Cluely's Roy Lee Got $15M From A16z - Then Stopped Talking About Revenue

Roy Lee knows how to get attention. Getting expelled from Columbia University for building a cheating tool?

That'll do it. Launching a startup with rage-bait marketing claiming to help people "cheat on everything"? Even better.

But four months after boasting that Cluely's ARR skyrocketed from $3 million to $7 million in a single week, Lee has gone quiet on the numbers.

"What I've learned is you should never share revenue numbers," Lee said onstage at TechCrunch Disrupt 2025 last week.

From Cheating Tool to... AI Meeting Assistant?

Cluely burst onto the scene in April with a provocative pitch: an AI tool that delivers undetectable information during online conversations.

Lee first made headlines when he was suspended from Columbia University for building a tool used for cheating on coding job interviews. He channeled that controversy into Cluely's launch strategy.

The virality worked.

Andreessen Horowitz backed the company with a $15 million Series A in June. Partner Bryan Kim said on the firm's podcast that he invested because Lee had figured out how to convert attention into paying customers.

But the product has been shape-shifting ever since. In late June, Cluely introduced an enterprise product for sales calls, customer support, and remote tutoring.

This week, the startup narrowed its focus again, repositioning as an AI meeting assistant and note-taker.

"The whole idea [was] let's launch something that barely works, and if we can get enough initial users, they will find out the use cases for us," Lee admitted onstage.

Translation: they're still figuring it out.

The Revenue Question Nobody's Answering

When Cluely launched its consumer product over the summer, Lee was eager to share metrics. "Every single person who has a meeting or an interview is testing this out," he told TechCrunch, flaunting that one-week jump to $7 million ARR.

Now? Silence.

Lee deflected questions about sales and retention, offering only: "I'll say we're doing better than I expected, but it's not the fastest growing company of all time."

His explanation for the pivot to secrecy? "If you're doing well, nobody is going to talk about how well you're doing, but if you're doing poorly, then everybody will only talk about how poorly you're doing."

That logic doesn't hold up in 2025's AI boom, where dozens of founders at fast-growing startups regularly share explosive ARR numbers. It's become standard practice - a signal of momentum and validation.

The Lesson: Virality Doesn't Equal Revenue

Lee himself admits that while startups should be thinking harder about social media virality, brand awareness alone won't sustain growth.

"I can't say if it's a mistake, but maybe we launched too early," he said.

Cluely's trajectory suggests a hard truth: attention is easy to manufacture (especially with controversy), but converting that attention into a sustainable business is an entirely different game.

The startup entered a crowded AI note-taking market, touting features like "sending follow-up emails" - hardly differentiated functionality. They're competing against established players with clearer positioning and proven retention.

Four months after claiming $7 million ARR, the fact that Lee won't share updated numbers speaks volumes. Either growth has stalled, or the company is dealing with retention issues - or both.

What Happens Next?

Cluely has $15 million in the bank and a founder who knows how to generate buzz. That buys time. But time to do what, exactly?

The company needs to pick a lane and prove it can retain customers, not just attract them. AI meeting assistants are a commodity market.

If Cluely's differentiation is "the founder once got expelled for building a cheating tool," that's not enough.

The real test comes next year. Can they convert their early users into long-term customers? Can they build features that justify switching costs? Can they move beyond the viral launch and become a real business?

Or will Cluely become a cautionary tale about mistaking attention for traction?