For three months, every conversation at this company ended in the same place.

October.

Why can't we get back to October? What changed since October? Who do we need to hire to recover what October looked like?

The CEO — I'll call him The Optimizer, because that's genuinely what he was trying to do — had a spreadsheet. Six KPIs. Color-coded. Financial model built around returning to the October baseline.

He was doing exactly what a CEO is supposed to do: managing to data, making decisions based on benchmarks, holding the team accountable for the gap.

So accountable, in fact, they'd blown through 3 marketing leaders in 1 quarter.

Nobody thought to ask whether October was real.

I came on in January. By February, I was reviewing the first full month of numbers with the team.

Cost per booked call: ~$300. Decent. 338 calls scheduled. Okay, not bad.

Then I kept reading.

58 calls actually finished. Out of 338. Resulting in 12 purchases. A $19,000+ cost per acquisition.

Worst numbers in the company's history.

I said it out loud on the call: "I think our pixel is poisoned."

Silence. The kind of silence where everyone's doing math in their head and not liking the answers.

But here's the thing... the pixel wasn’t the fire. It was the smoke alarm that finally went off.

When I got under the hood, it was like opening a closet and having everything fall on you at once.

Problem one: Cold traffic campaigns were routing new leads to the lead magnet funnel instead of the VSL. Ads say "book your strategy call," people land on a page offering a free course they never asked for. A team member caught it during a routine check: "Oh shit... they're going to the wrong place." Every conversion from those campaigns was measuring the wrong event.

Problem two: $40,000 spent on instant form campaigns with zero conditional logic. Meta was getting a signal every time someone filled out a form... whether they were a baller business owner or someone's uncle who fat-fingered an ad on the toilet. FORTY GRAND. To build a machine that finds form-fillers. Incredible.

Problem three: Google Tag Manager hadn't been touched in months. The person who built it was long gone. Tags on wrong funnels, some funnels tagged twice, no labels on anything. The retargeting meant to recapture qualified leads? Pulling from everyone who'd hit any page, regardless of how they got there.

The dashboard was showing numbers. Numbers feel like information even when they're not.

Here's why this matters beyond a pixel problem.

Every layer of measurement was broken. And every layer had been broken for long enough that the broken version was the baseline.

So when October looked like a golden month... most of those attributed results were sitting on a corrupted signal the platform had been training on for months. What looked like a winning month was mostly noise that had been labeled as signal. The Optimizer wasn't trying to recover October.

He was trying to rebuild a crime scene.

And this is where it gets painful to think about.

The CMO who got fired? She was optimizing to the numbers she had. She wasn't wrong about her strategy. She was wrong about what the numbers meant. Which is a totally different problem... one she had no way to see from her seat.

The sales team working garbage leads and getting blamed for low close rates? They weren't underperforming. They were calling people who'd accidentally clicked a form ad while scrolling Instagram.

Real people, real careers, real stress. All held accountable for numbers that were mostly fiction.

I've been the person in that seat. (More than once, if I'm honest.) The person brought in to "fix marketing" where marketing was never the thing that was broken. It's a special kind of awful — you can feel something's off, you're executing the playbook, and the numbers keep telling you you're failing. It messes with your head.

Every company I've been inside of has their own version of this doom loop. And it always runs the same way.

Something works. (Or appears to.) It becomes the baseline. The baseline becomes the target. When results dip, the team goes looking for someone to blame or something to change... but the one thing they never question is the baseline itself.

So they churn. New marketing lead. New agency. New strategy. New hire. Each one walks in, inherits the same broken measurement, gets held to the same fictional standard, and flames out. The company reads this as "we keep hiring the wrong people." What's actually happening is they keep solving yesterday's problem with yesterday's logic... using the same instruments that lied to them in the first place.

Three marketing leaders in one quarter. That's not accountability. That's a doom loop.

Drucker said it decades ago: "There is nothing quite so useless as doing with great efficiency something that should not be done at all." He was talking about this exact loop.

The only thing worse than a bad decision is a confident bad decision.

Bad data that looks like bad data is survivable. The number is obviously wrong, someone flags it, you fix it. Tuesday.

Bad data that looks like good data is a completely different animal.

When your measurement is broken and it isn't obvious, every downstream decision becomes a confident move in the wrong direction. You don't question the baseline. You try to recover it. You don't audit the pixel. You audit the CMO. You don't ask whether October was real. You ask who you need to fire to get back there.

And the brutal part? It all looks like rigorous management from the outside. Spreadsheets. Color-coded KPIs. Accountability conversations. I know there are founders reading this right now who would kill for that level of organization. It has all the right aesthetics of a company that knows what it's doing.

That's what makes the doom loop so dangerous. It doesn't look like chaos. It looks like discipline. It feels like you're doing the hard thing — making tough calls, holding people accountable, managing to the numbers. But if the numbers are wrong, discipline just means you're marching faster in the wrong direction.

I've written before about my 4-layer Problem Ping-Pong mental model. Layer 3 is: fix the seeing before the doing. One sentence. I hope you didn't move past it too fast.

This is what skipping Layer 3 actually costs. Not just "imperfect data." Months of confident decisions built on a foundation that was cracked before anyone showed up to work. People losing jobs over numbers that were mostly fiction. A financial model built around restoring a month that never existed the way anyone thought it did.

Here's what I'd ask if I were sitting across from you right now:
“What's your October?”

What's the number, the month, the benchmark your team keeps trying to recover... that you've never actually pressure-tested? Go pull the thread. Look at the inputs behind it. Ask your marketing leader where the conversions are actually coming from, not what the dashboard says they're coming from. You might not like what you find. But it's cheaper to find it now than to fire three people over it later.

Who in your company is responsible for making sure what you're measuring is real? Not who could do it if pressed.

Who owns it.

If the answer is "everyone, generally"...

It's nobody, specifically.

—Chris Piper
The Growth Operator