^{Why this story matters}

Most software engineers building side projects lose months chasing the idea they find interesting. Arvid Kahl lost them chasing ideas no one needed. The thing that changed everything wasn't a better idea. It was standing in the right kitchen.

The insight buried in this story is about proximity. Not networking, not market research, not interviews with strangers. Arvid already had someone next to him who experienced a real problem every day. He just had to stop looking past it.

What makes this worth studying isn't the exit number. It's the gap between what the product was (a feedback form tool for English teachers) and what it produced. The more boring the problem, the more people hate living with it. That tension is where the money hides.

‍

^{Where he started}

Arvid was a software engineer living in Berlin with a commute that consumed five hours of every working day. He was technically skilled in full-stack development and serverless architecture, but professionally miserable. The work was fine. Getting there and back was not.

His goal was specific and modest: build something that replaced his salary so he could quit the commute. He wasn't trying to build a company. He was trying to get his days back.

The problem was a history of failed attempts. Side projects that launched and immediately sank. He'd built things he thought were clever. Things that solved nothing specific for no one in particular. Each one confirmed a fear he carried: that he was "just a dev" who couldn't figure out the business side.

‍

^{What they tried first}

Before FeedbackPanda, there were what Arvid called projects that "imploded on impact." The pattern was consistent. He'd pick something technically interesting, build it well, and release it into a void.

The failure wasn't in the code. It was in the premise. He was building for everyone, which is another way of building for no one. Tools that were cool. Tools that solved problems he found intellectually interesting rather than problems people were actively suffering through.

No specific audience meant no feedback loop. No feedback loop meant no way to improve. And no improvement meant abandonment. His and the users'.

‍

^{How they got their first real client}

Arvid stopped looking outward and looked across the kitchen table.

His partner, Danielle, was an online English teacher working on the VIPKid platform. Every day, after teaching, she had to write individual feedback for each student. It was mandatory. It was repetitive. It took two hours of unpaid time that she could never get back.

He didn't do market research. He didn't run ads. He built one tool that solved one problem for one person he already knew. Then asked her whether it worked.

The MVP took one week to build. It automated the feedback process down to five minutes. Danielle was in the community of VIPKid teachers. She shared it. Teachers in that community understood the problem immediately because they lived it every day. They didn't need to be convinced. They needed the tool to exist.

The first customers came through word of mouth inside a tight-knit professional community. No cold outreach. No ad spend. Just a solution circulating among people with identical problems.

_{What the work actually looks like}

‍

At its core, FeedbackPanda did one thing: it turned a two-hour task into a five-minute one.

Teachers would log in, select a student, and generate personalized feedback using templates and automation rather than starting from scratch every time. The tool lived inside their existing workflow on the VIPKid platform. It didn't require a habit change. It replaced something painful with something fast.

Arvid handled development. Danielle ran customer support and understood the users in a way Arvid couldn't. She was one of them. Between the two of them, they handled everything: features, bugs, questions, edge cases, and the slow growth of a business that was adding customers without adding headcount.

A typical week was building, responding, and shipping. No management layer. No process overhead. Just two people keeping a product alive and growing.

‍

^{The tradeoffs}

They never hired.

At five thousand customers, with a two-person team handling all of it: development, support, maintenance, and growth. The pressure had become structural. Not a bad week. Not a rough patch. A permanent state of too much to do with too few people to do it.

By the time they sold in 2019, they were on the edge of burnout. The exit wasn't just a financial decision. It was a necessity.

What they gained was clear: time back, no commute, full control over the work. What they gave up was less visible. Years of headspace consumed by a support queue that only grew. The thing that made FeedbackPanda work, staying lean and close to the customer, was also the thing that nearly broke them. Small teams are powerful until they're not. They stayed small for too long.

‍

^{The number that matters}

FeedbackPanda was sold in 2019 for a mid-seven-figure sum after reaching $20,000 in monthly recurring revenue within nine months of launch.

‍

^{What's easy to miss}

The problem Arvid solved was not glamorous. Student feedback forms for an online English teaching platform. Nobody was writing essays about it. Nobody was following a hashtag for it. It was invisible to anyone who didn't have to do it every day.

That's the point.

The most durable small software businesses often solve problems that people find too boring to discuss publicly. Which means the problem has existed for years with no competition, no press, no solution. Danielle wasn't complaining about it in forums. She was just doing it. Grinding through it. Assuming it was just part of the job.

When you solve a problem that people have accepted as permanent, their reaction isn't "interesting." It's relief. That's a different kind of customer entirely.

‍

^{Buildzone takeaway}

Arvid didn't redesign his work by finding a brilliant idea. He found it by paying attention to what was already happening in his own home.

The commute was what he wanted to escape. But FeedbackPanda wasn't built around that goal. It was built around Danielle's Tuesday afternoons. Her problem. Her community. Her feedback on whether the thing he built actually worked. The business came from the life he was already living, not from a plan he designed in advance.

By the time he sold, the commute was gone. So was the ceiling. What came after, the writing, the teaching, the newsletter, was built the same way. Not from strategy, but from what he'd actually learned by living through it.

The business followed the life. Not the other way around.

‍

^{At a Glance}

• Name: Arvid Kahl
• One-line descriptor: Software engineer who turned his partner's unpaid admin work into a $20k/month SaaS. Sold it three years later and walked away.
• Career before: Full-stack software engineer, Berlin (5-hour daily commute)
• What he built: FeedbackPanda, a student feedback automation tool for online English teachers on the VIPKid platform
• Revenue model: SaaS subscription
• Customers: 5,000+ at time of exit
• Team size: Two people (Arvid + partner Danielle). Never hired.
• Location dependence: None. Fully remote from day one
• Tools used: Full-stack web stack, serverless architecture
• What didn't work: Building "cool" tools for no specific audience; staying a two-person team past the point of sanity
• Transition timeline: 1 week to MVP → $20k MRR within 9 months → mid-7-figure exit in 2019

‍