Key Takeaways

  • Linear runs zero A/B tests or split experiments, choosing instead to trust team intuition trained on direct customer context.
  • AI is commoditizing code generation, shifting the primary product bottleneck from software output to decision quality.
  • Karri Saarinen defines intuition as neurological training built through continuous immersion, not mystical inspiration.
  • Linear maintains judgment across teams using weekly rituals like Quality Wednesdays, Feature Roasts, and AI customer feedback summaries.
  • Great teams win by compounding context across everyone building the product rather than measuring micro-optimizations in dashboards.

The Software Factory Trap

Most software companies operate like factories. They measure story points, velocity, and code commits. When they want to decide what to build next, they spin up an A/B test and let split-run metrics dictate the roadmap.

Saarinen thinks this approach produces low-learning organizations. When automated tools generate code in seconds, high output stops being an advantage. “It's more that we have this obsession to look at what is the output, the software or the code,” Saarinen said. “I look the other direction: what is actually creating the software? Who is making the decisions? And that's the people doing that.”

If code generation is cheap, the only scarce resource is judgment. Companies that rely on testing dashboards surrender their judgment to local maxima. They ship hundreds of micro-tweaks that please an optimization algorithm while losing all taste for the whole product.

Intuition Is Just Neural Training

Founders often dismiss intuition as guesswork. Growth teams label it unscientific. Saarinen disagrees, arguing that intuition is the direct result of continuous exposure to customer reality.

“We at Linear, we don't run experiments,” Saarinen explained. “We tell people to use their intuition, but I often tell people that intuition is not some magical force like in Star Wars that happens to you. It is something that you have learned while working on something or listening to customers.”

When engineers and designers read raw bug reports, watch user recordings, and talk to buyers, their brains build an internal model of the user. “Intuition is basically your training in your brain that you have,” Saarinen said. “The more you can learn and listen and see this context, it will compound your personal understanding and then eventually compounds the whole team's understanding.”

Instead of outsourcing choices to split tests, Linear invests in mechanisms that pump real customer reality into everyone's head. They use AI-generated customer feedback briefings, hold weekly Feature Roasts to critique work in progress, and dedicate Quality Wednesdays to fixing product rough edges.

Managing Context Over Execution

As engineering execution gets automated, product management changes. Shipping tickets is no longer the job. The real job is building a shared brain across the company.

“In the end, I think more of the product organization becomes about managing this context,” Saarinen noted. “Learning from it, finding ways for the team to come together and learn from it, versus the actual execution of the code or building the thing.”

“Rather than focusing on the execution, we could start focusing on the customers, the explorations, the critiques, the quality, or even the reflection of what we do,” he added. When every builder shares the same rich context, you do not need permission layers, multi-variant test suites, or rigid roadmaps. People make good decisions on their own because their mental model matches the user.

What to Do With This

Cancel your next roadmap prioritization scoring session. Pick three engineers who have not spoken to a customer this month and put them on live support calls or give them five raw user interview recordings to summarize before Friday.