierarchy between them, and a handful of disciplines that keep the measure attached to the meaning.
The two languages of change
Watch what happens in the first month of any ambitious change effort and you will hear the switch occur. The launch is conducted in the language of ends. We exist to serve; we will be trusted; we will treat the person in front of us as a person and not a transaction. It is genuinely felt, and it does real work: it tells people why the disruption is worth enduring. Then the programme has to be run, and running it requires a different grammar. The plan needs milestones. The milestones need owners. The owners need something to be held to, and “trusted” is not something anyone can be held to at a fortnightly review. So the value is translated — into a service-level figure, a cost-to-serve reduction, a satisfaction score, a go-live date. The translation feels like nothing more than making the value operational. In fact it is a change of object.
A value is a claim about which results are worth having. A target is a claim about how much of one particular result we will produce by when. The first is a statement of direction and priority; the second is a quantity with a deadline. They are related — a good target should be a faithful sample of a value — but they are not the same kind of thing, and the difference matters enormously once the pressure comes on. When a value and its target diverge, the person on the floor has to choose which to serve, and the entire apparatus of the programme — the reporting line, the incentive, the escalation, the applause — is pointing at the target. The measure is a proxy, and every proxy is a loan against reality that comes due the moment someone is paid to move it.
We have a rich professional literature warning us of exactly this. The observation that a measure ceases to be a good measure once it becomes a target is now old enough to be a cliché, and older still is the insight that the more any quantitative indicator is used for decision-making, the more it will be gamed and the more it will distort the process it was meant to monitor. The generation of quality thinkers who shaped post-war management were emphatic on the point: numerical quotas drive out pride in the work, and fear of the number drives out the honesty on which any real improvement depends. None of this is new knowledge. What is striking is how completely we ignore it under delivery pressure — not because we have forgotten it, but because the structure we operate in keeps making the target the path of least resistance.
Why the target keeps winning
If target-driven behaviour were simply a failure of character, it would be rare, and it would cluster around cynical or lazy organisations. It is neither rare nor clustered. It appears in earnest, well-led programmes staffed by people who believe in the stated values. That should tell us the cause is structural rather than moral. At least five forces push the same way.
- Legibility. A target can be written on a slide; a value cannot. The organisation above the programme sees only what fits into a reporting template, and a template is a machine for converting a rich situation into a small number of comparable quantities. What cannot be reduced to the template is, for practical purposes, invisible to the people holding the programme to account.
- Delegability. You can hand a target down a hierarchy and hold each level to its slice. You cannot hand down a value in the same clean way, because a value has to be interpreted afresh in each situation, and interpretation cannot be audited as easily as attainment. Large organisations run on delegation, and delegation runs on things that survive being passed down a chain — which targets do and values do not.
- The reassurance of the number. A senior sponsor carrying real career risk wants to know the programme is safe. “We are living the values” cannot discharge that anxiety; “we are eighty-seven per cent to plan and amber-trending-green” can. The number is a sedative. It is administered most eagerly precisely when the underlying situation is most uncertain, which is exactly when it is least trustworthy.
- The audit trail. In a post-Enron, post-Sarbanes world, the instinct to document control has never been stronger, and control is documented in metrics. A committee that governs by measurable target leaves an evidence trail that protects it if the programme is later questioned. A committee that governs by judgement about values leaves only judgement — which, in a blame-conscious culture, feels dangerously exposed.
- The cadence trap. Values move on the timescale of years; the reporting cycle moves on the timescale of a fortnight. Anything that cannot show movement between this review and the next struggles to hold attention against things that can. The reporting rhythm silently privileges the short-cycle metric over the long-cycle purpose, not by anyone’s decision but by the simple physics of what can change in two weeks.
None of these forces is illegitimate. Legibility, delegation, reassurance, auditability, and rhythm are how large organisations function at all. That is the uncomfortable part. The drift toward the target is not a bug introduced by bad actors; it is the ordinary working of the delivery machine doing what it is built to do. Which means it will not be fixed by exhortation, by a better values statement, or by a more inspiring launch. It can only be resisted deliberately, and resistance has a cost that someone must be willing to pay.
The drift toward the target is not a failure of the machine. It is the machine, running correctly. Anything that must be resisted continuously will not be cured by a better slogan — only by a discipline someone is accountable for maintaining.
The strongest case for the target
It would be easy, and wrong, to stop there — to cast the target as villain and the value as victim. Any honest account has to meet the opposing case at its strongest, and the case for target-driven change is strong.
Start with the plain fact that targets have delivered. Consider what a rigorous, numerically-driven regime achieved in public services over the last decade. When a health system committed to a hard maximum wait in emergency departments and held itself to it relentlessly, waits that had been tolerated for years as regrettable but immovable began to fall. When elective waiting times were made a published, non-negotiable figure with consequences attached, the figure came down. One can dislike the coercion, the gaming, and the perverse effects — and they were real — while still admitting that a great deal of genuine human benefit was won by the crude instrument of a number with teeth. Waiting patients did not care about the theory. They waited less.
The target’s defenders make a further point that the values party finds hard to answer. Values, untranslated into measurable commitment, are cheap. Everyone can sign up to serving the customer, because signing up costs nothing and commits no one to anything in particular. It is precisely the discipline of the target — this figure, by this date, or someone answers for it — that converts a comfortable aspiration into a real allocation of effort and consequence. The tyranny of good intentions is that they permit endless motion without progress; the target, for all its brutality, forces the question of whether anything actually moved. A well-known industrial quality movement of recent decades built extraordinary performance gains on exactly this refusal to accept sentiment in place of measured result.
So the steelman is not merely “targets are sometimes useful.” It is this: measurement is the only known cure for the organisation’s deep capacity to feel virtuous while achieving nothing. Remove the target and you do not liberate the value; you remove the one mechanism that stops the value from becoming a comfortable lie. Any argument for values-driven transformation that cannot absorb this point is not worth making.
Where the seam splits
The synthesis has to be built on top of that concession, not in denial of it. And the place to build it is the point where a value and its chosen measure come apart under load. Consider a case — composite, but true to a hundred real ones.
An organisation resolves to transform its customer service. The stated value is unimpeachable and sincerely meant: resolve the customer’s problem, first time, so they do not have to come back. Everyone believes it. The programme, needing to be run, translates it. First-time resolution is hard to measure cleanly and slow to report, so the scorecard settles on the two figures that are easy and fast: average handling time per contact, and contacts handled per adviser per day. Both are proxies for efficiency, and efficiency is assumed to be a friend of the value. The transformation is now, in practice, driven by those two numbers, because those two numbers are what appear on the dashboard the steering group reads.
The results arrive, and they look like success. Average handling time falls from 6.2 minutes to 4.4. Contacts handled per adviser per day rise by roughly a fifth. The efficiency benefit is booked; the programme reports green; the sponsor relaxes. But underneath, the value the programme was created to serve is being eaten. Advisers, paid to be quick, close contacts before the underlying problem is truly settled, because the deep resolution costs the very minutes the target punishes. First-contact resolution slips. Repeat contact within a week climbs from around eighteen per cent to thirty-one. Total volume rises — because unresolved customers come back — and with it the escalations and the complaints. The true figure, the cost of actually resolving a customer’s problem rather than the cost of handling a call, goes up even as the cost per call goes down. The programme has optimised its proxy to the point of inverting its purpose, and every dashboard is green while it does so.
| What the value asked for | What the target rewarded | What the organisation got |
|---|---|---|
| Resolve the problem, first time | Shorten each contact | Faster contacts, unresolved problems |
| Fewer reasons to come back | More contacts handled per day | More repeat contact, higher true cost |
| Trust built over time | Movement visible this fortnight | A green report over a hollowing service |
Nothing here required a villain. Every actor behaved rationally given what they were measured on. This is the whole mechanism of the essay in a single scene: the value was real, the target was its faithful-looking proxy, and the proxy decayed the instant it became the thing people were paid to move. The gap between transformation intent and transformation reality is not usually a gap between what leaders said and what they secretly wanted. It is the gap that opens, silently and lawfully, between a value and the measure that was supposed to stand in for it.
What the target cannot see
Why does the proxy decay? Because a measure is a sample of a value, and a sample can always be satisfied without satisfying the thing it was sampled from. “Resolve first time” is a rich, situation-dependent judgement; “handle in under five minutes” is a thin, universal rule. The moment the thin rule carries the weight of consequence, effort flows to the rule and away from the richness it was meant to represent. This is not a failure of the particular measure. A better call-centre metric would have delayed the divergence, not prevented it, because the divergence is inherent in the act of substitution. Every proxy is a compression of reality, and every compression throws information away. Under enough pressure, the discarded information is always exactly the part that mattered.
This is where the value earns its keep, and it is a more precise role than the inspirational literature usually claims. The value is not primarily there to motivate, though it may. Its indispensable job is governance of the measure: to specify which results count, to say when a number has stopped standing for the thing it was chosen to represent, and to license the organisation to override the metric when the two diverge. Without a live value sitting above them, targets have no principle by which to be corrected; they simply accumulate authority until they are gamed to exhaustion and replaced by fresh targets that will meet the same fate. The value is the thing that remembers what the number was for.
“A value that cannot be measured at all is decoration. A measure that answers to no value is a countdown. The whole craft of transformation lives in the tension between them.”
Put the two failure modes side by side and the shape of the problem is clear. Values without measures give you the organisation that feels wonderful and changes nothing — motion without progress, sincerity without accountability, the perpetual relaunch of the same aspiration. Measures without values give you the contact centre above — accountability without meaning, precise attainment of a figure that has come loose from its purpose. Neither is a place to live. The task is not to choose between them but to get the relationship the right way up.
Toward an honest settlement
If the two are not equals, which governs which? The settlement I would defend is that the value is sovereign and the target is its servant — but a servant with real teeth, not a decorative one. Concretely, that means a small number of disciplines, each of which has a cost that leadership must consciously agree to bear.
- Name the value the target is a proxy for — on the same page as the target. Every headline measure on the dashboard should be written next to the value it stands in for, so that no one can read the number without being reminded what it was sampled from. The cost is that it makes the act of substitution visible, and visible substitution invites the awkward question of whether the proxy is still faithful. That awkwardness is the point.
- Pair every efficiency measure with a guardian of the value it could erode. A handling-time target travels with a first-contact-resolution measure; a cost-reduction target travels with a quality or trust measure that would move in the opposite direction if the value were being sacrificed. The guardian metric is not there to be optimised; it is there to scream. The cost is that the programme can no longer show an unambiguous win, because the guardian will often be flashing while the headline is green — which is precisely the information the single number was hiding.
- Review the measures, not only the performance against them. Most governance asks “are we hitting the numbers?” Almost none asks “are these still the right numbers?” A transformation should revisit, on a deliberate cadence, whether each target still faithfully represents its value or has begun to drift — and retire the ones that have. The cost is that it treats the scorecard as provisional rather than fixed, which is unsettling to anyone who wanted the comfort of a stable target to chase.
- Reserve, explicitly, the right to override the metric. There must be a named authority empowered to say: the number says green, the value says we are failing our customers, and the value wins. If no one holds that authority, the metric is not a servant but the sovereign, whatever the values statement claims. The cost is that it reintroduces judgement — and therefore exposure — into a system that adopted targets partly to escape judgement.
Notice what these disciplines share. Each one reintroduces friction that the pure target regime had smoothed away — the friction of visible substitution, of ambiguous results, of provisional measures, of exposed judgement. That friction is not waste. It is the tax an organisation pays to keep its measures honest, and the reason target-driven transformation is so seductive is precisely that it lets you stop paying it. A programme that has become suspiciously frictionless, where every report is clean and every number cooperates, is not usually a programme that has mastered its purpose. It is usually one that has quietly swapped its purpose for its proxy and has not yet noticed.
Closing reflection
We should be honest that this is hard, unglamorous, permanent work, and that it will never be finished. The drift toward the measurable is not an episode to be survived once but a current to be swum against for as long as the organisation exists. The moment the counter-pressure relaxes — a sponsor leaves, a delivery date looms, a difficult quarter arrives — the ground tilts back toward the target, because the target is always the easier master to serve.
But the choice was never really values versus targets, and the practitioners who framed it that way did the field a disservice. Values without targets are a sermon; targets without values are a countdown. The organisations that actually change, as opposed to those that merely feel as though they are changing, are the ones that hold the two in a deliberate and slightly uncomfortable tension — using the target for the accountability that values alone can never supply, and using the value to keep asking, relentlessly, whether the target still means what it was meant to mean. The green dashboard is not the enemy. Forgetting what the green was for is the enemy. And the whole discipline of transformation, in the end, comes down to refusing to forget.