The hum of the projector fan suddenly feels like a jet engine. No one is speaking. Chloe’s slide is still up on the screen, a masterpiece of pivot tables and regression analysis that glows with the undeniable, mathematically-sound conclusion that Project A will yield 49% more revenue. She spent 199 hours on it. And Stan, the Senior VP of Everything Important, just squinted, steepled his fingers, and said the words that vaporized all that work: ‘Interesting. But my gut tells me we should go with Project B.’

And just like that, the data was dead. The decision was made. The rest of the meeting would be a performance, a collective exercise in finding charts that could be twisted just enough to support Stan’s gut. We aren’t a data-driven company. No, that’s what we print on the recruiting posters. We are a data-supported company. We make decisions with the same old stew of ego, fear, and political calculus, then we send the analysts scurrying back to their caves to find data that proves we were right all along. The dashboard isn’t a compass; it’s a decoration. It’s the expensive art we hang on the wall to make ourselves look sophisticated while we still eat junk food on the floor.

I just spent an hour writing a whole paragraph detailing the cognitive biases at play here-confirmation bias, authority bias, the sunk cost fallacy-and then I deleted it. It felt academic and hollow. You don’t need a textbook to understand what’s happening. It’s about power. A spreadsheet cannot have a gut feeling. It has no vested interest in a project that was its pet idea nine months ago. Stan does. A spreadsheet doesn’t get a bonus if a certain initiative, regardless of its actual performance, is perceived as a success. Stan does. The data is impartial, and in a corporate hierarchy, impartiality is a threat to the people who have built their careers on being partial to their own ideas.

The Raw Data vs. The Executive Caption

I know a guy, Michael L.M., who works as a live closed captioning specialist for a major news network. His job is one of the purest forms of data transmission I can imagine. He hears words, and he transcribes them. Fast. There’s no room for interpretation. If a politician accidentally says ‘nucular’ instead of ‘nuclear,’ Michael types ‘nucular.’ His job isn’t to fix the error, or interpret the intent, or make the speaker sound smarter. His job is to present the raw data of the speech, exactly as it was delivered. He once told me about a new hire who lasted less than a day. The guy kept correcting the grammar of a senator during a live hearing. He was fired within 9 minutes. He thought his job was to improve the message; his job was to transmit it.

Most executives are that new hire. They see the raw data-the transcript of reality-and their first instinct is to ‘clean it up.’ They reframe it, they add their own captions, they massage it until it says what they wanted it to say in the first place. ‘Customer engagement is down 19%.’ That’s the raw data. The executive caption reads: ‘Due to a strategic realignment of resources, we have observed a temporary and expected modulation in our user interaction metrics as we pivot toward higher-value cohorts.’ See? It sounds much better. It just happens to be a lie.

I hate this. I find it intellectually offensive. And yet, I did the exact same thing nine years ago. I was running a marketing team for a software company, and we’d just launched an expensive campaign that was, by all accounts, my baby. I’d championed it for months. The initial data came in, and it was awful. Our primary metric, new subscriptions, was down. Not flat. Down. But I went digging. I kept digging until I found a vanity metric, something buried deep in the analytics platform-‘average time on solutions page’-that was up by a measly 9%. I took that one, single, glistening data point and I built an entire presentation around it. I stood in front of the leadership team and, with a straight face, presented a story of ‘deepening engagement’ and ‘quality over quantity.’ I used the data not as a flashlight to find the truth, but as a drunk uses a lamppost: for support, not illumination.

They bought it. Because they also wanted my project to be a success.

The Insidious Nature of Shared Delusion

This is the insidious nature of the data-supported lie. It’s often a shared delusion. Everyone tacitly agrees to believe the story because the alternative-admitting failure, changing course, challenging a powerful person’s gut-is far more uncomfortable. It’s far easier to build a narrative around a 9% increase in a meaningless metric than to confront the reality of a 19% drop in the one that keeps the lights on. Building a system with integrity, where the rules and the data are applied consistently, is brutally difficult. It requires a culture that values truth over comfort.

This principle applies everywhere, from corporate governance to the complex algorithms that run responsible entertainment platforms like gclubpros. In both cases, if the underlying system can be manipulated by ‘gut feel’ or favoritism, trust collapses entirely.

What a Genuinely Data-Driven Culture Looks Like

So what does a genuinely data-driven culture look like? It’s not about having more dashboards. In fact, it might be about having fewer. It’s about the behavior that happens when the data delivers bad news. What does Stan do when Chloe’s chart shows that his pet project, Project B, will cost the company $239,000 in lost opportunity? In a data-supported culture, he challenges the data. He asks for more cuts, different timeframes, he pokes at the methodology until he finds a crack he can use to dismiss the entire analysis. In a data-driven culture, he says, ‘Thank you, Chloe. This is not what I wanted to see, but it’s what we needed to know. Let’s dig into Project A.’

It’s also about empowering people like Michael L.M., the data transcribers. The people closest to the raw information. A truly data-driven organization doesn’t just listen to the analyst in the room; it gives them a voice that can’t be muted by the highest-paid person’s opinion. Their job is to deliver the unvarnished transcript of reality, no matter how inconvenient. A few years ago, we had a project manager, a quiet guy named David, who was responsible for tracking the bug reports on a new software launch. The executive team was getting ready to issue a press release patting ourselves on the back. 29 minutes before it was scheduled to go out, David sent an email to the entire leadership team with a subject line that just said: ‘STOP.’ The body of the email was a single chart showing a catastrophic, exponentially-growing memory leak that was bricking about 9% of our users’ machines. It was an inconvenient truth delivered without apology. It delayed the launch by weeks and cost a fortune in overtime. It also saved the company from total humiliation and a potential class-action lawsuit. David didn’t interpret. He just showed us the transcript. We need more Davids.