Tugging at the corner of cell AH44, I watch the revenue for June 2034 bloom into a figure so large it feels like a personal insult to my current bank balance. My fingertips are slightly numb from the repetitive rhythm of the trackpad. It’s 2:14 AM, and the laptop fan is whirring like a jet engine preparing for a takeoff that will never actually happen. This is the moment where the spreadsheet stops being a tool for calculation and starts becoming a work of speculative fiction, somewhere between a sci-fi novel and a prayer. I am projecting a churn rate of 1.4% for a company that currently has exactly 14 customers, three of whom are my cousins who I’m pretty sure will stop paying the moment I stop answering their texts.

The Hollow Victory of Conviction

Last week, I won an argument I was fundamentally wrong about. I was debating the merits of a 54-month scalability plan with a colleague, and I’d accidentally double-counted a specific tax credit. Even after I realized the mistake midway through the conversation, I didn’t stop. I doubled down. I was so precise with my delivery, so confident in my grasp of the underlying mechanics, that I watched the doubt leave his eyes. He believed me because I looked like a person who knew what the world would look like in 2034. It was a hollow victory. It’s the same hollow victory we all chase when we present these models to investors. We aren’t selling them a future; we are selling them the comfort of a person who is willing to lie to themselves with enough conviction to make it look like data.

Leo’s Honest Guess

I think about Leo R.J., a prison education coordinator I met a few years ago. Leo deals with a reality that is far more rigid and, frankly, more honest than the tech world. In the prison system, budgets are a battlefield of 14-point fonts and mandatory minimums. Leo once told me that the only way he could get funding for a vocational training program was to project recidivism rates 14 years into the future. He knew the data was a guess. The wardens knew it was a guess. But without the guess, the money didn’t move. The ritual of the forecast provided a safe container for the risk. It’s a performance of clairvoyance that punishes humility. If Leo had said, “I honestly don’t know how these 24 men will respond to a coding class in a decade,” the program would have been killed in 14 seconds. Instead, he gave them a chart. He gave them a number ending in 4. He gave them the fiction they needed to authorize the truth.

We are all doing a version of Leo’s dance. The financial model is a psychological test, not a predictive one. When an investor looks at your 2034 projections, they aren’t checking your math-they’re checking your grasp of the levers. They want to see if you understand how a 4% increase in retention impacts your cash flow in month 44. They want to know if you’ve thought about the friction. It’s about the wiring of the business, even if the house hasn’t been built yet. The danger is when we start to believe our own blueprints are the actual walls. We get so caught up in the elegance of the formula that we forget that next Tuesday might involve a global server outage or a competitor launching a better version of our product for $4 less than our cost.

The Texture of Reality

I remember another time I was wrong. I had projected that marketing costs would plateau after the first 14 months of operation. My logic was that organic word-of-mouth would take over. It was a beautiful theory that ignored everything we know about human attention spans and the ruthlessness of the ad-bidding market. I defended that plateau for 44 minutes in a board meeting. I won that argument too. And then, 14 months later, the reality hit, and the CAC didn’t plateau; it spiked by 64%. The model was a shield that kept us from seeing the sword coming. We had built a plan for a world that didn’t exist, and because the plan was in a spreadsheet, we treated it as more real than the market feedback we were actually receiving.

This is the paradox of the forecast. You must build it to prove you are serious, but you must be prepared to burn it the moment the ink is dry. It’s a map of a territory that is constantly shifting its borders. If you follow the map too closely, you’ll drive right off a cliff that wasn’t there when the satellite took the photo. The most successful founders I’ve met are the ones who can hold the 5-year vision in one hand and the 14-day reality in the other without letting the two touch. They treat the model as a living document, a series of ‘what-ifs’ rather than ‘will-bes’.

Maybe the real value of the 2 AM Excel session isn’t the output. Maybe it’s the process of forcing your brain to inhabit a future that doesn’t exist yet, so that when pieces of it start to arrive, you aren’t completely surprised. It’s a mental rehearsal. Like Leo R.J. and his 14-year projections, we are just trying to find a way to justify taking the next step in the dark. We are looking for a reason to believe that the risk is calculated, even when the calculator is being operated by a person who hasn’t slept in 24 hours.

I close the laptop. The blue light fades from my retinas, leaving a ghost of the hockey stick burned into my vision. I’ll wake up in 4 hours and look at it again. I’ll probably change that churn rate back to 2.4% just to feel a bit more honest. Or maybe I’ll leave it at 1.4% and keep telling myself the story until I believe it. After all, if I don’t believe in my fictional 2034, who will? The forecast isn’t about the money. It’s about the audacity to imagine a world where things actually go according to plan, even if only for 44 minutes at a time.