Craft as a Startup Strategy: Linear CEO Karri Saarinen's Playbook

For Karri Saarinen, CEO and co-founder of the software development platform Linear, the old startup mantra of "move fast and break things" has lost its edge. In a mature, design-conscious market, he argues that impenetrable quality is the only reliable differentiator. "You have to match that level of capability or quality, otherwise people just don't pay attention," Saarinen says. That focus on craft solved what he calls a startup's biggest challenge: getting people to notice and care. Here are his ten rules for building with craft.

Leadership and Team Structure

1. Commit to quality at the leadership level. A culture of quality requires buy-in from the top. Leaders must set the tone that craft is the most important priority. This commitment creates a permission structure for the entire company. Saarinen notes that companies often fail when they claim craft matters without a clear plan for implementing it. Reflecting on his time at Coinbase, he saw how design made the crypto space more approachable, which was essential for going mainstream.

2. When building teams, go small and aim high. Larger teams bring more opinions and deliberation, which dilutes execution quality. Small, high-caliber teams produce good work quickly. The key is hiring craft-oriented people who already do great work. It is about setting a standard, not measuring output.

3. Do away with handoff. Over-specialized teams—where a PM decides, a designer visualizes, and an engineer builds—create silos. At Linear, teams are connected, and everyone shares responsibility for quality. "There's no 'handoff to dev.' You're never off the hook," says Saarinen. When everyone understands how designs are implemented, the result is more likely to be high quality.

4. Resist creating specialized product teams. Defined product teams simplify leadership but create artificial silos in quality and culture. One area of the org chart may produce high-quality output while another lags. At Linear, responsibilities are rotated to keep ideas fresh rather than having someone work on search every day. "Users don't care about your org chart—they care about the experience," Saarinen says.

Execution and Design Philosophy

5. Treat the spec as a baseline, not the goal. In a market full of high-quality competitors, excellence requires going beyond the specifications. "A door is a door, and if it opens, it technically functions," Saarinen explains. Meeting the spec is easy; delivering craft is hard. A team aiming for quality views the spec as the starting line, not the finish.

6. Quality is not perfection. Details need to be correct before a public release, but the product doesn't have to be perfect—that's what beta users are for. It's fine to iterate toward polish, but not to show unfinished work to customers. While defects are countable, excellence is harder to quantify. For Saarinen, success is seeing Linear mentioned organically in conversations about quality solutions.

7. The best design is opinionated. A great product is designed for a specific person. Products built for everyone tend to be adequately good at many things but great at nothing. A clear value proposition is needed to win a market's trust. The more specific a product's purpose, the better it performs for its intended users.

8. The simplest way to increase quality is to reduce scope. Teams that struggle with quality are often trying to do too much at once. The solution is to start with less: build one part of a feature exceptionally well, gather feedback, and move forward. Saarinen rejects the idea that quality requires slowing down. "It's about continuously refining a product to meet a standard," he says.

Process and Decision-Making

9. Don't get locked into one way of doing things. Linear doesn't follow a singular process. Instead, the company establishes values and principles so team members understand the why behind their work. Engineers and designers are connected directly with users through shared Slack channels. Responsibility is pushed onto the team, giving individuals freedom to make decisions against a clear standards bar.

10. Data can be a crutch. Linear does not make decisions based on A/B tests or other experiments. Instead, it relies on developing and trusting intuition. Wrong choices become thought experiments: Why did it go wrong? Did team members truly listen to users? This approach keeps responsibility with the team without offering easy outs.

To provide the best experience, you must surprise users—and neither data nor users themselves can tell you how. Ultimately, success depends on hiring people who care about craft and can make informed decisions based on expertise.