An engineer at a Series-A startup opens one browser tab. Inside it: how many people signed up this morning, a replay of the user who rage-clicked the checkout button, the feature flag holding back a half-finished release, and the SQL to ask why any of it happened. A year ago that was six tools, six logins, and six invoices. Now it is PostHog. That collapse - many tools into one - is the whole story.
The tab that ate the analytics stack
PostHog sells software to the people who build software. Its platform bundles product analytics, web analytics, session replay, error tracking, feature flags, experiments, surveys, a customer data platform and a data warehouse into a single product. In 2025 the company raised a $75 million Series E led by Peak XV Partners at a $1.4 billion valuation, which made the eight-letter word "unicorn" technically accurate and slightly beside the point.
The interesting number is not the valuation. It is that more than 190,000 companies now run on PostHog, and roughly 65% of every Y Combinator batch reaches for it by default. For a category that used to be owned by polished enterprise dashboards, that is a strange place to win.
Your data, shipped to strangers
Back in early 2020, co-founders James Hawkins and Tim Glaser were building something else entirely inside Y Combinator's Winter 2020 batch. They kept hitting the same wall: to understand which features people used, they had to ship their users' behavioral data off to a third party. For a privacy-minded engineer, that is an uncomfortable trade - you give away the thing you most want to control in order to learn anything about it.
The other half of the problem was sprawl. Knowing what users did meant one vendor; replaying their sessions meant another; running an experiment meant a third; storing the raw events meant a fourth. Each tool spoke its own dialect and sent its own bill. The stack was less a toolkit than a tax.
Open the box, and let people read it
Hawkins and Glaser made a bet that sounds reckless until it works: build the analytics tool as open source, let companies self-host it so the data never leaves their walls, and run the entire company in public. The code, the roadmap, the compensation framework and the strategy all live where anyone can read them. The company handbook - including how PostHog pays people and how it emails investors - is published on GitHub.
That last part is either radical honesty or excellent marketing. It is, of course, both. Transparency is cheap to claim and expensive to keep, and PostHog kept it. The launch landed at number one on Hacker News in February 2020, which is the developer equivalent of a standing ovation from a notoriously hard-to-please crowd. The seed money followed within weeks.
The team behind the hedgehog
James Hawkins (CEO) and Tim Glaser (CTO) run a company that is 100% remote across more than 20 countries. Nobody tracks hours. People are judged on what ships.
A roughly 110-person company that treats "we'll figure it out in writing" as an operating system.
Nine tools wearing one trench coat
The pitch is simple: stop stitching. Capture an event once, and every PostHog product can use it. The funnel that shows where users drop off links to the session replay of someone dropping off, which links to the feature flag you flip to fix it, which links to the experiment that proves the fix worked. There is also Max, an AI assistant that writes queries and helps debug, because every platform in 2025 is contractually required to have one.
Product Analytics
Funnels, paths, retention and SQL on autocaptured events.
Session Replay
Watch real web and mobile sessions to find the bug.
Feature Flags
Ship to a cohort, toggle remotely, roll back fast.
Experiments
A/B and multivariate tests tied to real metrics.
Surveys
In-app NPS, CSAT and custom questions.
Data Warehouse + CDP
Import sources in a click; query and pipe out events.
The business model fits the philosophy. PostHog Cloud is free for the first million events a month, then charges per product by usage. Over 90% of customers never pay a cent - they sit on the free tier until they grow, and then they grow into the bill. It is patient capitalism disguised as generosity.
Six years, from dorm to unicorn
A milestone reel. Note how the gap between rounds keeps shrinking.
The numbers, and what they argue
PostHog's strongest evidence is not a logo wall - it deliberately doesn't have one. It is reach. The platform is used by roughly 250,000 engineers across 190,000-plus companies, a count that dwarfs the customer rosters of incumbents that have been selling analytics far longer. Adoption spread the way developer tools actually spread: one engineer tries it, likes it, and brings it to the next job.
Companies on platform: PostHog vs. the usual story
Approximate, drawn from public company figures. The point is the order of magnitude, not the decimal.
The backers read the same chart. Stripe led the Series D; Peak XV led the Series E; Accel, GV and Y Combinator were there from the early rounds. That is roughly $194 million raised across seven rounds, which is a lot of conviction for a company that gives most of its software away.
Build every piece of SaaS a product needs
Ask PostHog what it is for and the answer is unfussy: "We're here to help product engineers build successful products." The longer version is more ambitious, and a little cheeky - "We're going to build every piece of SaaS you need to make your product successful." Most companies would call that scope creep. PostHog calls it the plan.
The Series E is funded toward what the founders call "Act 2": pushing past analytics into deeper developer tooling, including AI features that can read your usage data and open the pull request to act on it. The thesis is that the loop from "what happened" to "here is the fix" should close inside one tool. It is the same idea as the founding bet, scaled up - fewer tabs, more ownership.
The case for owning your own loop
Every team that builds software faces the same quiet question: how much of the truth about your own product are you willing to rent from someone else? For years the honest answer was "most of it," because the alternative was building your own stack. PostHog's argument is that you no longer have to choose - you can have the convenience of a hosted tool and keep control of the data, the code, and the bill.
That is the bet worth watching. If software keeps eating the world, the tools that measure software become infrastructure, and infrastructure you can read and own beats infrastructure you merely lease. A hedgehog is a small thing with sharp spines that is surprisingly hard to pick up. The metaphor was not an accident.
Back to that single browser tab. The engineer flips the feature flag, watches the next session replay come in clean, and closes the laptop. Six tools became one; the data never left the building; nobody sent an invoice this month. That is what PostHog changed - not the dashboard, but who gets to own it.
Find PostHog
Figures are drawn from public reporting and may be approximate. Funding, valuation and customer counts reflect data available as of 2026.