The Process That Annoys You Most Is Usually the Wrong One to Build

Ask an experienced operator which process in their organisation most needs fixing and the answer comes fast, with feeling. It is the quarterly reconciliation that ruins a week. The escalation that always arrives on a Friday. The report nobody can produce without three people and a shared screen.

Those are real problems and they are usually the wrong ones to build a business on.

Why pain is a misleading signal

A process becomes memorable through a combination of difficulty and infrequency. The quarterly reconciliation is agonising partly because it happens four times a year, which means nobody ever gets good at it, no tooling is ever justified, and every occurrence feels like the first.

That same infrequency is what makes it a poor product. Four events a year gives you a product somebody thinks about four times a year, which is a product they forget they are paying for and cancel at renewal. It also gives you four opportunities annually to learn anything, which is a brutally slow feedback loop.

Meanwhile the process that runs two hundred times a day is invisible precisely because it has been smoothed into the background. Nobody complains about it. Everybody depends on it.

The arithmetic nobody does

“A task that takes four hours quarterly costs sixteen hours a year. A task that takes ninety seconds three hundred times a day costs over three thousand, and nobody has ever sworn about it.”

Value to a buyer is roughly frequency multiplied by the cost of each occurrence, and operators systematically overweight the second term because that is the one they feel.

A task that takes four hours and happens quarterly costs sixteen hours a year. A task that takes ninety seconds and happens three hundred times a day costs over three thousand hours a year. The second is two hundred times larger and nobody has ever sworn about it.

The high-frequency process also has properties that make it a much better product. It generates enough events to show whether your product works within weeks rather than years. It is boring enough that incumbents have ignored it. And it is embedded deeply enough in the day that once it runs through your product, removing it is disruptive, which is a moat you did not have to build.

A task that takes four hours quarterly costs sixteen hours a year.

How to find yours

Stop asking what frustrates you and start asking what you do without thinking.

The practical method is to track a week. Not in detail, just a note each time you switch context, and particularly each time you move data from one place to another by hand. The things that appear repeatedly, that you never mentioned to anyone because they are too small to mention, are the candidates.

Then check the second condition, which is whether the task has a definable correct outcome. Many high-frequency tasks are judgement calls in disguise, and those are much harder to productise than they appear. The ones to want are frequent, boring, and unambiguous.

The objection

The obvious response is that small frequent tasks are small, and nobody will pay much to remove ninety seconds.

They will not, framed that way. They will pay to remove three thousand hours a year, which is the same thing stated at the scale a buyer actually experiences it. The error is not in the opportunity, it is in presenting it per occurrence rather than per year, and that is a positioning problem rather than a product one.

The frustrating quarterly process still deserves fixing. Fix it for your employer, where the sixteen hours are worth recovering. Build the business on the thing nobody noticed.

See operators building from workflow pain for the starting point, internal process experts turning hidden knowledge into product for the knowledge angle, and boring vertical tools beating AI wrappers for why unglamorous wins. The operator track covers the rest.