Most software products are built through a process of Accretion. We add a feature because a competitor has it. We add another because a single customer asked for it. We add a third because the lead developer felt like experimenting with a new framework.
The result is a “Franken-Product”—a bloated, confusing mess that provides less value the more it grows.
The Sovereign Architect knows that every new feature is not an “asset.” It is a Liability. It is more code to maintain, more bugs to fix, and more cognitive load for the user.
The Value-Complexity Trap
There is a common delusion that “More Features = More Value.” In reality, the relationship is a bell curve. Beyond a certain point, every new feature decreases the total utility of the product by making the core value harder to find.
The “iPhone” Fallacy
In the original 2012 notes, there was a question: “Do customers know what they want?” The answer is usually No. They know their Problems, but they are terrible at imagining Solutions. If you ask a customer what they want, they will ask for “Faster Horses” (more features). If you observe their behavior, you will see they need a “Car” (a different outcome).
The Protocol: The Power of Less
To build software that matters in a world of infinite noise, you must apply the Protocol of Subtraction:
1. The “Kill-a-Feature” Bounty For every new feature you propose, you must identify one existing feature to remove or hide. This forces the team to acknowledge that “Surface Area” is a finite resource.
2. Optimize for “Time to Value” (TTV) The only metric that matters in 2026 is how quickly a user can achieve their desired outcome after opening your app. If a feature adds 10 seconds to the TTV, it must provide 10x the value to justify its existence.
3. The “No” by Default Most founders are afraid of saying No to customers. But saying Yes to everyone is the fastest way to build a product for no one. Your roadmap should be a graveyard of “Good Ideas” that were sacrificed to keep the “Great Idea” alive.
#DhandheKaFunda: Excellence in software is not when there is nothing left to add, but when there is nothing left to take away. Build a tool, not a toy.