The illusion of progress is one of leadership's most expensive traps

One of Goldratt's most powerful ideas is also one of the most uncomfortable:

Optimizing something that is not the constraint is a waste of time. And sometimes, it makes the whole situation worse.

This article is Part 2 of our Theory of Constraints series. Part 1 laid out the core idea, every system is limited by one constraint, and gave you three questions to find yours. Part 3 walks Goldratt's five focusing steps through one continuous case. Here, we stay inside the trap itself: the feeling of moving forward while the system stands still.

Why do we optimize the wrong things?

Because we naturally like improving what is visible, measurable, rewarding, or easy to discuss… even when that is not where the system actually wins or loses.

A shipped feature demos well. A grown pipeline fills a dashboard. A new ritual shows the team we are "doing something." The constraint, meanwhile, is usually invisible, awkward to measure, and uncomfortable to discuss, which is precisely why it stays untouched. In our coaching sessions, we rarely meet leaders who lack ideas for improvement. We constantly meet leaders whose ten improvements are all happening away from the one point that is capping the system.

Three mini-cases show how the illusion works in practice: in sales, in product, and in management.

Mini-case 1: sales, pouring more leads into a blocked pipeline

A sales leader we worked with faced flat revenue and responded the intuitive way: more top-of-funnel. New SDR, new outbound sequences, a bigger events budget. Six months later, the pipeline had nearly doubled. Revenue had not moved.

The real constraint was the team's ability to move decisions forward: deals entered discovery and quietly died there, because conversations stayed shallow. Salesforce's State of Sales research points at the same gap: 86% of B2B buyers say they are more likely to buy when the seller understands their goals, and 59% say most sellers do not take that time. More leads did not fix that. They made it worse:

• more complexity,
• more noise in the pipeline,
• more frustration,
• more fatigue,
• and more false reassurance that "things are moving."

Everything looked like growth. Nothing compounded. The turnaround began when the team cut its active opportunity count and rebuilt deal reviews around one question: what decision moved forward this week?

Mini-case 2: product, shipping faster into weak adoption

A product team doubled its delivery pace over two quarters. More releases, more tickets closed, a roadmap that looked impressively alive. Activation and retention did not move at all.

The constraint was not build speed. It was the first fifteen minutes of the user's experience: most new users never reached the moment where the product's value became obvious. Every new feature shipped past that point was invisible to the majority of users, and each one made the product heavier and the onboarding noisier. Velocity, the metric everyone celebrated, was a local optimization.

The team froze new features for six weeks and worked only on time-to-first-value. Usage rose without a single new capability being added. The roadmap had not been the engine of progress; it had been the feeling of progress.

Mini-case 3: management, adding rituals to escape the fog

A scale-up leadership team felt execution slipping: missed commitments, slow decisions, tension between departments. Their response was additive: more hiring, a new weekly sync, two new tools, a training program.

Each addition was defensible. None touched the constraint, which was that nobody knew who owned which decision. And that is the general pattern:

• hiring more people does not solve a lack of clarity,
• adding meetings does not solve weak ownership,
• buying more tools does not solve poor decision-making,
• offering more training does not solve the absence of real-world application.

When the team finally wrote down decision rights, who decides, who is consulted, who is informed, they cancelled two of the new meetings within a month. The system had not needed more inputs. It had needed one blockage removed.

The activity vs impact test

Leadership becomes blurry when we confuse:

• activity with impact,
• movement with progress,
• local optimization with systemic improvement.

This is a common trap for smart, committed, hard-working leaders. They see many things that could be improved, so they improve many things. But until the real constraint is addressed, the system remains capped.

Here is a test we use in coaching sessions. Take any initiative currently consuming your team's energy and ask four questions:

1. If this succeeds perfectly, what number moves? Name the metric. If it is a local metric, leads generated, features shipped, meetings held, keep asking until you reach a system-level outcome, or admit you cannot.

2. Does it touch the constraint? If you cannot name your constraint, stop here: that is the real finding, and the three questions in Part 1 will help.

3. What would happen if we stopped? If the honest answer is "nothing much, for a while," you have been funding motion, not progress.

4. Why did we choose it? Because it matters, or because it is visible, measurable, rewarding, or easy to discuss?

Then ask the summary question, the uncomfortable one:

Among everything we are currently improving, what is truly increasing the system's overall performance… and what is mostly creating a feeling of progress?

It is uncomfortable because it forces us to admit that some useful actions are not strategic actions.

What should we measure instead?

If activity metrics lie, what tells the truth? Goldratt's answer was throughput: what the system as a whole actually delivers toward its goal, per unit of time. In practice, that means shifting your dashboard's center of gravity.

• In sales, not leads generated but decisions advanced: how many opportunities crossed a real commitment threshold this week.
• In product, not features shipped but users reaching first value, and how fast.
• In management, not meetings held but decisions made, communicated, and still standing a month later.

And one measurement almost nobody tracks: the queue in front of the constraint. Its length and age tell you more about next quarter than most KPIs on your dashboard. When the queue shortens, the system is genuinely speeding up. When activity rises and the queue does not move, you are watching the illusion in real time.

The discipline of stopping

That admission is exactly what separates leaders who manage activity… from leaders who manage value creation.

Stopping is harder than starting. Every initiative has an owner, a history, and a constituency; killing it feels like a broken promise. This is where the illusion of progress meets its cousin, the dilution trap: unable to stop anything, we spread attention across everything, and the constraint receives the same thin slice of energy as the noise. The alternative is not brutality: it is honesty. Park the useful-but-not-strategic work explicitly. Say out loud: this is good, and it is not now.

Sometimes the best next move is not to do more. It is to stop optimizing what is not actually unlocking anything.

If you want to run the activity-vs-impact test on your own portfolio, doing it with a sparring partner keeps you honest: our own projects are the hardest to judge. Practice it with Lumia, your AI coach: list your top five initiatives and let the four questions sort them.