From Google to $1M ARR Every 10 Days: Zach Lloyd's Warp Story

Zach Lloyd took everything Google teaches you — massive scale, unlimited engineers, instant distribution — and learned it was the perfect preparation for failure. His comeback? A terminal that became an agentic development environment now growing at $1M ARR every 10 days.

Key Facts

• Warp ARR growth rate: $1M ARR added every 10 days (as of recording) (Zach Lloyd)
• Total funding raised: ~$70M across Seed, Series A, and Series B (Zach Lloyd)
• Series B size: $50M, raised pre-revenue (Zach Lloyd)
• Day-one waitlist signups: 10,000 developers signed up the first day Warp was posted on Hacker News (Zach Lloyd)
• Time to first $1M ARR: 300 days before the agentic development pivot accelerated growth (Zach Lloyd)

Why Google Is the Worst Preparation for Startup Founders

Google optimizes for scale, not distribution. Engineers there never worry about whether anyone will use what they build — millions of users show up automatically. That assumption is lethal at a startup, where getting people to care is the hardest problem.

Zach Lloyd spent eight years at Google, ultimately leading engineering for the entire Google Docs suite. By any measure, it was an elite career. But when he left in 2015 to build a photo sharing app, he discovered that Google had trained him to be blind to the most important startup skill: distribution.

The photo app, Camera Club, worked well technically. It attracted a few thousand users. But it never became a real business — and Zach admits he never seriously thought about why anyone would use it. He'd spent nearly a decade in an environment where launches automatically reached millions of users, where the best engineers in the world applied to work with you, and where the business model was someone else's problem.

That first experience set up every lesson that followed. The failure wasn't a disaster — it was data.

"Google is an amazing company that sets you up in a uniquely bad way to be a startup founder. Anything you launch at Google is going to get millions of people trying it. That is not the case for a startup. The challenge in a lot of ways is not can you build a cool product — it's can you get people to care." — Zach Lloyd

The SelfMade Post-Mortem That Built Warp's Operating System

After leaving SelfMade — a social media editing startup he co-founded as CTO — Zach spent months writing a brutally honest post-mortem documenting every failure. The principles he extracted became Warp's 'How We Work' document, which governs hiring, communication, and product process to this day.

SelfMade started as a consumer photo-editing marketplace and pivoted into a tech-enabled social media service for small businesses. It reached mid-single-digit millions in revenue and made it through a Series A. But the model was operationally heavy, the unit economics were difficult, and Zach found himself increasingly disconnected from the mission.

The breaking point was cognitive dissonance. He could no longer in good conscience sell employees and investors on a business he didn't believe in. Leaving was painful — described as 'a bit like a divorce' — but necessary. What he did next was unusual: he wrote everything down.

The post-mortem wasn't published — it's too personal, too critical, documents too many real interactions with investors and co-founders. But from it, Zach extracted a set of principles he still runs Warp by: only work on problems you genuinely care about, build pure software not services, hire only product-obsessed people, and document everything in advance so the team never wastes time debating operating norms.

"It got to a point where I no longer believe in the mission or the viability of the business and my co-founders still did. I ended up feeling too much cognitive dissonance to sell a thing to employees and future investors that I fully didn't believe in." — Zach Lloyd

"We have this thing called How We Work — a public Notion document. It's all the principles of the company written down. We spent so much time debating this stuff and wasting time at SelfMade. Much better to just have it written down and let everyone look at it." — Zach Lloyd

• Only work on a problem you deeply care about — startups are too hard otherwise.
• Build software, not services — operational complexity destroys margins and speed.
• Hire product-thinkers in every role, including engineering.
• Write down every operating principle before conflict forces the conversation.

How Warp Got 10,000 Signups on Day One (Without Friends Caring)

Warp validated demand by posting on Hacker News with a simple waitlist — not by testing with friends. The product framing was clear: a terminal built in Rust with a consumer-grade UX. Ten thousand developers signed up the first day, proving real market pull existed.

The original insight behind Warp was straightforward: developer tools had never received the same product polish that consumer software had. Slack replaced WebEx. Notion replaced clunky wikis. Zoom replaced WebEx. Why hadn't anyone done the same for the terminal — the black screen with green text that every developer uses daily but almost no one has tried to redesign?

Zach's answer was Warp. Built in Rust, designed with consumer-grade UX, and focused on removing the 'magic spell' problem — the fact that the terminal is powerful but only if you already know the exact commands. The initial friends-and-family test flopped. Nobody close to the team wanted to switch tools. But a Hacker News post with a waitlist link generated 10,000 sign-ups in a single day.

The lesson Zach draws is about signal quality. Friends can give falsely positive feedback out of loyalty, or falsely negative feedback because they're not the target user. Putting something in front of the actual audience — developers who hang out on Hacker News — gave real signal fast.

"We posted on Hacker News that we were building this new terminal. It was built in Rust, it had all these features. We posted it with a waitlist — just sign up, give us your email if you're interested — and we got 10,000 developers to sign up for it the first day." — Zach Lloyd

Raising $50M Pre-Revenue: What Sequoia and Figma's CEO Were Betting On

Warp raised a $50M Series B from Sequoia with no meaningful revenue. The bet was on developer love: hundreds of thousands of free users who retained and engaged daily. Investors like Dylan Field of Figma believed that deep developer adoption would create a large business opportunity, regardless of the exact model.

Warp's fundraising arc is striking. A $6M seed from GV in 2020. A $15–17M Series A led by Dylan Field, founder and CEO of Figma, who invested as an angel. Then a $50M Series B from Sequoia — all before Warp had cracked a scalable revenue model.

The logic was consistent across all three rounds: developers are an unusually influential and hard-to-reach market. A product with strong retention among hundreds of thousands of developers is rare and valuable. The exact monetization path could be figured out later. That thesis required investors with long time horizons and genuine conviction in the product — not investors optimizing for short-term metrics.

Zach is candid about the fundraising environment: he acknowledges it was 'different times' for some of the pre-revenue rounds. But he also attributes the success to knowing his investors personally for years and to the trust that comes with being a second-time founder with a credible first chapter.

"Dylan really believed in the product and the team. Our investors for Warp were very much of the mindset of: if you build something that's really useful for hundreds of thousands or millions of developers, that's going to be a big business opportunity one way or the other." — Zach Lloyd

The Agentic Development Pivot That Turned Warp Into a Growth Machine

In early 2024, Warp launched 'Agent Mode' — letting developers instruct the terminal in plain English rather than command syntax. By June 2024, Warp repositioned as a full 'Agentic Development Environment,' competing with Cursor and Claude Code. That pivot turned a 300-day crawl to $1M ARR into $1M ARR every 10 days.

Warp had integrated AI features before ChatGPT launched, using OpenAI's Codex model to translate English into terminal commands. But the team initially thought about AI narrowly — a chat panel bolted onto the side of the app. The bigger insight came when they realized the entire interface should be reimagined around English-language interaction.

Agent Mode — a term Zach claims Warp coined before it was widely adopted — let developers simply tell the terminal what to do: fix this Git problem, run this deployment, debug why the servers are crashing. That was useful but still terminal-scoped. The June 2024 launch of the Agentic Development Environment went further, adding code editing, file browsing, diff review, and benchmark-level coding performance. Warp is now ranked first on Terminal Bench and third on SWE-Bench.

The competitive framing is deliberately contrarian. Cursor and Claude Code are the obvious competitors — both fast-growing, well-funded, and popular. Zach's view is that this is exactly where you want to compete. The demand for agentic development tools is so large and so early that being in that market is more important than avoiding the competition.

"It's not the developer tools market anymore. It's the market for automating the production of software — something closer to the market for paying software developers to develop software. There's just a very, very low penetration of these agentic tools into that market." — Zach Lloyd

"It's better to be in a market like those are two super fast growing tools. You kind of want to go to where the competition is. There's competition there because the demand for it is insane." — Zach Lloyd

• Agent Mode coined by Warp before the term was widely used — lets developers use plain English in the terminal.
• June 2024 ADE launch added coding, file editing, and diff review alongside terminal workflows.
• #1 on Terminal Bench, #3 on SWE-Bench among coding benchmarks.
• Growth curve went near-vertical after the ADE launch and has continued accelerating.

Warp vs. Traditional Terminal vs. Cursor / Claude Code