The 5 Biggest Failures That Taught Me to Build a Better Business

In 2021, I'll be celebrating 10 years since my first lines of code went live on the web, so what better way to mark the occasion than a little retrospective of my failed projects?

The entrepreneurial ecosystem keeps hammering home (a bit too much) that you have to "bounce back" from your failures, that it's startup culture, and that you need to fail enough before succeeding. As usual, this is advice that's much easier to give than to put into practice yourself... But here are my 5 biggest failures in the digital space over the past 10 years:

2011 - Robotschool

or the misconception that writing content is quick and easy.

Back in high school, I spent a good chunk of my free time on Le Site Du Zéro (now OpenClassrooms) discovering web development tutorials that were so well written (thanks @MathieuNebra). The problem was that my curiosity didn't stop at programming: electronics, mechanics, music... I was thirsty to learn everything. Unfortunately, I couldn't find any platform that gathered tutorials on all these topics and was as well written as the ones on SdZ, so I decided to try my luck. At the end of high school, I launched Robotschool, my first web platform.

After several months of development (I was teaching myself to code at the time) and a very approximate design (I was also teaching myself Photoshop), I finally got to the stage of writing the first tutorials. I had a plan: find several tutorials on the same subject, follow them and absorb them, synthesize them into clear and precise content, and finally format the final piece. The methodology seemed simple to me but... what a mistake!

After spending several long weeks writing 3 tutorials of rather average quality, I decided to stop everything, realizing how naive I had been about writing large amounts of content on the internet.

Writing content is a job, or at least a passion you need to have to spend entire hours (sometimes days) keeping up a blog, a newsletter, or tutorials. Creating content can quickly become demotivating, especially if it becomes a daily task, and this is an issue we still see today with some companies that set up a blog and abandon it within a few weeks!

Bonus: my circle at the time had decided, to support me, to click on the site's ads in order to generate a bit more revenue, something Google flagged as fraud and which got my AdSense account permanently banned!

In fact, this technique resurfaced not long ago as a way to blackmail companies that rely on AdSense as a revenue source! (Article)

2012 - RPG Facts

or the false good idea of user-generated content.

If massively creating content wasn't for me, maybe motivating a community to produce it autonomously was the solution? It was from this line of thinking, and from my passion at the time for tabletop role-playing games, that the RPG Facts platform was born.

In 2012, the site viedemerde.fr counted more than 4 million visitors per month, the community was very active, it created the content and moderated new posts, the platform ran autonomously, the dream.

What I didn't know at the time was that the famous VDM had started out as a simple blog between friends, and that the platform as I knew it had come much later, which could have saved me several weeks of development.

For my part, I launched the site with all its features (registration, publishing, moderation, voting...) topped off with a few stories already written, before realizing that even though my target users loved sharing their stories, the stories were very long and role-players preferred to share them out loud rather than spend hours putting them in writing.

In short, I had a community ready to "consume" content, but too little motivated to "create" that same content — this is what I call a "desynchronized" community.

Bonus: This was the second time I lost weeks of work building a full platform without having tested the market. It was only a bit later that I discovered the value of creating an MVP (Minimum Viable Product) through the best-seller "The Lean Startup" by Eric Ries.

2014 - Eurekube

or when customers exist but aren't ready to pay.

The idea of solving problems by harnessing the collective thinking power of an entire community has always fascinated me, especially since discovering the reCAPTCHA initiative: a bot-blocking system on the internet that also happened to digitize the New York Times archives.

This is what's called crowdsourcing.

During my master's at the University of Nantes, and after taking several months of courses on entrepreneurship, I launched "Eurekube," a new project that allowed small businesses to find solutions to their problems thanks to suggestions from the community; the idea being, eventually, to provide extra income to the most active members who proposed the best solutions.

Why target small businesses? Because a few competitors were already positioned on large accounts, and the process of writing solutions for big companies was too tedious to turn into a simple, dynamic, and appealing application.

The concept started with an MVP in the form of a Facebook page bringing together a community of 400 students, and I had finally managed to create a concept that generated revenue — but what revenue?

Among my first 3 clients: a bar, a pizza truck (yes, really) and a convention for amateur video creators, total revenue amounted to €30 and 3 free tickets to attend the convention. After several weeks of relentless prospecting, the conclusion was clear: the community was interested in this kind of project, businesses were also interested in gaining a bit of local visibility, but they were absolutely not ready to pay for this type of service, and certainly not enough to sustain a platform like this one.

2015 - Peetch

or the art of knowing when to say stop.

There are adventures so beautiful and so complicated at the same time that it's hard to talk about them as a mistake one would have liked to avoid, and this project is one of them.

Peetch came to life in March 2015 following a Startup Weekend where our team of seven people and our collaborative story-creation app won first prize among all the projects.

For 3 years, the project gave us incredible experiences, such as meeting Sébastien Forest (founder of Allo Resto), our business angel and mentor, for whom I still have a very particular admiration today; our participation in the European Innovation Academy, which brings together 600 entrepreneurs from around the world; joining Centech (one of the best incubators in the world); as well as the chance to go pitch our project at StartupFest in Montreal.

All of this was made possible by months of relentless work from the whole team, and the product we had built had truly won over our users, but from the very first months of the project's life, a problem was lurking in the shadows and would eventually blow the project apart into a thousand pieces.

The attitude of one of the members of the founding team, already responsible for several crises that had put the project's survival at risk, had become completely incompatible with the rest of the team, creating conflicts, unilateral decisions, and a loss of motivation for most of us.

It's hard to say today whether removing that person from the project early on would have let us go further in this adventure, or even let us live such experiences, but I now know that if one member of the team isn't aligned with the rest, a decision needs to be made as quickly as possible, or you risk watching the project slowly die out.

Worth noting: more than 20% of startups fail because of a team problem!

2018 - Kalico

or the story of numbers that deceive.

In 2016, as Facebook Messenger was evolving, I discovered a new buzzword: chatbots. Passionate about everything related to robots and automation, I threw myself into this new experiment and developed, within a few days, the outline of a personal shopping assistant for high-tech products.

Held back by other, more important projects, I set my assistant aside for many months, while noticing that a few curious users kept asking it for advice.

In 2018, finding a bit of time to experiment again, I picked the project back up from scratch and gave it a new identity, new products (smartphones), and above all a new platform: a dedicated web application. The beta was ready in two weeks flat, and I decided to throw it to the wolves of the community always on the lookout for new apps: ProductHunt.

Within a week, Kalico reached its first 500 users from 20 different countries, with a completion rate of the user journey close to 70%.

All the conditions seemed to be in place for the app to succeed, and yet after the first week, the number of daily users started to drop, never stopping. But why?

Our mistake was basing the rest of the project on the numbers and feedback from the launch; all our subsequent decisions were skewed and we followed the wrong leads. We thought we were aligned with our market and our users, when in fact we had simply looked at the numbers that most reinforced our own ideas.

The numbers were very good, but they were very good thanks to the visibility Product Hunt had given the product, and even though the completion rate was very satisfying, it was so among a category of users used to testing apps, whereas, on the contrary, Kalico's goal was to help people who struggle with technology choose a smartphone.

Choose your KPIs carefully, but above all choose carefully how you're going to analyze them and remember to put them back into context. Traction shouldn't be measured right when an article comes out about the product — it should be measured when only the users representative of your actual community are talking about it.

---

Of course, a single mistake is rarely the sole cause of a project's failure — rather, it's the accumulation of these mistakes that drives it straight into a wall! I chose to tell you here about the mistakes I've come to understand and now know how to avoid, so that you can steer clear of repeating them.

If you'd like to discuss technical and strategic support for your project, drop me an email at nicolas@brondin.com and we'll make sure you avoid as many mistakes like these — and many others — as possible.