“When a measure becomes a target, it ceases to be a good measure.”
This quote from British economist Charles Goodhart could, on its own, sum up a large part of contemporary managerial excesses.
Initially applied to economics and monetary policy, Goodhart’s law has gradually found a much broader application: organizations obsessed with indicators, dashboards, and KPIs. In a world governed by data, it stigmatizes a form of collective blindness, which is why it deserves to be brought back into the spotlight today.
In short:
- Goodhart’s law shows that when an indicator becomes a target, it loses its relevance and can lead to counterproductive behavior.
- In businesses, an obsession with indicators transforms measurement tools into ends in themselves, distorting the original meaning of initiatives (e.g., well-being at work, use of digital tools, diversity).
- This managerial drift reflects a confusion between measurement and meaning, where data-driven management creates a form of bureaucracy and undermines real performance.
- To remedy this, it is essential to link each indicator to a clear intention, adapt them to the context, co-construct them with teams, and cultivate discernment rather than dependence on figures.
- Goodhart’s law reminds us that indicators must remain a means to an end, and not become an end in themselves.
An economic law that became an organizational principle
Initially formulated in the 1970s, Goodhart’s law was intended to criticize the way British governments used economic indicators to make decisions. But it quickly went beyond this framework because its central insight is almost universal: when you turn an indicator into a target, you push people to optimize the measurement, even if it means betraying its original meaning.
In other words, the system breaks down as soon as we forget why we are measuring.
What was originally intended as a decision-making tool to help us quantify and objectify a situation then becomes an end in itself, and with it, the behaviors, trade-offs, and even the values of the organization gradually change.
When indicators replace original intentions
Modern businesses, with their penchant for data and measuring everything, provide the perfect playground for observing the effects of Goodhart’s law in real life. And there are plenty of examples.
Employee experience: the barometer trap
Many businesses have embraced the topic of engagement and quality of life at work. This is legitimate and necessary. But very quickly, annual or quarterly barometers become the only compass. We no longer seek to understand or improve the reality of the situation, but to get a better score. Everything is done to influence responses, smooth out differences, and show that scores are improving. And I can tell you that there are a thousand ways to get the answers you want. In doing so, we stray from the ultimate goal: to create an environment that is truly more fulfilling and conducive to individual and collective success, not just more “satisfying on paper”.
Digital workplace: the reign of unproductive activity
Another common example is collaborative tools. Their use is measured extensively (number of connections, messages exchanged, files shared, etc.) in the name of so-called effective collaboration, but this ends up encouraging absurd behavior: over-solicitation, useless chatter, digital “noise” (How to motivate your employees to blow hot air instead of being productive (thank you Microsoft)). The indicator values volume, not relevance, because we measure the use of tools and not what they help to accomplish. We therefore work less to produce and more to use tools, which become not a means but an end in themselves.
Diversity and inclusion: the temptation of quotas
When it comes to diversity, intentions are often sincere, but indicators quickly become quantified targets: percentage of women, percentage of diverse profiles, etc. Here again, there is a risk of “playing the numbers game”: artificially creating diversity in the figures without transforming practices and culture. The goal is no longer to be fairer, but to tick boxes.
We may end up with people being promoted or recruited for the wrong reasons or, worse still, extremely competent people who will be burdened by the illegitimacy of the “quota effect” despite themselves.
A deeply cultural managerial drift
What Goodhart’s law reveals is a problem that runs much deeper than the poor design of an indicator. It points to a systemic tendency to confuse steering and control, measurement and meaning, visibility and impact.
By trying to make everything measurable, objectifiable, and traceable, we end up shifting priorities. The role of the manager, who is supposed to be the guarantor of meaning, framework, and performance, is reduced to that of a guardian of dashboards, and employees end up making reporting their main activity.
The promise of data-driven management has turned into algorithmic bureaucracy and, more broadly, the promise of a quantified organization can, despite its appeal, be completely misused and lead to a number of abuses (The quantified organization: Grail or Big Brother?).
How to get out of it?
Fortunately, nothing is irreversible as long as you dare to question certain certainties.
1. Return to the purpose
Each indicator must be linked to a clear and useful intention. Are we measuring to understand, to improve, or just to reassure ourselves (In KPI, K stands for Key)? If we can no longer answer this question, we have crossed over to the dark side of measurement.
2. Change metrics when the context changes
A good indicator quickly becomes outdated. You have to accept this, revise it, or even get rid of it. Managerial courage also means knowing when to give up data that has become useless or even counterproductive.
If, for example, you have been impressed by Moderna’s recent announcements (HR and IT merger: Moderna redesigns its organization for and with AI) and are intrigued by their plan to reinvent work organization around the concept of flow, be aware that this will certainly require you to throw all your old indicators in the trash (Thinking of work as a flow: appealing, but is it realistic?).
3. Involve those who do the work
The best indicators are developed jointly with operational teams. They are the ones who know what makes sense, what is useful, and what is worth monitoring. Otherwise, you end up with indicators that are disconnected from reality.
Let’s take this a step further: not everyone, depending on their role, needs the same indicators, and it can be really dangerous to subject the field to indicators set by a manager or senior management. Not everyone needs to know the same things, and not all indicators have the same value depending on who is using them.
4. Strengthen judgment rather than reliance on figures
An indicator is only ever a signal, not the truth. It cannot replace discernment, dialogue, or contextual analysis. Good management begins where the figures end.
Bottom line
Goodhart’s law is not inevitable, but rather a warning. It tells us that any system becomes meaningless if we forget why it was designed. And in a world saturated with indicators, where artificial intelligence promises to measure, anticipate, and optimize everything, it is vital not to succumb to the illusion of management without human judgment.
It is not a question of abandoning indicators, but of putting them back in their rightful place: as a tool at the service of a project, rather than a project enslaved to its tool.
Image credit: Image generated by artificial intelligence via ChatGPT (OpenAI)
