The author argues that software carries the perspective of its creators, reflecting their experience, values, and way of thinking. They note that while industry discourse often treats software as a factory output—gathering requirements, building features, measuring results—the reality is that two teams with similar inputs can produce different products because of how they interpret what they see and their individual styles. The author points out that founder background and company culture heavily influence this outcome.
At Linear, the author says they have valued intuition over strict process or data. They clarify that intuition is not arbitrary preference but develops through repeated work, customer conversations, and learning from mistakes. This intuition represents a point of view that guides judgment rather than hiding behind faceless metrics. Before launching Linear, the team spoke with people across various companies and roles about their work habits. Those discussions, combined with the team’s own frustrations, helped broaden their perspective beyond personal experience.
The post explains that a product’s character emerges over time from what its makers choose to pay attention to, what they decide to omit, and what they continue refining after a feature technically works. The author likens this to a product’s “soul,” suggesting that users can sense when someone cares about the overall cohesion of the product. They warn that a feature can be technically sound yet harm the product if it does not align with a shared vision of what the product should be. Maintaining that vision requires someone to make a call and stay close to customers to verify its correctness.
The author observes that many successful products now feel messy, as more effort goes into adding new capabilities than into improving the core problem. Because adding capabilities is easier than rethinking existing ones, products become collections of abilities lacking a unified point of view. They speculate that this trend may stem from work organized around autonomous agents or individuals operating like small software factories, where each change makes sense in isolation but together creates a disjointed whole reminiscent of Frankenstein’s monster rather than a carefully sculpted piece.
Without a shared perspective or soul, the product becomes an average of disparate outputs, diluting the original choices that gave it character. While more things may work, it becomes harder to sense a human presence behind the software. The author concludes that as software creation becomes easier, they hope it will give people more space to develop and follow a deep understanding of problems, shaping solutions where the pieces belong together, so that users can feel the creator behind the software rather than just the factory.