Shutting down my baby

[07-2021 / English / 1688 words]

Update: April 2022 - this post made it to HackerNews!

I was always dreaming of running my own company. When I started to learn programming, this idea took the shape of founding a consumer-oriented (B2C) internet-startup. After being accepted in the entrepreneurship association “Austrian Startups”, I decided to commit to my dream fulltime – not knowing what storm of intense work, and what emotional rollercoaster of successes and failures would await me. The past ten months have been an immensely satisfying, frustrating, empowering, humbling, and enlightening time for me. Now, looking back at this time, I realize how much I have changed by working on the fast-track like this. I have made countless new experiences, have learned new things about myself, have organized a hackathon, and have found new friends in the Austrian entrepreneurial ecosystem. It was a lot of fun. However, there were a lot of difficult parts in my startup journey, too. In this post I will share the key lessons that I learned in my attempt to build MechanicalQuill, a textual machine-learning platform for hobby writers.

#1 People don’t care about your product.

Like so many startup founders, I was driven by the motivation to build a product that I would like to use myself (or so I thought). My goal was to allow authors to automatically generate new creative text based on text they had written so far (basically a literary autocomplete). So I built a prototype of that: a dynamic website with integrated machine-learning models. It took me a long time. And, unfortunately, the work really only started at this point. I reached out to technology-savvy authors in my personal network, and also online via different writing blogs to find my first potential customers. The feedback I received was generally favorable, and people thought of my product as a cool gadget. But potentially paying for it? Nah. It turned out that writers actually like to write themselves, and the current level of technology was just not good enough to make them really satisfied with whatever output they got from my product. I had learned the hard way that what really counts is to build a product that people love. As a result of that, I tried pivoting to a different customer base: hobby writers. However, to reach this more broad and heterogeneous customer base, my product was not accessible enough yet.

#2 People can’t read your mind.

Even though nearly everybody on this planet has heard the “machine learning” buzzword by now, the less technologically experienced hobby writers that I talked to didn’t have a real understanding of what they could expect from my product. This fundamental knowledge divide really made me think about how to reach this group of people. As a result, I improved the product’s web user interface, and added descriptive use-cases and visual explanations as to why they should use my website. Not knowing much about marketing, I set out to build a customer acquisition funnel. I made call-to-action buttons, introduced the yellow bird mascot Quilly, and generally tried my best at creating an approachable, fun brand that I thought might appeal to my target customer group. I knew by now that the only way I could build something that people love was to talk to them – a lot. I knew that every minute I spend talking to a potential user was time well spent. Nevertheless, doing user interviews and all the stuff around it (initial pitch, answering some questions, and scheduling a meeting) created a different kind of stress for me.

#3 If you don’t do it yourself, nobody does it.

In my experience, working with at least one co-founder is infinitely better than working completely alone. At every part of the road, I felt an extreme need to prioritize and multitask. There’s just so much to do! Even worse, the work that needs to be done, often has a high “switching cost” – it’s not only coding, or designing, or emailing. It’s everything, in an often haphazard mix driven mainly by prioritization concerns. Speaking in the words of startup guru Paul Graham, I was trying to work on a managers’ and a makers’ schedule at the same time. I ended up splitting my work days into half-day blocks as much as possible. In the mornings I did my focused work on machine learning and web design problems, and in the afternoons and evenings I did all the “smaller” tasks that required less focus, like designing new explanatory graphics and talking to potential customers. But even though I eventually had built a routine for myself that somehow worked, it wasn’t a routine that made me happy.

#4 It can get lonely.

If I ever attempt to found a startup again, I won’t do it alone. Being a part of a good team can be so much fun. I know how it feels to collaborate with highly motivated, likeminded people from my work in the development cooperation association “Engineers Without Borders Austria”. We planned and built a 300 square meter education center in Togo, Western Africa that now trains 20 young women per year in tailoring. It was a brutally hard, completely underfunded project that was on the brink of failure several times. But it was all OK – because everybody in the team was passionate about the project, and ready to extend a hand to struggling colleagues. When founding MechanicalQuill, I wasn’t completely alone. I was part of the encouraging entrepreneurial network of AustrianStartups, and I had found a great supportive mentor in Michael Ionita. Nevertheless, it wasn’t enough for me. Whenever I started my workday looking at yesterday’s notes from myself, addressed to myself, I felt my thoughts drifting to how much better everything would be with the magic of being part of a supportive team. However, just when I had won over my friend Alexander Simunics to give working with me on MechanicalQuill a try, and collaborate on some parts of the startup, my fate was sealed by a different matter.

#5 You need (some) money to make it work.

As the optimist that I am, I had started my startup journey with only a low amount of personal savings from my last job. I didn’t have any illusions that my product would achieve profitability in the timeframe dictated by my financial situation. I was very hopeful, though, that after building a small prototype, I would have a chance at getting outside investment to continue working on my product. The funding landscape in Austria is very different from more mature entrepreneurial ecosystems. I had heard that for very small, early stage startups like mine, that don't have achieved product-market-fit yet, getting an angel investment is quite hard. Therefore, I decided to try my luck with governmental grant programs, instead. The Austria Wirtschaftsservice and the Forschungsförderungsgesellschaft are grant agencies that essentially hand out free money – provided, that they see promise in a startup idea. It turned out to be a long and time-intense process to prove to them that my idea is worth a grant. I ended up spending more than a month of work on the applications, and several months waiting. Upon receiving the rejection letters, I knew that my baby was dead. But it wasn’t all for nothing.

#6 You learn a lot.

Over the course of the last 10 months I had to wear many hats. My responsibilities were a cross-section of everything that makes a (young) company a company. I was a full-stack developer, building a responsive machine-learning web application from the ground up (you can read more about my tech stack and architecture choices in an earlier blog post). I was a UX designer, working on user stories, interface design, and copy writing. I was a CFO, preparing documents on budget planning, and addressable market size. And I was a CEO, communicating and pitching my product to potential users, peers, and governmental agencies. Most of the times, this array of responsibilities left me feeling overwhelmed. I was juggling too many balls, and – inevitably – was almost always falling short of my envisioned goals in each problem area. While this constant overload was extremely stressful, it also left me with a better understanding of the different tasks involved in building a lean, web-based company.

#7 It’s a roller coaster marathon.

There’s a reason why so few startups succeed – and why even less try it in the first place. The work on my startup took over my life. A mixture of financial stress and lack of accountability to a supervisor or colleague pushed me to overwork at the expense of everything else. My personal life suffered severely, and I had less time than I would have liked with my wife, my friends, and my hobbies. I realized, that building a startup isn’t a roller coaster, it’s a roller coaster marathon. It goes up and down a lot, for a long time. Knowing how hard it is to be an entrepreneur first hand, I have all the more respect for the people that manage to succeed despite all these difficulties.

What now?

I want to walk a different road now. I would love to be able to dive deep into technical problems, improve my coding skills by collaborating with colleagues, and learn the ropes from a senior developer. I like to get passionate about my work, and want to bounce off ideas with others. This is why I see myself thriving in a fast-moving, innovative workplace that values strong teams, and cross-sectional communication.