Edition #4: We don't have time to fix what's broken

Edition #4 of The Exceptional OS. ~7-minute read.

Last fall, I was on a call with the founder of a services business doing about $12 million in revenue. We had been working through his operations for an hour. He had named six different broken things in the business clearly and precisely, by their first names, the way you talk about people you have lived with for years.

1️⃣ The handoff between sales and delivery that loses two days every time.

2️⃣ The reporting that requires three people to reconcile manually on the last Friday of the month.

3️⃣ The onboarding that varies depending on which project manager picks up the new client.

4️⃣ The vendor that bills late so finance is reconciling the wrong month.

5️⃣ The proposal template that produces a different price than the actual contract.

6️⃣ The Slack channel that everyone joined two years ago and nobody reads anymore, but where critical decisions sometimes get made.

He knew all of them. He was articulate about all of them.

Then he said it.

"We know all this. We just don't have time to stop and fix it."

I have heard that line in different forms from probably 200 founders over the years. Sometimes it is "we will get to it next quarter." Sometimes it is "we are in growth mode." Sometimes it is "the team is heads-down on a launch."

The variations do not matter. The underlying decision is the same. The operator has correctly identified the problem and has consciously decided to defer the fix.

I want to spend this letter on why that decision is more expensive than the spreadsheet shows, and on the one decision that breaks the pattern.

The conversation went like this.

I asked him how many hours per month the team spent on workarounds for those six things. Not solving the root cause, but just keeping the wheels turning while the broken thing kept being broken.

He thought about it. He has a senior team, so he pays attention to where their time goes. The number landed at roughly 100 hours per month across the team. Maybe a little more.

100 hours a month is 1,200 hours a year.

A senior team member at a $12M services company has roughly 1,500 to 1,600 productive hours per year after holidays, sick days, and time spent on things that are not direct work. 1,200 hours is 75% to 80% of one of those people. The founder was paying for the better part of a full senior employee, every year, indefinitely, to fight the same six fires he had just described to me in such precise detail.

He had not done that math. Founders rarely do. The math feels obvious in retrospect, in the same way every math problem feels obvious once somebody else solves it for you. In the moment, when you are looking at the bank balance and the sales pipeline and the team capacity and the customer escalations, the workaround feels cheap. You feel like you are saving money by not pausing to fix things.

You are not saving money. You are renting an FTE you cannot see.

What the math also does not show is that the cost is not flat. It compounds. I have watched the same pattern play out in enough companies that I can describe its shape.

Stage 1: Visible but tolerated. The broken thing is known. The team works around it. Everyone agrees it is annoying. Nobody is panicking. The workaround feels mature, even sophisticated. "This is how grown-up companies handle things."

Stage 2: Firefighting becomes the job. The workaround has multiplied. Now there are three or four workarounds layered on top of each other because the original workaround broke something downstream. Senior people are spending their week putting out fires instead of doing the work they were hired for. Their calendars no longer reflect their job descriptions.

Stage 3: A key person leaves. The one person who really understood how the workarounds connected together decides they have had enough. Maybe they leave for another job. Maybe they get sick. Maybe they just quietly disengage. Whichever it is, the institutional knowledge of how the broken thing is being held together walks out the door.

Stage 4: Panic. Things start breaking in production. The team hires three people. Two of them do not last because they cannot make sense of the workarounds without the knowledge that left. The third one inherits a system she did not build, designed by people she will never meet, and she is now expected to make it work.

Stage 5: The fix is now expensive. The thing that would have cost two engineers two weeks to fix at Stage 1 now costs a six-month migration, a consultant, a hiring round, and a customer-facing apology cycle. The bill is roughly three times what it would have been. Sometimes more.

The companies I have watched directly move from Stage 1 to Stage 5 in something like 90 days. That number is directional, not statistical. I have not measured it scientifically. But the pattern is real, and the compression is faster than founders expect. When you are inside it, Stage 1 feels stable for a long time. The transition to Stage 5 feels like it happened overnight.

It did not. You were just inside the slow part.

Here is what I told the founder.

You cannot fix six broken things this quarter. You have the budget for one. Not because the budget is small but because attention is scarce. Six fixes split across a busy team is six half-finished projects in 90 days, plus six new workarounds layered on top of the old ones.

So pick one.

Not the easiest. Not the loudest. The one whose Stage 5 will hurt the most. Look at his list. The sales-to-delivery handoff that loses two days every time. That one is a Stage 5 candidate; the first customer to discover the gap on a high-stakes project will cause a renewal crisis. Pick that one.

Now do three things with it.

- Name a specific owner. Not a committee. Not "the ops team." A person, with a calendar, who is accountable.

- Give them a deadline. Not "by end of quarter." A date.

- Carve out their time. Their other work has to give. If you cannot carve time, you have not actually picked anything. You have made a wish.

The other five fires keep burning this quarter. They will burn cheaper than the cost of pretending you are addressing all six.

Next quarter, you pick the second one. Same rules.

This is not a productivity playbook. It is a survival playbook. You are not optimizing the company. You are stopping the company from arriving at Stage 5 on six fronts simultaneously, which is what happens to most $10M to $30M services companies that try to fix everything at once and end up fixing nothing.

He picked the sales-to-delivery handoff. He named his head of operations as the owner. He gave her until the end of Q1. He cut three meetings from her calendar to make room.

Three months later, the handoff was clean. The team recovered about 80 hours per month of senior time. Some of that time went to the next thing on the list. Some of it went to actual rest, which is itself worth measuring.

The next quarter, he picked the reporting reconciliation. Same rules. Three months later, that one was fixed too.

A year in, four of the six were resolved. The other two had become smaller in proportion to everything else and could wait. He never tried to fix all six at once. He let the math do the work.

The company is not bigger as a result. The company is more durable. Those are not the same thing.

The expensive lesson is not that he had broken things. The expensive lesson is that he had a name for each of them and was still choosing every quarter to pay for the workaround instead of the fix. The math was running in the background the whole time.

It runs in the background of every operator who has not yet done it.

Reply with one thing. What is the one broken thing in your operation you have been "too busy to fix" for more than 90 days? You do not have to fix it this quarter. But naming it out loud, to your team or to me or to a notebook, is the move that turns a workaround into a decision.

I read every reply.

Iulian

P.S. If you read this and a specific broken process came to mind in the first 5 seconds, that is a Future-State Audit signal. The audit is a structured 60-minute conversation that names the broken things, sizes their math, and picks the one fix that breaks the pattern for your quarter. Details at iulianchiriac.com/audit when you are curious. No pressure.