The long, unglamorous road to startup product-market fit

Executive overview

Most startups that look like overnight successes spent years with nothing working. Siqi Chen, co-founder of Runway, built a finance platform after experiencing firsthand how spreadsheet opacity led to mass layoffs — including his own.

Runway's first hypothesis was wrong. They rebuilt around planning and forecasting for slightly larger teams, embedded engineers directly in customer Slack channels, and tripled revenue two months in a row after a viral website launch.

The gap between when you believe in something and when it starts working can be three to five years — and you have to stay irrational about it the whole time.

From engineering to entrepreneurship

• Built a Facebook game on the side in 2007 that became the platform's second-largest app
• First investor offered $5M to build games; Chen took it, reframing his social network as a gaming company
• Core mental model: code runs 10 million times a second — software multiplies individual leverage
• Saw building organisations as "higher-level programming": people, norms, culture replace keystrokes

The spreadsheet problem that started Runway

• At Postmates as VP of Product and Growth, Chen didn't know what "contribution margin" meant — and it was his team's key metric
• At Sandbox VR, built the company's first financial model; it revealed they'd run out of money in three months
• COVID hit while he had just relocated his family — 90% of the company was laid off, including him
• Core insight: business understanding affects everyone, not just executives
• Spreadsheets create a "wall of numbers" disconnected from the context (plans, campaigns, hires) that explain them
• Runway's first product connected business context to financial numbers in one view

The wrong hypothesis, then the right one

• First product let users pivot and slice expenses and revenue in flexible ways
• It worked technically but solved a low-value problem: spend breakdowns don't change month to month
• Customers wouldn't pay meaningful amounts for something with a static answer
• Real pain: deploying millions across hiring, marketing, and forecasting at growing companies
• Pivoted after ~one year to planning and forecasting for slightly larger teams

How Runway runs product development

• Product team makes almost no product decisions — their job is customer proximity
• Designers and engineers sit in every customer Slack channel
• After shipping, an engineer messages the customer directly: "I saw your problem, I built a thing — does this solve it?"
• Traditional PM role is distributed across the whole team; the product team creates the culture, not the roadmap

The viral website launch

• Tweeted "new brand, new website" with no fanfare; it went viral across Twitter
• Leads came in every few seconds; revenue tripled in both June and July
• Key design principle: put the product front and centre, not stock photos of handshakes
• Deliberately avoided referencing any design inspiration — copying kills memorability
• Stripe and Linear are remembered because they did something new first; imitators are forgotten

What it actually takes to survive the early years

• A startup is a company without product-market fit — that's the definition, not a phase to skip
• For Runway, the first two and a half years produced essentially nothing
• Dylan Field (Figma CEO, Runway investor): the most surprising thing about Figma's success was how few people stayed until it worked — even half the early team didn't believe it would
• You cannot know it's working until it starts working
• To keep going, you have to irrationally believe the outcome would be one of the most impactful things anyone could build
• That belief is what lets you inspire a team to continue for another day, week, month