TalkNotes: Sold for $200,000 in 11 Months — Why the $7,000-MRR AI Voice-Memo App Chose an All-Cash Deal

What makes this case worth reading

An app at $7,000 MRR (about ¥1.05 million) sold for $200,000 (about ¥30 million). That works out to roughly 2.4x ARR (this article’s own calculation, against $7,000 × 12 = $84,000). But the number itself isn’t the most interesting part. What stands out is that the whole run from launch to sale was just 11 months, and that developer Nico Jeannen turned down a higher offer of $300,000 in favor of $200,000 in cash, up front.

There’s a reason he didn’t maximize the sale price. Following that reason leads to the real turning point in this case — and it isn’t Product Hunt, and it isn’t the launch.

The 11-month timeline

What he built

TalkNotes transcribes voice memos with AI, then reformats the transcript into whatever shape you need, a blog post, an email, a task list. Its origin was highly personal: Nico was frustrated with the accuracy of Google Docs’ voice input. A classic starting point of building what he himself wanted to use.

Technically, the early version was simple, vanilla JS/CSS/HTML with a Node backend, and was later fully rebuilt on Nuxt.js. The reason: “to ship more interactive features, faster.” Swapping frameworks after revenue is already flowing is normally an investment that gets deferred, but here, feature-shipping speed itself was judged to be the bottleneck.

The turning point wasn’t Product Hunt

The initial $700, then winning Product of the Day on Product Hunt and hitting $1,500 MRR, up to this point, it’s a familiar launch success story. But $1,500 MRR alone doesn’t get you to a $200,000 sale.

The trajectory changed once he started running Facebook ads. What matters in Nico’s account is that he didn’t use ads purely as an acquisition tool. He built a loop: use ads to acquire users → watch where those users get stuck and what they actually want during onboarding → feed those learnings back into targeting and the product. Running this loop doubled MRR within a few months.

In numbers: $1,500 MRR to roughly $7,000 MRR. Of the $200,000 sale price, Product Hunt’s contribution accounts for something like the $1,500 slice, while the ad loop accounts for the roughly $5,500 slice. The turning point wasn’t “going viral”. It was “the acquisition machine put in place once the viral spike faded.”

Spend matters here: $30–$100/day translates to roughly $900–$3,000/month. At the point MRR hit $7,000, that’s a level that eats into margin, but framed against a sale, the picture changes. Pushing MRR from $1,500 to $7,000 likely moved the sale price by tens of thousands of dollars. The ad spend functioned not as an expense, but as an investment in the exit price.

Why he turned down $300,000

During the sale process, Nico received an offer of $300,000. But it came with conditions, either financed payments or an earnout. He turned it down and chose $200,000 in cash, all at once.

Giving up the $100,000 difference was a call grounded in an understanding of structure, not emotion. An earnout means “you get paid more later if certain performance targets are hit after the sale,” and whether those targets are hit depends on the buyer’s ability to run the business. A financed payment shifts the buyer’s credit risk onto the seller. Seen that way, the $300,000 was not a guaranteed amount but an expected value, and the authority to influence that expected value would no longer belong to him after the sale.

On top of that, the reason he decided to sell in the first place was burnout. May had brought a string of emergency bug fixes, and he was spent. If the goal is to step away from the product, an earnout that keeps him tethered to post-sale performance runs directly against that goal. The $100,000 gap reads as the price of “walking away completely.”

The deal nearly collapsed at escrow

The rawest part of this case is what happened at the settlement stage. Escrow processing through the Acquire.com deal stalled, and the transaction came close to falling apart entirely. Nico announced he would go public about it via a sponsored tweet and pursue legal action, and the refund was processed the following day. The final transfer of funds was completed as a direct transaction between the two parties.

The assumption that using a sale platform makes a deal safe doesn’t hold. A platform’s function is finding a buyer, not guaranteeing that funds actually land. The fact that a deal can still die between contract signature and payment clearing is practical information anyone considering a solo business sale shouldn’t overlook.

The seven years behind the “11 months”

The headline is “$200,000 in 11 months,” but Nico himself explicitly rejects framing this as an overnight success. In his own words, it’s the result of seven years of accumulation, 40+ project launches, most of them failures, 100-hour work weeks, and a life that cut deeply into social activity.

Strip that context away, and the meaning of the case flips. “Built an MVP in a week and made $30 million yen in 11 months” only holds together because it’s the 41st at-bat for someone who has already failed 40 times. The reason an MVP could be shipped in a week, and $700 made in the first 10 days after launch, is that he already knew which directories to post to for early traction. TalkNotes’ 11 months sit on top of the execution speed produced by seven years of failure.

What generalizes, and what doesn’t

What’s easy to reproduce is the design philosophy of having an acquisition mechanism ready for after the buzz fades. A Product Hunt top ranking is a one-time event, even in this case it topped out at $1,500 MRR. What built the growth beyond that was small but sustained paid ads, plus a loop feeding what was learned back into the product. Starting from $30/day, the capital barrier here is low.

Equally reproducible is the judgment axis of not comparing exit offers on price alone. Not the size of the offer, but certainty, timing, and whether you stay tied down after the sale. This lens applies regardless of business size.

Two things are harder to reproduce. One is the execution speed and instinct built from seven years and 40 projects, something only time can buy. The other is the timing of late 2023, when demand for AI voice/transcription products and buyer interest in the category were both elevated at once. A product at $7,000 MRR won’t always sell at roughly 2.4x ARR, whether a buyer exists at all in a given market is a variable outside the seller’s control.

And last: burnout is the one thing here nobody should hope to reproduce. That the trigger for the sale was burning out means this case is, at the same time it’s a success story, also a record of the limits of running a business alone.

Related reading

• ScrapingBee’s exit — how two developers grew a SaaS to the conditions under which it sold
• Damon Chen and PDF.ai — a solo-built AI tool taken to profitability in a short window