The archive · Developer & Business Tools · Strategic decision · 2011–2013
Segment.io's 2013 pivot bet: one analytics API instead of another analytics product
YC S11's founders pivoted from a failed analytics product to one unified API; by January 2013 it had thousands of signups and 300+ active projects.
Segment.io
What the business is
Segment.io sold a unified analytics API: developers install one client-side JavaScript or server-side library (Ruby, Node.js, Python, Java, .NET) and Segment routes the data to 20+ providers such as Google Analytics, KISSmetrics, Mixpanel, Chartbeat, HubSpot and Salesforce.
Starting capital：$600,000 in seed funding from NEA, General Catalyst and other angel investors.
How it started
Four founders — three of them MIT roommates, plus a Rhode Island School of Design student — dropped out to join Y Combinator's summer 2011 batch with a plan to build a competitor to Google Analytics or KISSmetrics. They struggled to get people to integrate with their service, but they had also built Analytics.js, a library wrapping multiple analytics services and APIs, and released it as open source on GitHub. "The open source version started growing by itself," said Peter Reinhardt, "so after a couple of months, we decided to pivot and build what people seemed to actually want: a beautiful, simple analytics API." The team made that shift in December 2012.
What happened
By the January 2013 launch, Segment.io supported 20 analytics providers on both client and server side, with a mobile solution planned. The hosted, client-side Analytics.js stayed free, while developers paid for server-side libraries: $30/month for 1 million server-side calls, or $150/month for 10 million plus premium integrations including HubSpot, Marketo, Omniture and Salesforce. The pitch was time: integrations that had taken some customers months with old SOAP-based enterprise APIs could now take hours, and adding or removing providers was a switch rather than a rewrite. Reinhardt called it an investment in being data-driven — removing the barrier to using analytics tools that teams avoided because setup was painful — and the company planned two to three new provider integrations per week.
No ending yet — it is still running.
Background
Segment.io started as a plan to compete with Google Analytics or KISSmetrics: four founders — three MIT roommates and a RISD student — dropped out to join Y Combinator's summer 2011 batch. The problem was adoption; nobody wanted to integrate with another analytics service. But the team had also built Analytics.js, an open-source wrapper around many analytics services, and released it on GitHub, where it grew without their help.
In December 2012 the founders pivoted: instead of selling analytics, they would sell the routing layer. Segment.io's simple API sends one event to 20+ providers — Google Analytics, KISSmetrics, Mixpanel, Chartbeat, HubSpot, Salesforce and more — on the client or the server, so developers stop hand-coding each provider's SDK and businesses can switch tools without rewriting code. Integrations that took customers months with old SOAP-based enterprise APIs could now take hours, and the company planned two to three new providers per week.
The hosted client-side Analytics.js stayed free while server-side libraries cost $30/month for 1 million calls or $150/month for 10 million plus premium integrations. As of the 2013-01-25 TechCrunch profile, roughly 1,800 developers had started with the open-source version, and the paid service had thousands of signups and over 300 active projects, backed by $600,000 in seed funding from NEA, General Catalyst and other angels.
What has to be true
- The original product had a distribution problem, not a skill problem: the team could build the integration layer, but nobody wanted yet another analytics destination.
- Open source was the demand signal: Analytics.js grew by itself to about 1,800 developers, proving the wrapper was wanted even when the hosted product was not.
- The pain was quantifiable: every analytics API differed, and old enterprise integrations like Salesforce or Marketo had taken customers months, so a unified API saved real developer time.
What can be applied
Your rejected tool can be the wedge: nobody wanted the team's analytics product, but developers wanted the wrapper — the pivot followed demand instead of the original plan.
Aftermath
As of 2013-01-25 Segment.io was live in its new form: thousands of signups, over 300 active projects, and about 1,800 developers on the free open-source Analytics.js, with $600,000 in seed funding from NEA, General Catalyst and other angels. The company was adding two to three analytics providers per week with Pardot next; the record stops at this launch, with later growth and milestones not covered by the source.
Sources
- YC-Backed Segment.io Lets Developers Integrate With Multiple Analytics Providers In Hours, Not Weeks
spotted an error? The archive wants to know.
Your turn
You just read one. Describe what you are building, and see who is betting on the same thing.
Free account · 3 free questions · no card