Long before there are departments, quarterly business reviews, dashboards, or executive scorecards, there is simply a problem worth solving. Every organization begins with a purpose. A founder sees an opportunity to make something better, gathers people around a common vision, and measures success in the most direct way possible: Are we making a meaningful difference for our customers?
In those early days, alignment comes naturally. The people making decisions are often the same people who speak with customers, build products, resolve problems, and celebrate successes. There is little need for elaborate governance or reminders of its purpose, as everyone understands why the organization exists. Decisions are not evaluated against a spreadsheet; they are evaluated against the mission itself.
As organizations grow, however, that simplicity begins to disappear.
New departments are created. Teams specialize. Managers are hired to oversee increasingly complex operations. The founder can no longer observe every customer interaction or personally guide every important decision. Leadership now faces a challenge familiar to every successful organization: How do you ensure thousands of individual decisions remain aligned with a single purpose?
The answer seems obvious. You measure them.
In many ways, modern management is built upon this idea. From Peter Drucker’s Management by Objectives and Kaplan and Norton’s The Balanced Scorecard, companies measure revenue growth, customer retention, product quality, operational efficiency, employee engagement, and profitability all attempts to answer the same fundamental question: Are we making progress toward our mission?
Measurements are not the problem, and in fact, they are critical to achieving the goals. Without them, organizations would have no reliable way to understand whether they were improving or drifting off course. The problem begins when something subtle happens, something so gradual that few organizations notice it until the consequences become impossible to ignore.
The measurements stop describing the mission, and they begin replacing it.
Charles Goodhart captured this phenomenon in a single sentence that has become known as Goodhart’s Law, one of the most quoted observations in economics and management:
When a measure becomes a target, it ceases to be a good measure.
Unfortunately, most discussions of Goodhart’s Law end there, but this is where mine begins. Goodhart explained what happens to measurements, but he did not explain what happens to organizations during their obsession with achieving them. That story begins with a simple but profoundly important shift in how decisions are made.
Goodhart’s Law has become especially relevant today, when nearly everything in organizations is tied to data and the measurement of some form of activity. With the growth of AI and the ability to mine even larger data sets, this risk is even greater.
In healthy organizations, measurements validate decisions already made in service of the mission. The mission comes first; the metrics confirm whether the organization is moving in the right direction. Over time, however, many organizations quietly reverse that relationship, and decisions are no longer evaluated according to the mission but are evaluated according to the metrics.
The distinction appears almost trivial, but it radically changes everything.
I didn’t fully appreciate this until years after building Hreflang Builder.
Looking back, the product appears to have evolved as most enterprise software does. New capabilities were added. More deployment options appeared. Reporting became more sophisticated. Automation expanded. By the time the platform matured, it had become significantly more capable than the original product. From the outside, it looked like feature-driven development, yet it wasn’t. As I have been analyzing my decisions and our development path for the KPI trap, I realized that every meaningful enhancement was driven by the same underlying question:
What is preventing this customer from succeeding?
Sometimes the answer was unnecessary complexity. Enterprise websites rarely fit neatly into a single pattern, so we developed increasingly sophisticated methods to map pages. We didn’t build dozens of mapping methods just to make for an impressive feature list; in fact, we often had to hide that fact because prospects became overwhelmed. They simply wanted to know, “Can you effectively map my URLs?”
Sometimes the obstacle was operational friction. Generating XML sitemaps was easy. Deploying them across hundreds of websites managed by different teams, content management systems, and release processes was not. Solving the technical problem without solving the deployment problem would have created elegant software that delivered little practical value.
Sometimes the obstacle was repetitive manual work. If customers found themselves performing the same task every week, the product had not solved the underlying problem. Automation was not another capability to advertise. It was simply the logical conclusion of asking why the manual work existed in the first place.
And sometimes the obstacle was invisible. Customers knew they had implemented hreflang correctly, but they struggled to demonstrate why the investment mattered. Reporting, validation, and diagnostics were not built because dashboards are fashionable. They were built because organizations need evidence that their efforts are producing business value.
Only much later did I realize that every significant decision had been filtered through the same four questions.
- Does it eliminate complexity?
- Does it remove operational friction?
- Does it automate repeatable work?
- Does it better demonstrate business value?
I now think of these as Mission Filters. At the time, they weren’t a framework but were simply how our product decisions were made. The product roadmap wasn’t driven by a desire to accumulate features but an intense desire to remove another obstacle standing between the customer and success.
Now imagine a different conversation. Imagine I walked into the quarterly planning meeting and announced a new objective. “This quarter our KPI is to ship twenty new features.”
Nothing else changes.
The customers are the same.
The engineers are the same.
The market is the same.
Only one sentence has changed.
Yet every subsequent roadmap discussion becomes fundamentally different.
An engineer proposes redesigning the mapping engine, reducing implementation time by seventy percent while making every existing capability dramatically easier to use. The first question is no longer, Will this help our customers succeed? The first question becomes, Does it count as a feature? The answer is no.
Another proposal introduces customizable interface themes, several additional export formats, and a collection of cosmetic improvements that marketing can announce as new capabilities.
Collectively, they count as six features. Those projects move to the top of the roadmap.
No one intended to build a worse product.
No one ignored the customer.
No one acted irrationally.
The organization simply began optimizing the system it had been asked to optimize.
That is the deeper lesson behind Goodhart’s Law. Metrics do not merely measure behavior; they shape it.
Every KPI is a behavioral design decision because every KPI tells intelligent people where to invest their effort. When the metric remains connected to the mission, those behaviors strengthen the organization. When the metric becomes the objective, the behaviors gradually drift away from the purpose the organization was created to serve.
I believe this is where many organizations unknowingly begin a much larger journey.
Departments start translating the mission into local objectives they can control. Revenue becomes quarterly targets. Customer experience becomes response time. Product innovation becomes release velocity. Marketing becomes impressions and clicks. Procurement becomes cost reduction. Each decision makes perfect sense within its own department. Collectively, however, something important begins to disappear.
The mission slowly fragments into proxies.
That fragmentation is not the end of the story.
It is the beginning.
In the next article, I explore what happens after organizations begin managing proxies instead of purpose, why intelligent people make decisions that ultimately undermine the customer experience, and how those perfectly rational decisions accumulate into what I call The KPI Trap.
