December 18, 2024 was the day the math stopped working for Maddox Schmidlkofer’s internship. DuckMath, the gaming site he’d built as a teenager, earned $240 that day — more than his eight-hour shift at the software company where he was interning — and he’d done no work on it at all. “That was the moment I knew I had to quit my internship, I put in my two weeks right then and there,” he said, according to a press release from FreezeNova, the company that would eventually buy the site.

Less than a year later, at 20 years old and still a computer science student at Purdue, Schmidlkofer sold DuckMath for $120,000. The site was pulling in roughly $15,000 a month. On paper, he sold a business for less than one year of its revenue. The strange part — and the reason this story is worth telling — is that he was probably right to do it.

From a high school workaround to 1.5 million users

DuckMath began in 2021, when Schmidlkofer was a high school student in Carmel, Indiana, as first told by AI & No-Code Exits. The premise was simple: an “unblocked games” site, a corner of the internet where students play browser games on school networks that filter everything else. “I saw somebody make a site like that on TikTok and they were younger than me,” he told the Purdue Exponent, and figured if they could do it, so could he. According to the exit database FounderSold, the first version was built on Google Sites when he was about 16.

The site itself was ordinary. The growth engine was not. Schmidlkofer posted short-form video every day — TikTok, YouTube Shorts, Instagram Reels, Snapchat — and studied what made competitors’ videos work. FreezeNova’s own announcement of the deal describes him posting 12 short-form videos a day and iterating on whatever hit. A viral wave sent DuckMath from 5,000 daily users to 150,000 in two weeks. By the time of the sale, the site served roughly 1.5 million monthly users and, by his own account, had reached more than 5 million students worldwide.

But the daily cadence was both the moat and the trap. The unblocked-games niche rewards constant fresh promotion, and Schmidlkofer’s traffic rose and fell with his output. He has acknowledged that revenue tracked his content almost one-to-one, and the buyer’s announcement credits the growth to that relentless daily posting. Which raised an uncomfortable question when buyers came calling: what exactly would they be purchasing?

He turned down a first offer of $100,000. Then, in November 2025, according to FreezeNova’s announcement, he sold to FreezeNova, the biggest competitor in his own niche, for $120,000. It wasn’t even his first exit: by his own account he’d previously sold a smaller project called MaddoxCloud for $15,000, making him a two-time founder-seller before finishing his degree.

The deal by the numbers

Why 0.7x wasn’t a fire sale

Here’s where the independent data sharpens the story. According to FounderSold, DuckMath sold at 0.7x annual revenue — a sub-1x multiple that looks like a fire sale until you consider what a multiple actually prices. A multiple assumes the earnings are a property of the asset, something that persists after the seller walks away. Schmidlkofer’s earnings were a property of his daily labor. Strip out twelve videos a day and the buyer isn’t purchasing a machine; they’re purchasing a treadmill with the runner missing.

The buyer choice mattered as much as the price. FreezeNova wasn’t a stranger inheriting his workload — it was the largest operator in the same category, with its own existing audience to plug DuckMath’s games into. The site’s core weakness, its dependence on one student’s posting habit, effectively disappeared inside the acquirer’s distribution. The asset was worth more to the one buyer whose model erased its flaw.

Context also softens the headline number. Per the Purdue Exponent, Schmidlkofer plans to invest roughly half the proceeds and keep the rest for his next projects. Add the earlier $15,000 MaddoxCloud sale, and a 20-year-old holds seed capital, a safety margin, and two closed negotiations’ worth of experience.

What AI and no-code founders should take from this

• Run the 30-day test before a buyer does. If you stopped working for a month, what happens to revenue? Schmidlkofer knew his answer and priced accordingly, rather than letting diligence discover it and kill the deal.
• Sell while the numbers are strong, not after the treadmill breaks you. A Purdue course load and a daily posting habit couldn’t coexist forever. He exited from a position of verifiable strength.
• Pick the buyer whose strengths cancel your weaknesses. A competitor with existing distribution valued DuckMath more accurately — and more durably — than any outside buyer could.
• Rejection doesn’t have to be dramatic. He declined $100,000 and closed at $120,000: a 20% improvement from patience, not brinkmanship.
• Let the lesson shape the next build. He has described his next project as Tallow, a native iOS app with subscription paywalls designed from day one — a product meant to earn without a daily video feeding it.

The question worth asking tonight

Schmidlkofer’s story isn’t really about games, or TikTok, or even the $120,000. It’s about a founder honest enough to see that he owned a job dressed up as an asset — and disciplined enough to sell it like one, to the one buyer who could turn it into more. So run his test on whatever you’re building: if you stepped away for thirty days, would the revenue wait for you, or would it leave with you? Either answer is fine. The only mistake is pricing one as the other when someone finally shows up with a check.

Sources

• He Sold a Site Making $150K a Year for $120K. — AI & No-Code Exits (original reporting)
• Purdue computer science student sells gaming site for $120k
• 20 Year Old Purdue Student Sells Viral Gaming Startup to Global Giant …
• Carmel resident creates, sells unblocked gaming platform
• DuckMath Acquired for $120,000: Exit Details | FounderSold
• Run 3 Freezenova Unblocked – Duckmath
• duck – Freezenova

AI Founder Stories is a weekly Feature Paper series. This feature builds on reporting first published by AI & No-Code Exits, with additional research by The Feature Paper. Browse the series at https://featurepaper.com/ai-founder-stories/.