The Idea:

In August 2019, James Hawkins and Tim Glaser quit their jobs and put their last $8K of salary into a business bank account. James, a Cambridge economics grad and former VP of Sales at an enterprise software company, booked two meetings a day with heads of sales. Tim, who'd run R&D, built whatever James was pitching. Then they watched it die.

A sales territory tool: 15 sales leaders agreed to try it, and one clicked the sign-up link. A predictive CRM: one customer at $20 a month, who never paid the invoice. A 1:1 meeting tool nobody opened. A technical debt tracker that got them into Y Combinator's W20 batch with 600 users, then stalled when nobody would pay. An engineering retention tool that lasted five days.

Five pivots in six months. But one frustration had followed them through every failure. Each time they wanted to understand how people used their products, they had to ship all their user data off to a third-party analytics tool. So they built the thing they wished existed: product analytics developers could host themselves, open source from day one.

They wrote the first line of code on January 23, 2020.

This was the beginning of PostHog.

The Execution:

• February 2020: Just four weeks after the first commit, PostHog launched on Hacker News. It racked up 300+ deployments in a couple of days and passed 1,500 GitHub stars within a fortnight.
• 2020: Fresh out of YC, the founders raised a $3.025M seed from YC's Continuity Fund and 1984 Ventures in April. By December, GV had led a $9M Series A, taking total funding to $12M with staff in ten countries.
• June 2021: A $15M Series B landed ahead of schedule, led by YC. Three months later, PostHog hit product-market fit for its paid tier with a team of 30.
• 2022: While the tech market crashed, PostHog grew revenue 6x into the millions, with CAC payback under two months and just 38 staff.
• March 2023: The gut punch. PostHog lost its then-largest customer, who moved their source of truth into a data warehouse. Rather than retreat, the founders decided to build their own warehouse. They finished the year having quadrupled revenue while adding three net hires.
• July 2024: With 54% of YC's first 2024 batch already using it, PostHog cut prices by up to 80% on its two most popular products, existing customers included. Growth sped up.
• December 2024: PostHog passed 100,000 customers, reached low tens of millions in ARR and ended the year cashflow positive with around 80 people.
• June 2025: A tweet from Patrick Collison praising PostHog's website turned, 18 months later, into a $70M Series D led by Stripe at a ~$920M valuation.
• September 2025: Peak XV led a $75M Series E at $1.4B, making PostHog a unicorn. James called it the start of "Act 2".
• Today: Around 200 people, a product list spanning analytics, session replay, feature flags, error tracking, logs and AI observability, and PostHog Desktop billed as its biggest launch yet. The new pitch: self-driving products.

The lesson? PostHog's hardest pivot was the one with the most traction: a technical debt tool with 600 users and a YC badge. Most founders would have tweaked it, scraped together a small seed and spent years propping it up. James and Tim walked away because they didn't believe in it, and the winning idea came from a frustration they'd felt across all five failures. Traction you don't believe in is just a slower way to fail.