The projector hums with a low-frequency vibration that feels like it’s trying to shake the fillings out of my teeth… ‘Wei, I need you to pull the numbers. We need the data to show the board exactly why this is the only logical move. Have a deck ready with at least 13 slides by Friday.’
Data is the makeup we put on the corpse of a dead decision.
– The realization of the Foley Artist in the machine.
I’ve spent the last 3 years in this building, and I’ve realized that being a ‘Data Scientist’ in a corporate environment is often just a fancy term for being a Foley artist. I think about my friend, Wei T.J., quite a bit in these moments. Wei T.J. is a professional Foley artist for independent films. He spends his days in a dark studio surrounded by piles of junk-old shoes, rusted chains, dried celery, and sheets of corrugated metal. If a character in a movie walks across a gravel path, Wei T.J. isn’t recording actual gravel. He’s probably squeezing a leather pouch filled with 23 marbles. It sounds ‘more real’ than the real thing. It provides the texture the audience expects so they can stay immersed in the story.
My job at this company is fundamentally the same. Jenkins has a story he wants to tell the board, and I am being asked to provide the ‘crunch’ of the gravel. I’m not searching for the truth; I’m recording the sound of a bone breaking using a head of lettuce because the actual truth is too quiet, too messy, or too inconvenient to be heard.
Data-Justified, Not Data-Driven
It’s a cynical way to live, isn’t it? I’ve tried to turn my brain off and on again, hoping to reset this perspective, but the glitch persists. Every time I open a raw CSV file, I see the possibilities. Data, in its purest form, should be a mirror. It should show us the wrinkles, the blemishes, and the structural weaknesses of our strategies.
Truth & Discovery
Justification & Shield
But in the 83 meetings I’ve attended this quarter, data has been used as a shield. It’s used to provide ‘cover’ for gut feelings. If a VP wants to launch a product, the data will miraculously find a way to support it. If they want to kill a project, the same data will be tortured until it confesses to a lack of ROI. We are obsessed with being ‘data-driven,’ but we are actually just data-justified. The decision has been made in the gut, in the ego, or in the golf course conversation, and the numbers are just the formal attire required for the wedding.
The 63-Hour P-Hack
I remember one specific mistake I made early on. I actually tried to tell the truth. We were looking at a 23 percent drop in user retention for our legacy platform. I brought a heat map to the leadership team that showed exactly where we were losing people. It was glaring. It was a 53-page report of failure. I thought I was doing a great job. I thought, ‘This is it, they’ll see the problem and we can fix it.’
Jenkins looked at the screen for 3 seconds, cleared his throat, and said, ‘The database must be corrupted. Wei, go back and filter for users who have been with us for more than 13 months. I think you’ll find the story is different there.’ He didn’t want the truth. He wanted a version of the truth that allowed him to keep his current roadmap. I spent the next 63 hours p-hacking my way to his desired conclusion. By the time I was done, I had filtered out so much reality that the remaining data points were basically just a digital ghost story. I felt like Wei T.J. recording the sound of a thunderstorm by shaking a piece of tin. It sounded like thunder to the board, but I knew it was just a cheap trick in a dark room.
The Cathedral of Logic on a Foundation of Whims
When data is used as propaganda, it loses its utility as a tool for discovery. If we already know the answer, why bother with the math? It creates a culture of profound cynicism.
The junior analysts know the game. The managers know the game. Even the board likely knows the game, but everyone agrees to pretend that the charts are objective. We’ve built a cathedral of logic on a foundation of whims. It’s exhausting. You can only turn your integrity off and on again so many times before the power supply fries. I’ve reached a point where I can predict the outcome of a ‘data-deep-dive’ before the first SQL query is even written. If the project belongs to a favorite of the CEO, the numbers will be 103 percent positive. If it belongs to a rival, we’ll find 333 reasons why it’s a systemic risk.
The Uncompromising Auditor: Engineering Reality
There is a massive difference between this corporate theater and the world of actual, tangible reality. In engineering, physics is the ultimate auditor. You can’t ask an engineer to ‘pull the numbers’ to show that a bridge will stand when the gravity says otherwise. This is why I find companies like prefab house supplierso fascinating from a distance. Their work is grounded in the uncompromising reality of modular construction and engineering.
A bolt doesn’t care about your quarterly goals. A load-bearing wall doesn’t change its properties because a VP had a dream about the European market.
In the corporate world, however, we treat data as if it were soft. We bend it. We stretch it. We apply filters like we’re on Instagram, trying to make the profit margins look a little bit more vibrant than they actually are. I recently sat through a presentation where a director claimed we had a 93 percent customer satisfaction rate. When I looked at the raw data later, I realized they had only surveyed 3 people who had already given us five-star reviews. It was a statistical hallucination.
The Sound of Silence
Wei T.J. once told me that the hardest sound to fake is the sound of silence. He said that when a movie scene is truly silent, the audience gets uncomfortable. They start to hear the popcorn in their teeth or the person breathing next to them. So, he has to record ‘room tone’-a subtle, almost imperceptible hum that tells the brain, ‘It’s okay, the story is still happening.’
Corporate data is often just room tone. It’s a background hum of charts and percentages that keeps the investors from getting uncomfortable. We are terrified of the silence that comes with saying, ‘We don’t know.’
We are terrified of the gap between our ambition and our evidence. So we fill that gap with 43-page PDFs and 233-row spreadsheets. I’ve started to wonder what would happen if we just stopped. What if, the next time Jenkins asked me to justify a pre-made decision, I just stood up and made the sound of a car crash with my mouth? It would be just as honest as the report I’m about to write.
Walking Toward the Cliff
I’ll go back to my desk, I’ll open Excel, and I’ll start looking for the 13 percent of the market that makes his European dream look like a stroke of genius. I’ll be the Foley artist he needs me to be. I’ll make sure the gravel crunches perfectly under his feet as he walks toward a cliff.
Desired Cultural Shift
40% Progress
We need to reward the analyst who finds the 33 reasons why a project will fail, rather than the one who finds 43 reasons why it might succeed.
But maybe there’s a way out. Maybe the ‘turned it off and on again’ solution isn’t about the computer, but about the culture. We need to start valuing the data that tells us we’re wrong. We need to move away from ‘pulling numbers’ and toward ‘asking questions.’ Until then, we’re all just in a dark room, squeezing leather pouches and pretending we’re walking on mountains.