Peter Reinhardt on Segment’s Pivots and Charm Industrial’s Carbon Removal

概览

This episode traces Peter Reinhardt’s path from a failed classroom feedback startup at YC to Segment, and then from Segment’s acquisition by Twilio to Charm Industrial. The core thread is how a team repeatedly found real demand only after confronting that its original assumptions were wrong.

A major lesson from the Segment story is that traction came from a small, neglected open-source analytics library, not from the large analytics product the team had spent a year building. Peter also describes how learning sales, pricing, customer discovery, and transparent leadership changed the company’s trajectory.

The second half focuses on Charm Industrial, Peter’s carbon removal company. He explains Charm’s model of converting biomass into bio-oil and injecting it underground, why demand for carbon removal emerged faster than expected, and why permitting and regulation have been the company’s biggest bottleneck.

分段落总结

[00:00] Peter Reinhardt’s Background

[事实] Peter Reinhardt is introduced as the founder of Segment, a YC Summer 2011 company, and the founder of Charm Industrial, a carbon removal company. [事实] The conversation begins with Segment’s origin story, which started as a different company called ClassMetric. [推测] The episode frames Peter’s career as a series of pivots from software to hard tech rather than a straight-line founder story.

[00:50] ClassMetric and the Classroom Feedback Idea

[事实] ClassMetric gave students a button to anonymously indicate confusion during lectures, while professors saw a graph of confusion over time. [事实] Peter says the idea seemed right in theory, but people may not have actually wanted it. [事实] The founders were MIT and Rhode Island School of Design students who were influenced by Hacker News and Paul Graham’s essays.

[02:19] YC Interview and Weak Founder-Idea Fit

[事实] The YC interviewers liked the founder group but did not love the ClassMetric idea. [事实] Paul Graham suggested another idea during the interview, which indicated concern about the proposed startup. [事实] Peter recalls Robert Morris telling Paul Graham during the interview that he would not use the product, despite having earlier sounded more positive in a customer discovery meeting. [推测] YC accepted the team more for founder quality and pivot potential than for the original classroom product.

[04:05] YC Summer and Failed Classroom Launch

[事实] The team spent the YC summer building the classroom application and tested it in summer classes at Stanford. [事实] After Demo Day, they raised about $600,000 and moved back to Boston to launch near universities. [事实] They got several classrooms to adopt the product, but students mostly used laptops for Gmail, Twitter, Flickr, eBay, and other distractions during lectures. [推测] Observing real classrooms reversed Peter’s view of technology in lectures and convinced the team the product was harmful or poorly matched to the environment.

[07:30] The Analytics Tool Pivot

[事实] After shutting down the classroom idea, the team told investors they would pivot to building an analytics tool competing with products like Google Analytics. [事实] Two investors took their money back after the pivot. [事实] The team spent about a year building the analytics product and did not test B2B sales or go-to-market well. [事实] Peter says customer interest turned out to be idle curiosity rather than purchase intent.

[09:36] Low Morale Before Segment Emerged

[事实] By late 2012, Peter had just returned from his honeymoon and the team recognized the analytics product was probably not working. [事实] Peter says his morale was extremely low and he was having panic attacks about once a week. [事实] He was 23 and had dropped out rather than finishing his degree. [推测] The emotional pressure came from having raised money, spent a year building, and still not having meaningful users or sales.

[10:38] The Analytics.js Manifesto

[事实] Co-founder Ian wrote an “analytics.js manifesto” arguing that their small open-source routing library was the best thing they had built. [事实] The library had been created earlier to send ClassMetric data to Mixpanel, Google Analytics, and KISSmetrics. [事实] Peter initially resisted the idea because the library was open source and only about 200 lines of code. [事实] The team decided to build a landing page and post it to Hacker News rather than disappear into another long build cycle.

[12:30] Hacker News Traction

[事实] Peter expected the Hacker News launch to fail quickly so the team could move on. [事实] Instead, analytics.js reached number one on Hacker News, got a few hundred upvotes, thousands of GitHub stars, and thousands of email signups. [事实] That reaction made the decision to pursue the idea much easier. [推测] The launch gave the team stronger evidence of demand than either ClassMetric or the analytics product had produced.

[13:15] Paul Graham’s Harsh Office Hours

[事实] Around that time, Peter remembers office hours with Paul Graham, who encouraged them to pursue analytics.js. [事实] The hosts recall Paul telling the team they had spent about half a million dollars and had nothing to show for it. [事实] Peter says the comment hurt but felt real. [推测] The direct feedback helped force the team to stop defending the previous year’s work and focus on the thing showing actual pull.

[14:23] From Open Source Library to Hosted Product

[事实] The team found that users did not just want the open-source library; they wanted a hosted version where they could choose destinations by pushing buttons. [事实] That hosted routing service became the product, service, and company behind Segment. [事实] Growth was very linear for a period, which Paul Graham noticed during later office hours.

[16:31] Fundraising and Investor Updates

[事实] When Segment was running low on cash, Peter emailed investors to say they finally had something working. [事实] He had not been sending investor updates, which he calls a huge mistake. [事实] The hosts note that founders going quiet is usually a bad sign to investors. [推测] The lack of updates made it harder for investors to understand the company’s progress after the long difficult period.

[17:08] What Segment Solved

[事实] Segment helped software companies and mobile apps collect customer behavior data and send it to analytics, email, advertising, and data warehouse tools. [事实] Without Segment, teams often had to ask engineers to integrate each new tool individually. [事实] Segment let marketers or other users push a button to send data to services like Salesforce, Adobe, Google, or other destinations. [推测] Segment solved an unglamorous infrastructure problem that was highly valuable inside companies even if consumers never saw it.

[19:07] Pricing Lessons

[事实] Segment initially charged $10 per month. [事实] A customer warned that $10 per month was too low to sustain the service and could make him migrate away. [事实] A sales consultant told Peter they should be charging $120,000 per year rather than $120 per year. [事实] Their first major sales negotiation with Nat Friedman’s Xamarin ended at $18,000 per year, which still dramatically changed Peter’s sense of value-based pricing.

[22:32] Learning Sales from Chris Surdean

[事实] Segment’s first sales hire, Rafael Parker, pushed Peter to hire Chris Surdean as the second sales rep. [事实] Peter initially doubted Chris could sell to developers because of his style and presentation. [事实] Chris became one of Segment’s top sales reps for years. [事实] Peter learned from watching Chris that strong sales often means pushing through awkward discovery to understand the customer’s real pain before pitching.

[27:10] Twilio Acquisition Logic

[事实] Twilio’s Jeff Lawson had reached out for years saying Twilio and Segment would be better together. [事实] His argument was that Twilio had the communication layer and Segment had the data layer. [事实] Peter initially did not buy the argument, but around 2019 and early 2020 he saw customers asking for those capabilities to work together. [推测] Customer behavior, more than strategic theory alone, made the acquisition logic credible to Peter.

[28:43] Data Warehouses and Another Growth Curve

[事实] Around 2014, Segment launched a product that sent data to cloud data warehouses such as Redshift. [事实] Peter realized the opportunity after several New York customer meetings showed the same desire to store raw data in a warehouse. [事实] Segment’s revenue grew from about $2.5 million to $10 million the next year. [事实] Peter credits Chris Surdean not only with teaching him sales discovery but also with changing how he thought about product discovery.

[29:48] Pricing Crisis and Leadership Lesson

[事实] After rapid growth, Segment hit a pricing and packaging problem that caused growth to slow and even shrink in one month. [事实] Peter was stressed and kept the problem too much inside his own head. [事实] A part-time CFO and HR leader told him that employees did not know the company’s priorities and pushed him to communicate the problem clearly. [事实] Peter held an all-hands meeting explaining the situation and what needed to be done.

[31:23] The Team Rescues the Year

[事实] Peter asked a small product team to urgently ship a pricing and packaging update in a few days. [事实] He felt guilty for years about putting that pressure on them. [事实] After the acquisition, people on that team told him the moment was their favorite at the company because they got to help rescue it. [事实] Segment hit $20 million ARR at about 11:50 p.m. on December 31.

[35:18] Carbon Offsets Spark Charm

[事实] Peter started buying carbon offsets in 2015. [事实] In 2016, he tried to evaluate whether the offsets had actually helped and found the process opaque and unsatisfying. [事实] He later bought things like refrigerant destruction because they had a more tangible supply chain, though he says they were not scalable enough. [推测] Charm grew partly from Peter’s frustration as a customer who wanted transparent, durable climate impact.

[36:15] Charm’s Original Industrial Idea

[事实] In 2017, Peter spent Saturdays exploring ways to turn existing industrial processes from carbon emitters into carbon sinks. [事实] The idea evolved into Charm Industrial, originally related to “char farm industrial.” [事实] The concept involved taking biomass from farms, leaving behind char, and using bio-oil as an industrial feedstock to replace natural gas, coal, or similar inputs. [事实] Charm incorporated in February 2018 and brought on engineers as full-time co-founders.

[37:42] Peter’s Hard Tech Motivation

[事实] Peter says he was always more of a hard tech and hardware person than a software person. [事实] He fell into software largely because his MIT roommates were software engineers and wanted to start a company. [事实] Climate was relatively new to him, but he understood the problem from the customer side. [推测] Charm aligned more directly with Peter’s original technical interests than Segment did.

[40:44] Leaving Twilio for Charm

[事实] Peter told Segment board members that Charm helped him maintain the stamina to keep running an enterprise software company. [事实] He worked on Charm roughly half a day to a day per week while still running Segment and later working inside Twilio. [事实] In fall 2021, COP26 meetings for Charm conflicted directly with Twilio leadership responsibilities. [事实] Peter decided he could no longer execute both roles and left Twilio.

[42:06] What Charm Does

[事实] Charm puts carbon back underground in the form of liquid biomass oil. [事实] Customers such as Stripe, Microsoft, or JP Morgan can pay Charm to remove tons of carbon they cannot eliminate through emissions reductions. [事实] Charm currently procures wildfire residues in Colorado, converts the biomass into bio-oil, ships it to Louisiana, and injects it nearly a mile underground. [事实] Peter describes the future use of bio-oil as a feedstock for hard-to-decarbonize sectors such as steel, cement, aviation fuel, and shipping fuel.

[44:24] Carbon Removal Demand

[事实] Charm originally expected bio-oil to be used mainly in industrial processes such as steelmaking. [事实] The company discovered that many buyers wanted carbon removal itself. [事实] Demand started with Stripe and spread to Shopify, Microsoft, banks, and other tech companies. [事实] Peter says Charm is under contract for roughly 200,000 to 300,000 tons of material to put underground.

[45:10] Underground Capacity and Orphan Wells

[事实] Peter says the underground capacity for injection is shockingly large. [事实] One well may have around half a million tons of capacity. [事实] Charm uses orphaned oil and gas wells, which are abandoned wells that states inherit as liabilities when operators go bankrupt. [事实] Peter says Louisiana has about 22,000 relevant orphaned wells and the United States has between two and five million orphaned oil wells overall.

[47:25] Carbon Removal Scale

[事实] Peter says cheap electricity or fusion does not directly help Charm much because Charm gets much of its energy and carbon from biomass. [事实] He says decarbonizing the grid and vehicles still leaves an overshoot of carbon dioxide in the atmosphere. [事实] He cites an IPCC-style need for about 10 billion tons per year of removal by 2050. [事实] He compares that need to reversing the scale of the entire oil and gas industry.

[50:03] Regulation as the Biggest Barrier

[事实] Peter says the biggest barrier over the last five years has been regulatory. [事实] It took about four years for federal and state agencies to agree which permitting pathway Charm’s bio-oil injection should follow. [事实] The first Louisiana permit took about 14 months after that. [事实] Peter argues that the delay had large company costs and public health costs because biomass that could have been processed was instead burned.

[54:06] Policy and Recent Progress

[事实] Peter says Charm has a few talented policy people and wonders whether the company was under-invested in policy. [事实] He says the regulatory side has made significant progress this year. [事实] He says the current EPA administrators are eager to accelerate innovative companies doing interesting things. [推测] The discussion suggests Peter sees policy work as central to scaling climate technology, not just as a compliance function.

[55:37] Why Charm Energizes Peter

[事实] Peter says he loves working on Charm. [事实] Most of Charm’s engineering and operations are in Colorado, and he travels there weekly. [事实] He enjoys walking through the plant, watching engineers make progress, and seeing bio-oil being produced. [事实] He contrasts the physical manifestation of Charm’s work with Segment’s software work.

[57:20] Charm’s Bet-the-Company Hardware Moment

[事实] Peter describes a period when Charm was about six weeks from running out of cash. [事实] He personally put in additional money to extend the runway by a couple of months. [事实] A new pyrolyzer ran far longer than previous attempts, reaching about 36 hours. [事实] Around the same period, Charm got pyrolysis working, landed a major Frontier contract, and later raised its Series B.

[59:13] Transparency for Carbon Removal Buyers

[事实] The hosts say they use Charm to offset flights. [事实] Peter says customers can see delivery information through Charm’s ledger. [事实] He says he wanted extreme transparency and a FedEx-style delivery history as a customer. [事实] At the time of the conversation, customers still needed to do their own flight calculation or choose a subscription model.

[60:37] Charm’s Current Inflection Point

[事实] Peter says the August permit was a major transformation point for Charm. [事实] After the permit, the problem shifted from waiting for regulatory approval to executing and reducing costs. [事实] He lists shorter transportation distances, larger throughput machines, more machines, and similar engineering work as next priorities. [事实] Charm has customers including Stripe, Figma, Shopify, and other tech companies.

[62:21] Frontier and Advanced Market Commitments

[事实] Stripe organized Frontier, an advanced market commitment that pooled about $1 billion of purchasing power for carbon removal. [事实] Charm was the first company Frontier signed a purchase agreement with. [事实] Peter describes the Charm purchase as roughly an order-of-magnitude $50 million purchase. [推测] Frontier helped create early demand certainty for carbon removal companies that otherwise would face difficult scale-up economics.

[63:00] Peter’s Policy Ask

[事实] Peter says one of the most important takeaways for voters is the importance of deregulation. [事实] He says many climate tech CEOs face regulatory barriers more than they receive government support. [事实] He argues that society has made building real-world things shockingly hard. [事实] He connects the same abundance argument to California housing policy and reducing barriers to construction.

[65:24] Hosts’ Closing Reflections

[事实] After Peter leaves, the hosts say they enjoyed catching up with him and admire his transition from a side project into his main work. [事实] They emphasize that Peter spent a decade on B2B software even though Charm appears closer to his passion. [事实] They describe the Segment pivot story as encouraging for founders still searching for product-market fit. [推测] The hosts view Peter’s story as a combination of persistence, luck, and willingness to follow unexpected evidence.

播客点评/总结

This episode is valuable because it does not flatten the Segment story into a simple success narrative. Peter describes the failed classroom idea, the painful analytics detour, weak pricing instincts, sales discomfort, and leadership mistakes before getting to the acquisition outcome.

The strongest parts are the concrete founder lessons: test demand quickly, distinguish polite interest from buying intent, price based on customer value, let sales discovery shape product thinking, and tell the team the real problem when the company is in trouble.

The Charm section is useful for listeners interested in climate tech because it explains carbon removal as an operational, regulatory, and engineering problem rather than only a scientific idea. Peter’s emphasis on permitting bottlenecks and transparent delivery gives the discussion a practical edge.

[推测] The episode is best suited for founders, startup operators, climate tech investors, and listeners interested in how entrepreneurial skill transfers from software to hard tech. Its main limitation is that the claims about regulation, policy, and climate impact are presented mainly from Peter’s perspective rather than debated with outside evidence.