Hot off of Segment’s $3.2B sale to Twilio in 2020, I set an ambitious goal for our DC policy work @CharmIndustrial: win access to a $80/ton CO₂ removal tax credit within the year. After failing in ‘21 and 😂 about it, we learned. Now we have exciting progress to share today! /1 Silicon Valley runs best on fast, aggressive, unilateral decision-making fueled by trial and error. That’s what I lived and breathed with Segment and now Charm for over a decade. I approached DC the same way: 50+ Charm tours & meetings with lawmakers, staff & influencers. /2

As @Segment grew from 0 to >$200M ARR, I had three health scares. Each scare made me realize that good physical health is *required* to push hard and win as a founder, and you have to own your own healthcare. The medical system simply won’t do it for you. /1 First, the b.s. about stress… mental health is a hot topic right now, but not everything is mental. /2

One of my favorite learnings from Twilio CEO @jeffiel was that the most explosive developer products are the ones that solve a *biz dev* problem for developers. A 🧵1/5 The absolute last thing a developer wants is to sit in meetings with telecoms just to send a text (solved by @twilio), or spend months with credit card co’s and banks just to accept payments (solved by @stripe @Plaid). Just gimme an API key immediately after I sign up. 2/5

When @Segment was over $50m ARR we realized we were spending a fortune on servers, and it was growing like topsy, faster than revenue. Gross margin was free-falling into the low 50%'s (software benchmark was 80%). It didn't feel like we were in control, which I loathed. A 🧵 1/11 Around this time, Sandy Smith joined as our first CFO. She had spent years working on improving gross margins at web infrastructure companies like @Akamai and @Twilio. In those hardcore infrastructure co's, gross margins were a game of a 1-2%-point improvements each year. 2/11

After raising >$300m in capital across 14 funding rounds for @Segment and @CharmIndustrial, my fundraising process has gotten a LOT better. A thread 1/17 During @ycombinator we got programmed to focus exclusively on customers and code until demo day. No investors before then. 2/17

After we found product market fit @Segment in Dec 2012, we nearly ran out of money because we were so scared to ask customers to pay. It took two gracious customers, a cocky sales advisor and a lot of fernet to learn the true value of our product. A 🧵 with 📸 1/18 The underlying challenge was two-fold: (1) analytics.js was an open source library and we came from the open source mindset, so our anchor was $0, (2) we were all fresh out of college, had a few hundred bucks to our names, and had never sold anything. 2/18

In 2019 my co-founder @logiclow tried to convince me not to invest in @CharmIndustrial because it seemed hopeless. Instead, he had two breakthroughs and 2y later, Charm delivered 90% of global permanent carbon removals. What happened?! a🧵1/20 The beginning: in 2015 we decided to measure and offset our emissions @Segment. We purchased some Indonesian rainforest protection offsets, felt good, and moved on. But… 2/20

Every startup founder is “crushing it”, but in 2011 I sure wasn’t. After @ycombinator demo day in August 2011 I got jealous of a friend’s fast traction, forgot to eat for three days, and nearly let it destroy @Segment when we were still just a classroom lecture tool. A🧵 1/20 During the summer 2011 batch of @ycombinator we rushed to get our classroom lecture tool built and into the hands of professors. We never considered pivoting, but our traction with professors was not impressive. 2/20

In the early days of Segment, right after we crossed $10m ARR our growth decelerated 8x from 300% yoy to 40% yoy in 6 months. What the hell happened? How did we reaccelerate back to 100% yoy? A thread 1/18 At the surface, everything was burning. Customers were churning off our highly profitable Redshift data loader for cheap alternatives. A new competitor was killing us on mobile-first deals. The sales team was getting incredibly demoralized. 2/18

As Segment grew from open source afterthought to the leading customer data platform, we "refound" product market fit several times. Surprising to me as a product person, our biggest learnings in PMF were from wildly uncomfortable questions from our sales team 👀 a thread 1/14 Our 2nd sales hire was @entrp4life, who we hired despite my protests to @RaphaelParker about a likely culture clash with engineering buyers. I couldn't have been more wrong. Not only did Surdi blow his number out of the water for years, but his customers loved him. 2/14

The toughest part of growing Segment from $0-200M+ ARR was letting go of product (which I understood) to focus on go-to-market (which I didn't). In 2017 I was really struggling with this. Five friends set me straight, a thread. 1/12 .@calvinfo @ivolo and I dropped out of MIT undergrad straight into Segment... and 6 years in we were now responsible for a growth stage startup with several hundred ppl. 2/12

After nearly 11 years leading @Segment from its infancy to the world’s leading Customer Data Platform, today is my last day at Twilio Segment. Starting Monday I will be working full-time on @CharmIndustrial, focused on permanent carbon removal via bio-oil sequestration. /1 Segment’s first two failed ideas were mine, so I’m incredibly grateful to have had my best friends @ivolo, @calvinfo and @ianstormtaylor as co-founders. Their idea to make open source analytics.js into a product sounded insane to me, but they couldn’t have been more right. /2

1/ Today is 8 years since we founded @Segment! Like many startups, our first two years were brutal. The last six have been an insane ride. We made so many mistakes along the way, and owe so much to our early customers. 2/ While attending MIT and Y Combinator in Summer 2011 we built a classroom lecture tool called ClassMetric. YC constantly told us to solve real problems for our customers, but we somehow ignored this and 80% of the students just opened up Facebook instead of using our tool.