The Work That Only Looks Like Progress
The most dangerous work is not the work you avoid. It is the work that feels productive while changing nothing.
An operator sits down on Monday with a free week and no external deadline. By Friday they have migrated the site to a new framework, redesigned the pricing page, moved the email list to a different provider and started a rebrand. It was a productive week by every internal measure. Nothing about the business is different.
This is the failure mode that does the most quiet damage to small internet businesses, and it is almost never recognised while it is happening. Neglect is easy to spot. Motion that produces nothing looks exactly like work, feels better than work, and can absorb years.
Why the substitute is so attractive
The tasks that get substituted in have a specific shape. They are fully under your control. They have a clear finish line. They produce something visible at the end. And critically, nobody outside can judge them, so they cannot fail publicly.
The work that actually moves a business has none of those properties. Talking to customers means hearing that the thing you built is not what they needed. Publishing means being ignored. Raising a price means finding out what people really think it is worth. Every one of these puts you in front of a verdict you do not control, and the mind is extremely good at finding urgent alternatives.
So the refactor wins. Not because anyone consciously decides the refactor is more important, but because it is the only item on the list that guarantees a good feeling by Friday.
The four usual suspects
The same categories come up again and again, and all four are legitimate work in the right circumstances. The problem is when they are chosen because they are comfortable rather than because they are the constraint.
- <strong>The redesign.</strong> Almost always defensible in the abstract, almost never the binding constraint. If nobody is visiting the page, its appearance is not the reason.
- <strong>The refactor.</strong> Justified when the current structure is actively blocking work you are trying to do. Not justified because you have read about a better way to organise it.
- <strong>The tool migration.</strong> Moving between services that do the same thing produces a week of work and a net change of zero. The new one is nicer. Nicer is not a result.
- <strong>The rebrand.</strong> The most seductive of the four, because it touches everything and feels foundational. A name change does not fix a product nobody is asking about.
The test that separates them
There is a question that cuts through this reliably, and it has to be asked before starting, not after: what specifically will be different if this succeeds, and how would I know?
Real work produces a crisp answer. People who could not find the thing can now find it. A step that used to require me now happens without me. A conversation I could not have before is now possible. Substitute work produces an answer that dissolves under pressure — it will be cleaner, it will be more professional, it will be easier to maintain later.
"Easier to maintain later" deserves particular suspicion. It is the most common justification for work with no near-term consequence, and it is very often true and completely irrelevant at the same time.
Naming the constraint out loud
The structural fix is not discipline. Discipline fails against something this comfortable. The fix is to keep a written statement of what is currently blocking the business, and to check proposed work against it.
It should be one sentence and it should be uncomfortable to write. Nobody knows this exists. People arrive and do not understand what it does. People understand it and do not want it. People want it and will not pay. Those are different constraints and they call for completely different work. Most substitute projects become obviously irrelevant the moment they are held next to the actual sentence.
The sentence also stops the other failure mode, which is doing genuinely useful work on a problem you do not have yet. Optimising a signup flow is real work. It is not real work when the signup page has had no visitors this month.
When the comfortable work is the right work
This should not turn into a rule that maintenance is always avoidance. Sometimes the refactor is exactly right — when the structure is genuinely preventing changes you keep needing to make. Sometimes the redesign is right, when people are arriving and leaving confused. Sometimes the migration is right, when the current tool is failing in ways that cost you customers.
The distinction is not the category of work. It is whether you can name the thing it unblocks. A refactor that lets you ship a feature you have been unable to ship is progress. A refactor that lets you feel better about code nobody is running is a hobby, and it is worth being honest about which one you are doing.
The cost is not the week
It is tempting to treat a substituted week as a small loss. One week, a nicer site, no harm done. The real cost is compounding, and it shows up in two ways.
The first is that the constraint stays in place while you get further from the moment when acting on it was cheap. The second is subtler: every substituted week produces the feeling of having tried. After enough of them, an operator concludes that the business does not work, when what actually happened is that the business was never tested. The verdict they are reacting to was never delivered, because the work that would have delivered it kept getting postponed for something more comfortable.
That is the argument for asking the question at the start of the week rather than at the end of the year.
Common questions
How do I tell avoidance from legitimate maintenance?
Ask what the work unblocks, specifically. Legitimate maintenance has a concrete thing on the other side of it that you have actually been trying to do and could not. If the justification is only that the result will be cleaner or easier later, it is probably not the constraint right now.
What if I genuinely do not know what my constraint is?
Then finding out is the work, and it is usually a matter of talking to people rather than building. Not knowing the constraint is itself a diagnosis — and it is the one situation where every building project is a guess, no matter how well executed.
Isn't some comfortable work necessary to avoid burning out?
Yes, and it is worth doing deliberately rather than pretending it is strategy. The damage comes from mislabelling it. A week spent on something pleasant because you needed a break is fine; the same week logged as progress is what accumulates into a wrong conclusion about the business.
Every property in the Don Gastón portfolio is independently live — built, deployed, and operated by one person.
See what's live in the portfolio