Goodhart’s Law

Introduction

Goodhart’s Law states: “When a measure becomes a target, it ceases to be a good measure.”

The law highlights a common problem in management: when people are judged or rewarded based on a specific metric, they often optimise the metric rather than the underlying goal the metric was meant to represent.

In short, the metric becomes target and behaviour shifts so that the metric loses meaning.

Change programs rely heavily on metrics, KPIs and dashboards to track adoption and progress.

However, Goodhart’s Law warns that if these metrics are poorly designed, they can produce false signals of success.

This means that once people know they’re being measured against a certain metric, they often game the system to hit the target rather than genuinely improving performance.

Impact

Instead of genuine change, organisations may see:

• superficial compliance
• gaming of metrics
• “checkbox” behaviour
• inflated reports of progress

This can make leaders believe change is succeeding when real behaviours have not changed.

Some Examples:

• Education (If schools are judged only by test scores, teachers may “teach to the test” instead of nurturing real learning.)
• Business (If customer service reps are measured by call length, they may hang up quickly to meet the target rather than solving the problem.)

Some Common Goodhart Effects in Change Programs

1. Training Completion Metrics

Target:
“100% of staff complete the change training module.”

Behavioural effect:
Employees rush through the training just to complete it.

Result:
Training completion is high but understanding and capability remain low.

2. System Adoption Metrics

Target:
“Employees must log into the new system weekly.”

Behavioural effect:
People log in briefly to meet the target but continue working in old systems.

Result:
Adoption statistics look good, but work practices haven't changed.

3. Project Milestones

Target:
“Teams must deliver implementation milestones on schedule.”

Behavioural effect:
Teams mark tasks as completed prematurely or reduce scope.

Result:
The change appears on track but quality and effectiveness suffer.

Some Reasons for This Happening in Change Initiatives

Goodhart’s Law interacts with several behavioural factors:

• Status quo bias, ie people prefer old habits
• Loss aversion, ie people fear losing competence or efficiency
• Social proof, ie if others game metrics, it spreads

Together, these pressures encourage metric optimisation rather than genuine change.

Some Practical Applications for Change Leaders

1. Measure Outcomes, Not Just Activity

Activity metrics are easiest to game.

Instead of measuring:

• training attendance
• logins
• meetings held

Measure:

• performance improvement
• behavioural adoption
• customer outcomes

An example:

2. Use Multiple Measures

One metric can easily be gamed.

Use triangulation:

• behavioural indicators
• performance outcomes
• qualitative feedback

Example:

• usage statistics
• employee feedback
• customer results

3. Combine Quantitative and Qualitative Data

Numbers alone can hide problems.

Add:

• interviews
• pulse surveys
• observational feedback
• stories of real adoption

This reveals whether the change is truly embedded.

4. Update Metrics Over Time

Early change metrics often measure activity.
Later metrics should measure impact.

progression example:

5. Watch for “Gaming Signals”

Signs that Goodhart’s Law is occurring include:

• perfect-looking dashboards
• sudden spikes near reporting deadlines
• behaviour reverting after reporting periods
• staff saying metrics “don’t reflect reality”

These are red flags.

An example: Digital Transformation

A company introduces a new CRM system.

Initial metric

% of employees logging into the CRM.

Goodhart effect

Staff log in but continue managing customers via spreadsheets.

Better metrics

• % of customer interactions recorded in CRM
• reduction in response time
• improved customer satisfaction

Now the metric reflects real behaviour change, not superficial activity.

Summary for Change Leaders

Goodhart’s Law reminds us:

Metrics are signals, not the goal.

In change management, success is not:

• training completed
• systems installed
• dashboards showing adoption

Success is people consistently working in the new way.

Goodhart’s Law teaches change leaders to measure what truly matters, design metrics carefully and look beyond numbers to actual behaviour change.

(main sources: Charles Goodhart, 1975; Jerry Muller, 2018)