


























TOPSHOT - This photograph shows a figurine in front of the logo of the US artificial intelligence safety and research company Anthropic during a photo session in Paris on February 13, 2026. (Photo by Joel Saget / AFP via Getty Images)
AFP via Getty Images
Earlier today, Mutiny CEO Jaleh Rezaei posted that they terminated every customer contract, cut the team to 15 people, and scrapped a SaaS product they launched in November that took seven years to build. The company had $72 million in backing from Sequoia, Tiger Global, Insight Partners, and Y Combinator, eight figures in ARR, and enterprise logos like Uber and Snowflake. She burned it anyway. Four months later, MRR is growing at 188% week-over-week, 12 times faster than Mutiny's original SaaS launch in 2018.
Sequoia reposted Rezaei’s account of the decision on X. The fund’s board member Bogomil Balkansky told Rezaei, according to her account, that Sequoia had invested in the founders, not the product; and that whatever the company was worth at that moment was "nothing relative to what you'll become." That kind of conviction is what distinguishes tier-one VC from the rest of the market: the willingness to absorb a deliberate write-down on an existing business in service of a larger structural bet.
For investors with exposure to B2B SaaS, the Mutiny case forces an uncomfortable question. If a company with proven product-market fit, $72 million raised, and a blue-chip customer base could not execute an AI transition without killing the original business, how many SaaS companies currently threading a similar needle are actually succeeding? Sequoia’s "Services as Software" thesis explicitly anticipates that AI agents will replace workflow software. Mutiny is one of the first companies to act on that thesis rather quickly.
Mutiny’s initial response to the AI transition was rational on paper; keep the SaaS revenue as a cash cow, build the AI product alongside it and sell the new product to existing customers. Rezaei describes trying every variant: splitting the team, removing executives to go founder-mode, reducing headcount by 30%, and stopping new SaaS sales. None of it moved fast enough.
Three structural problems kept surfacing. First, the operating modes were incompatible. A scaled SaaS business runs on autonomy and process; pre-product-market-fit AI development requires centralized, fast decision-making by founders. Second, SaaS and agents have fundamentally different technical architectures. Mutiny had to rethink its data models and APIs from scratch to optimize LLM performance, a rebuild that couldn't coexist with maintaining enterprise-grade SaaS infrastructure. Third, go-to-market lived in the old world while product lived in the new one. Enterprise customers like Doordash and Notion kept pulling engineering attention into support fires.
Rezaei's summary: she and co-founder Nikhil Mathew put in 150% every day and got 30% out.
The rebuilt product is an AI agent for enterprise go-to-market teams. Salespeople spend roughly 70% of their time on custom preparation work; building case studies, ROI reports, ABM campaigns, deal rooms - rather than selling. The agent ingests a company's brand guidelines, case studies, and resources, then generates personalized, on-brand assets in minutes. The business model flipped from sales-led to product-led growth.
The results after launch were immediate. Over 1,000 signups from Rippling, Snowflake, DHL, Amazon, and Google arrived with zero marketing spend. Mutiny hit every general availability metric by the end of its launch week.
The Mutiny playbook is not obviously replicable. Most SaaS companies cannot self-fund a clean break from their revenue base, and most boards will not sign off on terminating every customer contract. What Rezaei had that most founders do not was a Sequoia board member willing to absorb the short-term loss, a team small enough to move in person, and a customer acquisition signal, 1,000 enterprise signups at launch, that validated the bet within weeks.
But the strategic lesson is transferable; the companies that will lose the AI transition are not the ones that failed to invest in AI features. They are the ones that treated AI as a layer on top of an existing SaaS architecture rather than a reason to rethink the architecture entirely. Mutiny's 12x MRR growth rate is the data point that makes that argument hard to dismiss.
For investors, the implication is obvious: the SaaS companies with the most defensible AI positions may be the ones willing to cannibalize themselves first. Founders who treat their existing ARR as a constraint rather than a foundation are the ones worth backing.
此内容由惯性聚合(RSS阅读器)自动聚合整理,仅供阅读参考。 原文来自 — 版权归原作者所有。