Lessons learned from my second failed startup

[02-2023 / English / 1056 words]

Many things have happened since I shut down my last startup: I started working as a software engineer, found a co-founder, and launched a new startup in the online education sector. We applied to YC, and got rejected. I doubled down and quit my job to work full-time on the startup. We pre-launched our service, but didn’t get traction, and decided to shut everything down.

Looking back on this intense year, the thing that stands out to me the most is how I failed to learn from my own lessons that I had written down in my first startup’s post mortem. It’s kind of ironic that I was right all along, but didn’t listen to my own advice. But, nevertheless, I have also made new experiences that give me more clarity for how I want my future entrepreneurial path to look like. So, in the style of my last postmortem, I am describing the pitfalls that I ran into while working on Frobocode, and the lessons that I learned along the way.

#1 B2C is really, really hard (don’t do it).

My last startup failed because I ran out of money. This is pretty much the norm with bootstrapped Business-To-Consumer startups, which tend to take a longer time to make money than their Business-To-Business counterparts. I knew that. It’s hard for me to wrap my head around why I still chose to launch a B2C, again (and bootstrap it, again, too). I credit this to getting too excited about my idea. And of course, I also thought that this time it was different.

Now, with a second failed B2C startup under my belt, I am finally seeing that working on an “unsexy” problem in a B2B setting is actually the smart choice to make. I am skeptical whether the “classic” iterative startup approach can even work in the B2C space without a significant upfront investment that provides enough runway for product experimentation. I am done with getting blinded by sexy problems, and the fantasy of quickly gaining thousands of users. Instead, I think I am finally grown up enough to find a niche in the business field, where I can laser-focus all my attention on really making a difference for a small group of customers.

#2 Find a complementary co-founder.

The one thing that I got right this time was to team up with a co-founder. I am very glad to have worked with Yury - it was a lot of fun, and I’ve made a new friend in him. But what Yury and I quickly realized is that we have a pretty similar profile, with skills mostly in the technical field. We’re both self-taught programmers that are more interested in building than in talking to users. With Yury being slightly more interested in business topics than me, we agreed he’d take the CEO seat. Looking back, I recognize that having two programmers in a two-person company gave us a great boost in development speed. However, our lack of excitement about marketing and user research left us coding without a clear direction.

Even though it makes me feel like a programmer cliché, I only recently realized how hard it actually is to be “good” at the business part of a startup. I still want to be actively talking with users, and strive to apply principles from “The Mom Test” in practice. But, being somewhat of an introvert who is quickly drained by interactions with strangers, I had to face the truth that I simply have to expend way more energy to get – at best – results even approaching what a more extroverted person might achieve.

In the brutally hard world of startups, it is not enough to be decent at something. You have to be quite good to make a difference and, probably even more importantly, enjoy doing what you do in order to not get frustrated by setbacks. Knowing this, and being more conscious of where my strengths are now, I want my next co-founder to be having a complementary skillset, focused on the “business” side of things.

#3 Think hard about your business model.

With Frobocode, I wanted to revolutionize the code education space. Being a fan of choose-your-own-adventure books, I wanted to combine their nonlinear, engaging storytelling with a beginner-friendly course on web development. However, I quickly learned that the online course landscape, if seen through a startup lens, is fundamentally dominated by the chicken/egg problem. You don’t get customers without a course – after all, every shill on the internet can claim that they have the perfect resources to learn X. But building the course first, especially with the level of quality I had in mind, would have been a significant time investment upfront with no feedback in the building stage – exactly the opposite of efficient, fast customer discovery.

My takeaway is that the online course business is not really startup-friendly, since quality seems to not be scalable. The exception to that would be a platform where other people are incentivized to create online resources. Crowd-sourced course material, however, usually tends to be of low quality, or is structured into bite-sized portions that lack the cohesion of a well thought out course. In the end, a good programming course is like a good book: there needs to be strong research, clear structure, and well-crafted pacing. But I wanted even more – an engaging fictional narrative, artwork, and a rudimentary inventory system. With these ideas, I had left the startup world, and had entered the publishing world. I wanted to make an interactive book, and not build a scalable business.

I’ve learned the hard way that coming up with a viable business model should never be an afterthought. In the future, I want to set myself more realistic and feasible goals, and plan to choose a simple business model, like for example subscription-based pricing for enterprise software.

What now?

I am currently trying to build myself a predictable income stream in the form of a self-founded web dev company. But my eyes are still set on the startup world, and I am ready to enter the ring again if a good opportunity presents itself. Are you launching a B2B startup, and do you have a business background? If so, let’s talk! :)