We Got a Billion Views and Almost No Downloads
One billion views. Say that number out loud. It sounds like the kind of milestone that comes with champagne, a press release, and a line in your Series B deck that makes investors lean forward.
We hit it in three months.
And then we looked at our download numbers and felt something closer to vertigo than victory.
This is the story of how we built one of the most aggressive content distribution machines in startup history, what a billion views actually bought us, and why we're now doing something we never thought we'd need to do: writing a blog.
TL;DR: Cluely generated over 1 billion social media views through a 700+ person clipper network and 50+ content creator interns. We hit $7M ARR and 70,000+ signups. But the ratio of views to downloads revealed a brutal truth about viral marketing that most growth teams won't talk about publicly. Viral awareness and real product adoption are connected by a bridge that most startups never build. We almost didn't build it either.
The Number
Let's start with the raw output.
In our first three months of executing the content playbook, Cluely content — original, clipped, remixed, and reposted — accumulated over 1 billion total views across TikTok, YouTube Shorts, Instagram Reels, and X.
Here's what that looked like week by week:
- Week 1: 4 million views using our initial distribution playbook
- First 6 weeks: 38 million views on TikTok alone
- First 30 days of the clipper swarm: 24 million cumulative views
- Best single video: 7.8 million views (a dating scenario that had nothing to do with our core product)
- Peak production: 200+ videos published per day
Those numbers weren't an accident. They were the output of a system we designed from scratch — a system that prioritized eyeballs above almost everything else.
And for a while, it worked exactly as designed.
The Machine
Most startups hire engineers first. We hired content creators.
Before we had our fourth engineer, we had 50+ content creator interns producing videos full-time. This wasn't a side project or a "brand awareness initiative" someone put on a roadmap. Content was the company strategy. Our CEO Roy Lee was clear about this from day one:
"The only goal of the company is to get 1 billion eyeballs on Cluely."
Here's how the machine worked:
Layer 1: The Intern Army
Fifty-plus content creators, most of them working six-day weeks from housing we provided within a block of our San Francisco office. Their job was simple: produce short-form video content featuring Cluely in scenarios designed to be shareable. Interview prep. Sales calls. Exam cramming. Dating advice. If it involved a conversation where someone might want an unfair advantage, we filmed it.
The hiring criteria? You needed 100,000+ followers on at least one platform. Engineering skills were irrelevant. Influence was the only qualification that mattered.
Layer 2: The Clipper Network
This is where the scale came from. We built a network of 700+ outsourced content redistributors — clippers — primarily based in Eastern Europe. Each clipper received our raw content and reposted it across their own accounts, sometimes edited, sometimes verbatim, always with our branding intact.
The economics were straightforward: roughly $1 per 1,000 views. Bonuses kicked in at the million-view mark. Some clippers ran multiple accounts. The best ones were pulling in real money.
Layer 3: The Multi-Account Strategy
We didn't post content from one Cluely account and hope the algorithm smiled on us. We pushed content across dozens of accounts simultaneously. Roy's insight was arithmetic, not creative:
"If you make 100 different videos, one will go viral. If you repost that viral video across 100 accounts, 20-30 will go viral too."
At peak, our goal was 2,000+ content contributors, each posting four or more TikToks per day. The strategy was volume at a scale that most content teams would consider absurd.
Layer 4: The Controversy Engine
We didn't just want people to watch. We wanted them to argue. Cluely's brand was built on provocation — "cheating tool for literally everything" isn't a tagline designed to make people feel comfortable. It's designed to make them feel something.
Roy's rule of thumb: "If half the audience doesn't hate it, it's not viral enough."
This worked. Our content consistently split opinion, which is exactly what short-form algorithms reward. Engagement isn't about likes. It's about reactions. And nothing generates reactions like a 21-year-old Columbia dropout telling the internet that cheating is the future.
On X, Roy personally posted 1,000+ tweets in three months, reverse-engineering virality through sheer A/B testing volume. Each tweet was a data point. The ones that hit got amplified. The ones that didn't got analyzed and replaced.
The Conversion Gap
Here's the part nobody puts in their pitch deck.
Roy wrote about this himself in September 2025, in a moment of unusual candor for a founder mid-fundraise:
"49M views on a UGC video, less than 100 downloads."
Read that again. Forty-nine million people watched a video about Cluely. Fewer than a hundred of them downloaded the product.
That's a conversion rate so low it barely registers as a rounding error. It's 0.0002%. If you showed that number to a performance marketer, they'd assume the tracking was broken.
But the tracking wasn't broken. The number was real. And it exposed something fundamental about the relationship between viral awareness and product adoption.
What Was Actually Happening
When we mapped the user journey from viral video to product download, the funnel looked like this:
- View the video (millions)
- Feel entertained or outraged (most of them)
- Remember the brand name (some of them)
- Google "Cluely" (a fraction)
- Visit the website (a smaller fraction)
- Understand what the product does (even fewer)
- Download and install (almost nobody)
Steps 1 through 3 were working beautifully. We had 38% week-over-week increases in branded searches without spending a dollar on Google ads. People were hearing about Cluely. They were talking about Cluely. They were tweeting about Cluely.
But steps 4 through 7 were a wasteland. The bridge between "I've heard of this" and "I use this" didn't exist.
The Numbers That Did Work
It wasn't all bleak. Our viral strategy did produce real business results — just not at the ratio you'd expect from a billion views:
- 70,000+ users signed up shortly after our seed round
- We went from $0 to $3 million ARR in 67 days
- ARR then doubled to $7 million in a single week after our a16z Series A announcement
- We closed a $2.5 million annual enterprise deal with a public company
Those are real numbers. Serious numbers. The kind of numbers that get you a $120 million valuation and a check from Andreessen Horowitz.
But look at the ratio. A billion views produced 70,000 users. That's one user per 14,285 views. For context, a well-targeted Google search ad converts at roughly one in fifty. Our viral machine was 285 times less efficient per impression than paid search.
The viral strategy wasn't failing. It was succeeding — at generating awareness. It was just terrible at generating the specific kind of intent that turns a viewer into a user.
What Views Actually Buy You
So why did we keep doing it?
Because viral views buy you something that no other marketing channel can deliver at that price: cultural presence. When a billion people have seen your name, you're no longer an unknown startup. You're a reference point. Reporters write about you without being pitched. VCs take your call. Enterprise buyers have already heard of you before your sales rep sends the first email.
Our a16z Series A didn't happen because Bryan Kim and Eric Zhou saw a TikTok and clicked a download link. It happened because Cluely had become impossible to ignore. The content machine created a gravitational field around the brand that pulled in press coverage from TechCrunch, Fortune, Business Insider, The New York Times, and twenty-plus other outlets.
That coverage was organic. We didn't hire a PR agency until much later. The controversial brand, the founder story (expelled from Columbia for building an AI cheating tool, raised millions anyway), and the sheer volume of social proof made us newsworthy by default.
Viral marketing bought us brand equity at a cost that traditional advertising couldn't touch. What it couldn't buy us was the middle of the funnel.
The Missing Middle
Here's what we didn't have:
When someone saw a Cluely video and wanted to learn more, they Googled us. And what did they find?
Our homepage. A Wikipedia article that mentioned our data breach. Some GitHub clones offering free alternatives. A handful of reviews from competitors. And three blog posts — all written by Roy, none longer than 1,200 words, none targeting any search keyword, and the most recent one already months old.
We had no comparison pages explaining how Cluely differed from Otter.ai, Granola, or Final Round AI. No guides explaining how our real-time AI actually works. No content addressing the most common questions people asked after seeing our videos: Is it legal? Does it actually work? Is it safe?
Meanwhile, our competitors were eating our lunch in search. Final Round AI had published hundreds of SEO-optimized articles. Interview Sidekick had 190+ blog posts. Even smaller competitors like Sensei AI and Beyz AI were publishing reviews of us — controlling the narrative about our product on a channel we'd completely ignored.
The irony was painful. We'd built the most effective attention machine in our category, and then left the door wide open for competitors to capture the demand we'd created.
Someone would watch our TikTok, get curious, search "AI interview assistant" — and find Final Round AI's content ranking #1. We'd paid for the awareness. They'd built the capture mechanism.
Why We're Building a Blog Now
Roy had it right when he wrote this in September 2025:
"Virality can give people a reason to try your product, but it will not give them a reason to stay."
We'd add a corollary: virality can put your name in someone's head, but it can't be there when they go looking for answers.
That's what search content does. It's the difference between a billboard on a highway and a shop on the street someone is already walking down. Both have value. But only one of them is there at the moment of intent.
When someone searches "best AI meeting assistant 2026," they're not casually scrolling. They're actively evaluating. They're closer to a purchase decision in that single search than in a million passive video views. And if we're not showing up in those results, we're losing that person to whoever is.
Our social strategy generates awareness at the top. Our product converts at the bottom. What we were missing was everything in between: the educational content, the comparison pages, the guides, the trust signals — the entire layer of content that turns "I've heard of Cluely" into "I understand what Cluely does and why I should use it."
That's why we're investing in SEO and content marketing now. Not because we're abandoning what works — the content machine is still running. But because we've learned that awareness without capture is just expensive entertainment.
Lessons for Other Startups
We've spent the last year running one of the most unusual growth experiments in recent startup history. Here's what we'd tell other founders:
1. Viral Views Are Real But Hollow
A billion views is not a vanity metric. The brand awareness it creates has tangible downstream value in fundraising, press, recruiting, and enterprise sales. But if you treat views as a leading indicator of product growth, you'll spend six months celebrating a number that doesn't show up in your activation metrics.
2. You Cannot Brute-Force the Middle of the Funnel
We tried. We thought if we just generated enough awareness, some percentage would naturally trickle down to signups. Mathematically, that's true. But the conversion rate is so low that you need an absurd amount of top-of-funnel activity to move the needle — and the economics eventually break.
3. The Channel That Creates Demand and the Channel That Captures It Are Different
This is the fundamental lesson. TikTok creates demand. Google captures it. They are complementary, not interchangeable. A startup that does one without the other is running on one leg.
4. Content Creators and Content Marketers Are Different Roles
We had 50+ content creators making incredible short-form video. What we didn't have was a single content marketer writing search-optimized articles. These are different skills, different strategies, and different KPIs. We conflated "content" with "video" and missed an entire channel.
5. Your Competitors Will Capture Your Demand If You Don't
While we were generating a billion views, Final Round AI was publishing hundreds of articles targeting the exact keywords our videos were making people search for. They didn't need to create demand — we did that for them. They just needed to be there when the demand showed up in a search bar. This is the most expensive lesson we learned.
6. Build the Capture Layer Before You Need It
SEO compounds. A blog post published today might not rank for three to six months. If you wait until you need organic traffic to start building it, you're already twelve months behind. We should have started writing the day we started filming.
What Comes Next
We're not abandoning the playbook that got us here. The clipper network is still running. The content machine is still producing. The controversy engine is still generating the kind of polarizing attention that algorithms reward.
But we're building the other half now. You'll see comparison guides, product deep-dives, and articles that answer the questions people are actually searching for — like whether meeting bots are actually dead (spoiler: we think so) and how real-time AI assistance actually works under the hood.
A billion views taught us that attention is not the same as adoption. The next billion will be different — not because we'll stop creating attention, but because this time, we'll have somewhere to send it.
Roy stopped sharing ARR charts after we hit $7 million. What he didn't stop doing was obsessing over the gap between awareness and conversion. This blog is part of closing that gap.
Because in the end, the most interesting growth story isn't the one about how you got a billion people to look at you. It's the one about what you did when you realized most of them looked away.
FAQ
How did Cluely get a billion views so quickly?
Through a three-layer content distribution system: 50+ in-house content creator interns producing 200+ videos per day, a 700+ person outsourced clipper network reposting content across multiple accounts, and a multi-platform strategy spanning TikTok, YouTube Shorts, Instagram Reels, and X. The controversial "cheating tool" branding also drove significant organic sharing and media coverage.
What was the actual conversion rate from viral views to downloads?
Extremely low. Roy Lee publicly shared that one video with 49 million views generated fewer than 100 downloads — a conversion rate of roughly 0.0002%. Even aggregated across all content, the view-to-signup ratio was approximately 1 user per 14,285 views, compared to roughly 1-in-50 for well-targeted search advertising.
Did the viral strategy generate any real revenue?
Yes, substantial revenue. Cluely reached $3 million ARR in 67 days and doubled to $7 million ARR in a single week following the a16z Series A announcement. The company also closed a $2.5 million annual enterprise contract. The viral strategy was effective at generating brand awareness that drove press, fundraising, and enterprise interest — just not efficient at driving direct consumer downloads.
Why is viral marketing ROI so hard to measure?
Because viral content creates value across multiple channels simultaneously — brand recognition, press coverage, investor interest, recruiting pipeline, enterprise credibility — most of which don't show up in a direct attribution model. A TikTok view that doesn't lead to a download might still lead to a Fortune article that leads to an enterprise deal six months later. The ROI is real but distributed across time and channels in ways that resist simple measurement.
What is the difference between content marketing and viral marketing?
Viral marketing generates awareness through shareable, often emotional or controversial content designed for social algorithms. Content marketing creates search-optimized educational material designed to capture existing demand. Viral marketing creates demand at the top of the funnel; content marketing captures it in the middle. The most effective growth strategies combine both — using viral content to put the brand in people's heads and search content to be there when they act on that awareness.