# Measurement Is Not Management

{% hint style="info" %}
This is a living document. It is maintained and updated without notice. It's best to refresh pages to ensure delivery of the latest iteration from the CDN Content Delivery Network in your area
{% endhint %}

One tenet of business has been that if you cannot measure it, you cannot manage it. But the authors of "The Performance Management System: Turning Strategies Into Results" take that notion several steps further. They argue that good management means managing the drivers that produce change, not the results themselves.

In 1995, management consultants published an article in *Strategy+Business* titled "The Perform System: Turning Strategies into Results." The opening line deserves attention: **"Measurement is not management."**

The authors identified a fundamental problem in how businesses approach performance systems. Companies were obsessing over financial metrics and results **while ignoring the operational decisions that actually produced those results.** They called this phenomenon the difference between scorekeeping and understanding cause-and-effect.

Here's the critical distinction they made:

**Measurement** = collecting data, tracking results, keeping score\
**Management** = making decisions, taking actions, understanding what drives outcomes

Sound familiar?

The painting industry in 2025 perpetuates the same broken system that management experts identified as dysfunctional three decades ago. Job costing is pure measurement—it tracks labor hours, material costs, and profit margins. But it doesn't tell you *why* those numbers occurred or *how* to fix them.

The 1995 article identified four requirements for effective management systems:

1. **Set fair "stretch" targets** — Job costing can't do this without knowing what's actually achievable under controlled conditions.
2. **Manage the drivers, not the results** — Tracking that labor ran 20% over doesn't manage anything. Understanding that recoat times were violated, conditions weren't controlled, and sequence was random—that's managing drivers.
3. **Focus analysis on root causes** — Without standard work, job costing can't isolate whether the problem was the product, the sequence, the conditions, or the technique.
4. **Reward good decision-making, not just good results** — Job costing only shows results. It can't distinguish between a crew that got lucky with weather and a crew that controlled conditions properly.

The authors put it bluntly: **"Measurement itself is inert. The impact is derived from the way it is embedded into management processes."**

That's the part Slavik, Honan, and the rest of the job costing evangelists miss entirely. The spreadsheet doesn't solve problems—it reveals symptoms. And without a system to interpret those symptoms, you're just documenting failures you can't fix.

The article goes further. They compared their Performance Management System to the Balanced Scorecard—a popular business framework that tracks multiple metrics across different aspects of a company. Their critique applies perfectly to job costing in painting:

> "While it does measure variables that are important in achieving a company's strategic objectives, it does not explicitly link them to each other or to the company's financial measurement system."

Job costing tracks labor, materials, and profit. But it doesn't link those numbers to the operational decisions that produced them. Was the high labor cost due to poor sequencing? Bad product selection? Uncontrolled conditions? Job site variables? Personal preference? Random improvisation?

Without standard work, you can't tell. The numbers float in isolation, disconnected from the physics and chemistry performed by the people who actually govern production.

The authors identified this as a "scorecard" problem:

> "While it focuses on measures across four critical aspects of business positioning, it does not focus on how to achieve those results... A tennis player who watched the scoreboard intently, instead of the ball, the court, and the opponent, would not win a game, much less a match, no matter how well designed the scoreboard."

Job costing is watching the scoreboard instead of playing the game.

Here's what makes this particularly damning: the authors explicitly warn against exactly what the painting industry is doing. They call out systems that "focus narrowly on financial measures and results, rather than on the real drivers of business decisions."

That's job costing in a single sentence.

They explain why this fails: **"Without this shift, performance measurement is a marginal activity."** 

Translation: if you're measuring results without understanding and controlling the operational drivers that produce those results, you're wasting your time.

The irony is almost too perfect. Management consultants published this critique in 1995. They identified the exact failure mode that defines modern painting business advice. Thirty years later, the industry continues to promote the same flawed approach to contractors.

Job costing isn't new wisdom. It's old failure.

The solution the 1995 article proposed: **"Create a business model that explicitly links operating decisions to financial performance."** In painting, that means:

* Define standard work (the sequence, tools, conditions, and methods)
* Execute the standard consistently across crews
* Use job costing to verify adherence to the standard
* Treat deviations as signals requiring investigation

This is the correct order. Standard work comes first. Measurement comes second. Without that foundation, job costing is just expensive documentation of randomness.

Slavik claims job costing can "solve almost any problem or friction point you have in your business." Management experts explained 30 years ago why that's impossible. **Measurement reveals problems. Systems solve them.**

The painting industry is still trying to manage by staring at the scoreboard.

It should be understood that job costing without standards is not only ineffective but also a documented failure mode, which business consultants identified and solved in 1995.

This excerpt also appears in [Job Costing](https://jackpauhl.gitbook.io/archive/field-notes/operations-systems/job-costing) but serves well as standalone advice.