The most powerful way to build a business is to focus on a 'divine lever'—an action that is both a causal force for growth and is objectively good, serving demand purely without self-centered motives. This creates a sustainable, meaningful foundation for a company.
A primary cost of focusing on a 'divine lever' isn't financial delay, but social and professional illegibility. Your business decisions, shaped by this internal conviction, will not make sense to outsiders operating with traditional business logic, making it difficult to explain your strategy.
Unlike typical business activities that expend energy and require recharging, focusing on a 'divine lever' or an objective good is energizing. This intrinsic motivation pulls you through challenges and fuels obsession over the long term, creating a sustainable advantage against burnout.
Constantly focusing on your one 'divine lever' acts as a global optimizing function. It forces a unique and often strange set of business decisions that, while confusing to outsiders, creates a company shape that is perfectly and defensibly fit to serve demand over the long term.
Any degree of self-interest mixed with the goal of serving customers creates a 'decay function.' Over time, this small percentage of selfishness compounds through countless micro-decisions, inevitably degrading the quality and integrity of the final output. Purity of intent is a functional requirement for long-term excellence.
In crowded markets, founders mistakenly focus on other startups as primary competition. In reality, most customers are unaware of these players. The real battle is against the customer's status quo: their current tools like spreadsheets, hiring a person, or using an old system. Your job is to beat those options.
Founders often quit for the wrong reason: struggling to schedule meetings, which is merely a lack of data. The true signal to pivot or quit is when you've successfully engaged potential customers who have clear demand (pull) and they still explicitly reject your solution after multiple iterations.