The Floor Moves Quietly

Most teams don't lose their baselines in a dramatic failure. They lose them on a Thursday.

Companies begin fragile by physics, not by bad decisions. Process lives in people's heads. Handoffs fail. Systems drift. Most things only work because someone happened to be paying attention at the right moment, and that attention is never guaranteed twice. Watch a company scale from fifty to two hundred employees in eight months: the pricing spreadsheet that three people maintained quietly starts producing different numbers depending on who runs it and which tab they trust. No single moment breaks anything. But by month eight, nobody can explain how a deal gets priced, and the people who once could have are too buried to notice the question is even being asked.

Growth doesn't fix that fragility. It replicates it, over and again.

So you build a floor. You measure something. You make it non-negotiable.^[2]

We used to check a dashboard every morning. Two numbers that mattered for the quarter: pipeline coverage and conversion rate. Updated overnight. Green, amber, or red. It wasn't sophisticated, which was the point. It was the one place in the day where something didn't negotiate back.

Then a launch slipped. Some enterprise escalations stacked up. A board deck was due Friday. The dashboard was the least urgent thing on anyone's plate, and urgency is a perfect alibi.^[1] Each decision made sense in isolation. There was no obvious wrong turn. Just a week of right turns that left the floor somewhere lower than anyone remembered.

Nobody called a meeting about it. Nobody had to. The dashboard kept glowing, the numbers kept updating, the only thing that changed was that we stopped looking.

Entropy in a dashboard you've stared at for a year doesn't feel like loss. It feels like Tuesday.

For a while the explanations arrived on their own. Maybe we'd outgrown it. Maybe the real work was happening elsewhere. They came quickly, almost helpfully. The baseline didn't break. We simply stopped looking at it.

That's how it usually happens. A script fails silently. The weekly review gets canceled once and never reinstated. Someone leaves and takes the context with them. The metric survives longer than the story that made it meaningful. Eventually someone shares a screenshot in a channel meant to celebrate progress, and a single reply appears beneath it: "Wait, is this right?"

The question lands heavier than it should. Not because the number is wrong, but because no one can explain why it's right.^[4]

There's a particular silence in a standup meeting when drift has already taken hold. Someone mentions that a number looks off. A few people glance at each other. The project lead says "we should look into that" with the specific intonation that means nobody will. Three weeks later, a client calls about it, and the scramble begins with the phrase that always signals trouble: "Wait, wasn't this supposed to be handled?"

After a missed launch window, the retrospective reveals the predictable finding: the bug itself didn't sink the timeline. The assumption that someone else was handling a key dependency did. Everyone thought someone owned it. No one did. The room goes quiet for a beat, and then someone types a note about "clarifying ownership going forward," which is the professional version of admitting the system ran on goodwill and got unlucky.

Ownership doesn't need a hero. It needs a name and a timestamp.

I once tracked seventeen metrics in a quarter, which is a good way to feel rigorous while the one number that mattered goes quiet.

I had a small cypress tree in a planter on our back patio. It had no business surviving where I live. The winters were long. The light went thin. I wheeled it inside when frost threatened. Some years it fruited. Some years it dropped its leaves and stood bare for weeks, a stick in a pot.

It would've been easy to miss a watering. Nothing dramatic would've happened immediately. The tree wouldn't have died on any given Thursday.

It died anyway.

Not all at once. Slowly, leaves thinned, and branches hardened. By the time I admitted it was gone, the change had been underway for months. I'd been tending something already past the point of recovery. Care had turned into ritual. The floor had moved, and I was still standing where I thought it was.

Some of the most important work in any company doesn't have a name. An operations manager spends fifteen minutes every Thursday updating a shared doc with weird customer edge cases from the week: invoice adjustments, billing exceptions, the customer who somehow has two to five accounts depending on which system you check. Tedious work that disappears into company motion. Until a new hire joins and can handle the strange cases without escalating, because someone wrote them down. Nobody thanks the person who maintained that doc. Most people don't know it exists.

That work won't show up in sprint points. It keeps the week from unraveling, and its absence only becomes visible when someone stops doing it. The launch drags. The fix doesn't stick. The number's off again.

The dashboard eventually got retired. No one asked why. New metrics, new priorities, new colors. Maybe they were better. Once you stop measuring, you lose the ability to tell whether what replaced it is improvement or accommodation. You're simply somewhere new, and the organization has already written the story of how it got there.^[3]

What disappears with the baseline isn't the number. It's the friction. The thing that couldn't be talked around in a planning session or smoothed over in a board update. Dashboards don't get retired because they stop working. They get retired because they keep working, and eventually that becomes inconvenient.

Some mornings I don't check.

And the forgetting feels the way it always feels, which is to say it feels like nothing at all.