Amjad Masad: “We’ll hit $1B revenue this year”

Why does it matter? Because $2.5M to $250M in one year isn't a fluke — it's what the real thing looks like

Amjad Masad spent nearly a decade building Replit while his team lost faith, partners quietly scrubbed his logo from their websites, and a cavernous Foster City office sat half-empty. Then he shipped a shaky iPhone video of a half-broken agent and watched revenue explode 100x in twelve months. What he learned about founder psychology, where AI value actually accrues, and how to recognize the difference between signal and noise is blunter than most startup playbooks.

• Real PMF hits like a landmine — unmistakable and violent. Six or seven years into startups, Amjad had never felt it. If you're still debating whether you have it, you don't. • The leading indicator of terminal failure isn't your bank balance. It's the visible decay of your team's belief in you. • Foundation model companies have no durable moat. A commodity substrate is the best news app-layer entrepreneurs have gotten in years. • The next defensible million-dollar business is probably hiding in an industry still running on clipboards.

Watching your team lose faith in you is worse than running out of money

The bank balance wasn't the problem. "The belief that your team have in you, your vision, your leadership — when that goes away, you can see it in their eyes. And that is the most hurtful and depressing feeling." Amjad said it plainly, and Shaan stopped him cold: "You just described my childhood trauma as a founder. I didn't even realize it until you said those words."

Replit's headcount went from 120 to 60 in three months — most left voluntarily. The new Foster City office was cold and dark. Every night, whatever sleep he got was spent wondering who would quit the next day. Partners dropped Replit from their websites. Vendors quietly removed the logo. Conference invites dried up. The investors who once backed him went from "cut burn" to "go invest" — but only after the turn came.

One room was different. The war room, where a small team was building Replit Agent, operated like a different company entirely. "Everything is depressing except this one room — the mood is intensely different. Everyone is super pumped." Amjad bounced daily between a company bleeding out and a secret that was about to change everything.

The pressure after the breakthrough is its own burden. "If we had sold or failed before the breakout point, I would have been depressed — but not as bad as when you now have the wind in your back and if you fail, it is fully on you." Clear path to the basket, no defenders, and you air-ball it. That's the feeling.

Stop measuring startup health in runway months. Team belief is the actual leading indicator — it decays silently, long before the bank account moves.

If you're debating whether you have product-market fit, you don't

A 21-year-old working out of Replit's office had posted in a crypto mining subreddit about renting unused GPUs — and thousands of people showed up instantly. The demand side wasn't there yet, so he pivoted. Amjad asked why he was walking away from such strong supply-side signal. The kid said: "When we did that first thing, it was like we stepped on a landmine. Now I know what's possible — so I know these little signals aren't real."

Amjad went back to his desk and sat with that. Six or seven years of startups, and he'd never felt what this kid had stumbled onto on day one.

"You don't even know what you're looking for if you've never felt it. It's like love — am I in love? Maybe this is it. And then you feel it and you're like, oh. All that other stuff wasn't that. This is the thing."

Sam had heard the same thing from the Twitch CEO: it feels like pushing a boulder uphill every day, then waking up to find the boulder already moved — and now you're sprinting to catch it. The whole game flips from push to chase.

The danger is rationalization. Founders who've never felt the real thing will accept mediocre signals as confirmation, spending years on a path that was never going to break out. Amjad's frame is binary: "If it's not hitting, you're not hitting on some human nature or element. You're like finding a secret in the universe almost." Keep pivoting until you step on the landmine. It doesn't sneak up on you.

Replit's agent was half-broken at launch. Day one: $1M ARR. Day two: $2M.

Andrej Karpathy called it "the AGI moment." Researchers at OpenAI and Anthropic reached out afterward to say they hadn't known their own models were capable of what Replit's demo showed. The demo was a shaky iPhone video, shot in the office, posted to Twitter.

The engineering team wanted to wait. Amjad overruled them. "I don't care if it is semi-broken. If you get 50% of the time get amazing results, it's going to wow the world — because for the first time, an agent can not only write the code, debug it, create a database for you and deploy to the cloud. This has never been done."

The framing came from his Steam Deck — bought during the darkest stretch to take his mind off everything. Games ship "early previews": semi-broken products where users know what they're signing up for and feedback is part of the deal. Amjad repackaged the launch the same way. Don't subscribe if you can't handle bugs. But here's what's now possible.

September 2024. Tweet goes viral. Karpathy retweets. $1M ARR day one. $2M day two. Twelve months later, $2.5M had become $250M. Targeting $1B this year.

When you're first to demonstrate a genuinely new capability, a 50%-working proof-of-concept that goes viral beats a polished product arriving six months later into a crowded field. The window to define a category is short. Ship before you're ready.

LLMs have no moat — and that's the best news app-layer builders have gotten in years

Every model is one click from being swapped out. Amjad has read Seven Powers — Hamilton Helmer's theory of competitive advantages — and he's applied it directly: "A lot of developers every day, professional developers, are switching between these models in some cases, and if you're in Cursor it's literally one click away to switch to the next model."

The historical moats — network effects, economies of scale, lock-in — haven't materialized for any foundation model company. The only candidate Amjad sees is capital: you need continuous funding to train the current model, the next generation, and the one after. But capital doesn't block the other well-funded players, and it won't stop China from subsidizing its way in and potentially collapsing global pricing the way it did with EVs.

"If we ended up with a monopoly or just oligopoly and you're building on top of it, it is actually very hard to build a successful business." A monopoly would eventually come for you. But a commodity substrate is different — the value stays at the application layer, with whoever builds real workflows and real user relationships on top of cheap, swappable models.

Models are potential energy. Entrepreneurs convert it into products. That's where the value lands — and it looks increasingly permanent.

Zero-sum execution and market creation demand opposite strategies — and confusing the two costs years

Most founders are playing the wrong game — not because they chose the wrong market, but because they misread which type of market they're in.

Amjad invoked Clayton Christensen — who he noted recently passed — to draw the line. Sustaining technology makes existing things incrementally better; you compete for a slice of demand that already exists. Mars Candy is zero-sum. Most businesses are zero-sum. Disruptive technology creates demand that didn't exist because for the first time something becomes possible. Google, Facebook, PayPal, Replit Agent — market creation. "Those are moments of time businesses or technologies that create this explosion of demand because for the first time it became possible."

The strategic split is clean. Zero-sum market: write the business plan, project market share, execute relentlessly. "Be very good at execution. Build amazing processes. Out-compete through the day-to-day." Market creation: the playbook doesn't exist because the category doesn't exist yet. "Pivot, pivot, pivot until it hits. Because if it's not hitting, you're not hitting on some human nature or element."

Patience is a virtue in zero-sum fights. Relentless pivoting is the only move when you're trying to create something new. Apply the wrong approach to the wrong game and you burn years you won't get back. The first question isn't "what's my go-to-market?" It's "which game am I actually playing?"

Sales is a contact sport — Replit is converting more than half the company into salespeople

Four sales reps at end of last year. More than half the company in sales by end of this year. That's a jarring ratio for a company built on pure product culture, and Amjad is clear about why it took him so long to see it.

Consumer growth is weather. Ship something, run an A/B test, wait two weeks. Someone writes about you, you go viral, you die down. Half the inputs are outside your control.

Sales is different. "I can apply effort and whether we get the win or not is somewhat much more within our control." When a deal goes sideways, Amjad activates completely. "The idea of losing to some competitor is like the worst feeling in the world. I'll call whoever I need to call, I'll show up to their office, we'll do whatever we need. And we rarely lose deals."

Shaan clocked it immediately: deadlifts before it was cool, likes sales before it was cool, co-founded with his wife. "My guy's got range."

The VP of sales who joined during Replit's lowest moment — Patrick Purvis, ex-VP at ZoomInfo, independently wealthy and on sabbatical — said he came because democratizing software felt important. Amjad told him there was basically nothing to sell yet. He said "welcome to the team" anyway.

Tech founders who treat sales as a necessary evil are leaving the most controllable growth lever untouched. The outcome is within reach. Show up.

The next million-dollar startup is in an industry that still doesn't have software

There's a guy in England building management software for ice rinks. Already at $100K run rate, heading toward a million. No VC, small team.

Amjad's thesis is structural: Silicon Valley defaulted to hyper-scale because building software was expensive, and only venture economics could justify the cost. AI has broken that equation. A solo founder can now build a profitable, defensible vertical SaaS without raising money — and the niches that were previously too small to bother are suddenly real businesses.

Two discovery methods. Paul Graham's: live in the future and build what doesn't exist. Zuckerberg didn't theorize about social networking — his social life was already online, he just built the tool he wanted. Amjad's own version is lazier: "Go about your life with this laziness attitude. What are the things that are so annoying that you have to keep doing over and over again? There's probably a million, maybe a billion people that feel the same way." His first business was internet cafe software because he literally grew up in those cafes in Jordan.

Medvoy — a GLP-1 startup the New York Times called a billion-dollar one-person business — runs major parts of its stack on Replit. Magic School, a $500M company, started there. Try Nearby, which connects Gen Z to local restaurants via TikTok-style influencers, hit $100K ARR in weeks after launching.

Stop chasing platform-scale. Walk around and find the industries still running on spreadsheets. The economics finally work for the niche.

The capability overhang is real — and for most verticals, the window is still open

Amjad said it without hedging: "I think we are in the singularity." New models every few weeks. Fundamental capability shifts in autonomy, computer use, cybersecurity — each one creating a backlog of products that haven't been built yet. The potential energy is accumulating faster than entrepreneurs can convert it.

What Replit proved is that the category-defining moment is brutal and fast. Be first to show what's possible, and the market forms around you overnight. Everyone else is running catch-up from day one.

For most verticals, that first-mover slot is still unclaimed. Move.

---

Topics: AI, startups, founder psychology, product-market fit, Replit, vibe coding, LLM commoditization, enterprise sales, revenue growth, entrepreneurship, market creation, vertical SaaS