Superhuman

The Natural Language Interface Revolution Software complexity has reached a breaking point. As digital products bloat with features, the gap between human intent and machine execution widens. James Evans, the CEO of CommandBar, recognized this friction early. While most companies were busy building better dropdown menus, Evans aimed for a paradigm shift: a natural language interface for software. The vision wasn't just to add a search bar, but to fundamentally change how humans communicate with their tools. Starting in the Y Combinator batch of 2020, CommandBar began as a specific "command palette" interface—the Command K shortcut popularized by high-productivity tools like Superhuman and VS Code. Evans notes that early on, the product was a "wedge" into a much larger problem. By allowing users to describe what they wanted to do in plain English rather than hunting through navigation trees, CommandBar began bridging the gap between novice users and power users. This transition from rigid UI to fluid, intent-based interaction represents the next frontier of user experience (UX). Pivoting from Category Creation to Market Disruption One of the most profound insights Evans shares involves the strategic choice between creating a new category or entering an existing one. For the first few years, CommandBar positioned itself as a category creator—a "search bar for any app." While this resonated with tech-forward founders, it fell flat with 99% of the broader market. The breakthrough came when Evans decided to reposition CommandBar as a new entrant in the "Digital Adoption" space, a market historically dominated by legacy popup and tutorial tools. By entering the G2 grid for digital adoption, CommandBar didn't just change its label; it simplified the buyer's journey. Instead of spending fifteen minutes of a thirty-minute discovery call explaining what the product was, sales teams could spend ninety seconds positioning themselves against the "annoying popups" of incumbents. This positioning allowed them to capture existing demand from teams already looking to solve user friction, while offering a vastly superior, AI-driven solution. It proves that sometimes, you don't need to invent a new name for a problem; you just need to offer a better way to solve the one people already have. The Amplitude Pricing Strategy and Finding PMF Product-market fit (PMF) is often treated as a binary state, but for CommandBar, it was measured through the lens of pricing power. Evans credits the "Amplitude pricing game" for helping the company realize its true value. The strategy is simple: start at a baseline price and double it in every subsequent sales conversation until you hit significant resistance. This aggressive experimentation revealed that CommandBar's value proposition resonated far more deeply with companies of 500+ employees than with early-stage startups. Large enterprises have a "running start" with data—they already have the wikis, help docs, and user patterns that fuel an AI assistant. When Evans saw contract values climb above $100,000 without pushback, he knew he had moved beyond a "nice-to-have" widget into a mission-critical infrastructure component. This highlights a vital lesson for B2B founders: your product's fit is inextricably linked to the scale of the customer's problem. A small startup doesn't need to optimize its UX; it needs to find its first ten users. An enterprise with ten thousand users, however, loses millions in efficiency if those users are confused. Radical Leaness and the Philosophy of the Small Team Despite raising $24 million from top-tier VCs, Evans has kept his team at a lean 35 people. This is a deliberate rejection of the "growth at all costs" mentality that led to the tech layoffs of 2022 and 2023. Evans argues that speed isn't achieved by throwing headcount at problems. Instead, speed comes from high-quality, motivated individuals having more scope than they are comfortable with. Large teams often devolve into bureaucracies where the most talented people spend their time in meetings rather than building. His philosophy is rooted in his time at Bain Capital, where he observed six-person deal teams managing $10 billion buyouts. If a handful of people can manage the complexities of a multi-billion dollar acquisition, a lean team of engineers and product designers can certainly build a category-leading software company. The rule at CommandBar is simple: only hire when something is truly broken or a massive opportunity is being missed. This approach ensures that every hire is critical and every team member feels the weight of their impact. Scout Mindset and the Art of Anti-Selling Culture at CommandBar isn't about perks; it's about intellectual honesty. Evans champions the Scout Mindset, a concept from author Julia Galef. In a "soldier mindset," your goal is to defend your position. In a "scout mindset," your goal is to see the world as accurately as possible, even if it means admitting your previous map was wrong. This is integrated into their hiring process: candidates are challenged with new information to see if they are willing to abandon a wrong answer. To ensure long-term retention, CommandBar uses a technique called "anti-selling" during the offer stage. Evans will sit down with a candidate and tell them everything that is wrong with the company—the chaos, the shifting roles, and the sheer difficulty of the startup journey. If the candidate still opts in, they aren't just joining for a paycheck; they are signing up for the mission. This transparency builds a resilient team that doesn't buckle when the inevitable challenges of scaling arise. Maximizing the Luck Surface Area In the high-stakes world of venture-backed startups, the difference between a unicorn and a failure often comes down to luck. However, Evans views luck not as a lightning strike, but as a surface area that can be expanded. By generating more conversations, asking more open-ended questions, and being willing to change direction based on data, a founder increases the probability of a favorable outcome. The future of software lies in removing the friction between the user's brain and the machine's capabilities. As Large Language Models (LLMs) continue to evolve, the "smart layer" that CommandBar provides will become the standard interface for all digital tools. For Evans, the goal is clear: build the solution that makes every other piece of software better, and do it with a team that is honest enough to face the facts and bold enough to change the world.