CTO and Co-Founder of BubblyDoo, an e-commerce startup creating personalized products for children.
This is my experience being CTO of BubblyDoo for 7 years, from being an entrepreneurial student at 21 to leaving as a CTO at 28 years old. Opinions are my own.
In 2019 I co-founded BubblyDoo. I was the CTO, and for 7 years I created technology that allowed us to sell hundreds of thousands of personalized children's books around the world. There's a lot to talk about. We did many things and we had to figure out most of it ourselves, from product creation to management to marketing to raising funding and everything else you do in a startup.
In this post I tell the story of BubblyDoo from my perspective. I focus on our milestones, pivots, our business and my story and learnings as a founder.
I wrote a separate post about
, and I'm still writing a post about the technology we created.
What is BubblyDoo?
We create personalized books for children. We sell them online on our website, print them and deliver them. We deliver in 25+ countries, we have 10+ printing partners around the world, and we did €5 million in revenue in 2025. Our first unique selling point is that we use licenses, so your kid can be in the story with their favorite character. The other one is that our personalization is very extensive.
I'd recommend taking a look for yourself
.
The early days
2019: Getting together and the idea
When I was on exchange in Singapore in 2018, I received a LinkedIn message from a guy I vaguely knew from the neighborhood. He wanted to talk about a startup he was working on. After meeting again on a ski holiday in 2019, we finally planned a call together with his friend. That's how I got to know Toine and Grégory.
They had an idea but they needed someone with programming skills to bring it to life. What they pitched sounded challenging and interesting, and I decided to join.
The idea started out simple: we wanted to create personalized books to make reading more attractive for children. We wanted to put the child in a story in a world with characters we'd invent, and then create a series of books from those characters, each one about a lesson or a moral. The first story was about eating healthily.
Our first book, The Color Monster
After a while we also found Laura. She was a graphic designer and illustrator, and now we had everything we needed to create our product. From June 2019, we were officially the 4 founders of Bubbly-Doo.
From the technical side, there were quite some things to do. Most importantly, we needed a way to illustrate a book and make it personalizable. But we also needed a website, a dashboard for handling our orders and a printing facility that would print these books.
I should have written my master thesis in the first half of 2019, but I was too excited by what we were building. We were all in our masters programs at that time. After half a year, our minimum viable product was done.
We set a deadline: launch on the 20th of October 2019 (of course this was moved a few times, but that was the final decision). I had class that week, I barely slept and kept coding during my classes, but it worked out and we launched.
BubblyDoo website at launch
We only underestimated a bit how people would actually get to know our product, and so we only sold to friends and family the first day. The week after we started contacting influencers and we launched some Google and Facebook ads.
The marketing wouldn't be solved for a while: we only sold about 750 books in Flanders and in the Netherlands in 2019. Definitely not enough to start paying ourselves.
Being naive
If someone with e-commerce experience would have looked at our plans, they probably wouldn't have loved it. New e-commerce businesses require high ad spend and great margins, and we didn't have that kind of capital and our margins were pretty terrible. But maybe it's better that we didn't know that: we might not even have started. Looking back, it's great we just had the self-confidence to focus on the product and the strategy, get it out there and improve from there.
2020: More books and Germany
In 2020 I went on exchange to Shanghai. 2 weeks after arriving, the pandemic broke out. Most of my class decided to stay in China, but our classes were online from that moment. In the end it was a good decision to stay, because the situation in Shanghai in the first half of 2020 was arguably better than the situation in Belgium.
I spent a lot of time coding from my bed and from the Baker & Spice deli on Changshou road, and we managed to work in different timezones for a while. I made the website ready for localization and to have more than one product. And in May 2020 we sold our first book in Germany.
To launch multiple products I needed to change quite some things, and in September we launched the sequel to our first book, this time focused on loneliness, a relevant topic especially during the pandemic.
Soon after other books followed by freelance illustrators: one about making friends, one about Sinterklaas (a regional holiday) and one for Christmas.
To try to get ourselves out there we also launched the free downloadable "Covid Hero Poster", where the personalized kid follows 4 hygiene rules, like keeping a distance of 1 crocodile from other people, and we managed to get
with that.
In November we were selected "Student Startup of the Year" by startups.be. We got a few months in startup hub The Beacon for free, so we moved to our first proper office.
By the end of the year we ended up selling about 7000 books, of course still at a net loss. Better but definitely not yet on track to become a successful company.
After surviving for 1 year on a bullet loan that we got as part of being accepted into Birdhouse, an accelerator, we also closed our first funding round and got our first angel investors.
With some agency help, we also rebranded. (They luckily agreed we should drop the hyphen and become BubblyDoo.)
Looking for product market fit
2021: Licensing, board games and realizing what kind of company we are
Instead of only focusing on books, we wanted to provide other means of creating family fun. That's why we started working on board games as well. This was difficult, as our personalization software was still quite primitive and the board game we were making could have multiple children and parent characters, each with tons of options. Because we didn't have a good foundation to build on, we quickly got stuck in development hell.
It was also becoming clear that our current business model wasn't good enough. We could sell our books at a positive gross margin without a problem, but when taking marketing costs into account, margins turned deeply negative. Our customer acquisition cost was just too high. There is too much competition on Facebook ads and our product just didn't resonate enough with people to outbid other advertisers. At the same time, other types of advertising (influencer, content, OOH, etc.) didn't really work out for us.
We could work on the gross margin, but the customer acquisition was definitely the main problem. How can we expect to sell a novel concept like a personalized book, online, for a premium price, on a website no one has ever heard of, using a character no one knows? We needed a new way to convince people to buy our books.
The answer came in 2021: licensing. We closed a deal with Studio 100, a big Belgian entertainment company, to use their characters in our books. The value proposition became much simpler: buy a book with characters you and your kid already know, but with your kid added to the adventure. An amazing gift. We could focus on creating the product, while the marketing got easier.
This worked and licensing became our unique selling point. We started closing more licensing deals.
We launched our board game, but it didn't immediately attract enough customers and the margins were much worse than those for books. We decided not to give up and to launch more games, mostly under license, and also card games.
Board game and licensed books
We also looked into retail: putting screens in toys or books stores where customers can order books directly from us and pay at the cashier. This didn't get far due to low interest from retailers.
I also hired our first full-time developer to work on launching products, still a job for a programmer back then.
Platform or not?
A question we were struggling with at that point was how our personalization software should be positioned. Should we open it up for external people to work with, so that BubblyDoo could become some kind of marketplace for personalized products? Of course, that would require a lot more engineering than for a purely internal tool.
In the end, we considered that non-viable, and we kept everything internal. We didn't have the capacity yet to even build a good tool for ourselves, and we doubted there would be a market for such a niche product.
2022: Other kinds of products and betting on A/B tests
On the business model side, we needed better conversion rates. We decided that the best way to improve those was A/B testing. I'm still not convinced it is the best idea to use A/B testing for optimizing a website that is obviously not mature yet, but more about that later.
We hired two more developers, one to focus on frontend and one for backend, a marketing creative and an illustrator. We also hired some students and freelancers.
I was kind of responsible for operations, customer support and the fulfillment of orders, so we created a better dashboard to manage that.
We decided to launch many new products in many new countries and "see what sticks". Our go-to-market speed seemed fast enough to use this strategy without too much upfront research.
By the end of 2022 we were live in a lot of countries. The US, the UK, Italy and France were added.
That Q4 was tough, our teams were all crunching and working long days and sometimes nights to get our products ready for Black Friday and the holiday season. The games were still being built on shaky foundations, each time going through development hell.
At the same time, we had so many printing facilities in so many countries all using custom integrations that there were bound to be issues as well. The whole team had to switch to customer support because some products were seriously delayed due to bad communication.
The licensing pivot paid off. We launched our first PAW Patrol book on the 25th of November, right on time for the holidays. It became our bestseller overnight. Without that deal, we probably wouldn't have survived the following year.
The hard work and the licensing deals made us cross an important milestone: 1 million euros in revenue. The margins were not positive yet, but we were on the way up.
Complementary founders
The four of us were really complementary. I think we were incredibly lucky that we found each other. We disagreed a lot, but never about the fundamentals. We each had our own area: strategy & business, marketing, product, technology. I don't know how we would have done it without each other. We were in it together, and I am very grateful for the 7 years of building BubblyDoo.
About A/B testing
We bet very big on A/B testing. I think this was a mistake.
Our strategy was increasing our conversion rate by continuous improvement of our website through A/B testing. If we would find a 1% gain every month, it would compound and get very large over time. This means our customer acquisition cost, our most important cost, would drop significantly.
The problem was that we also thought A/B testing can replace understanding of users and discovery of friction points. There was no vision behind our what our website should be like, only "the one that converts best". For me, this was disorienting.
We assumed that our inexperienced designers could identify improvements out of thin air, maybe by sometimes looking at screen recordings. We spent too much time bending over backwards in code to measure obvious things, like the effect of improving speed, instead of just improving it.
I think we were too distracted by reading things like "Google changes button color, gains $200M in revenue", while forgetting Google is building on a huge foundation.
The other problem was that we just expected too much from it. We got stressed when our conversion rate didn't grow as predicted.
In my opinion, A/B testing can work, but as a part of the toolkit to improve your website and product, next to vision, discovery, and user interviews.
Grow or (almost) die
We needed to improve our unit economics and grow the revenue to cover our fixed expenses. This turned out to be harder than expected.
2023: More types of products
We went on a weekend team trip to a house in France which was great. Our team had a bunch of different people, each one with different backgrounds and beliefs, but all passionate people. Our hiring strategy was poorly defined, because we didn't have any experience, and because finding a developer in Belgium takes more effort and money than finding other profiles. We didn't have big budgets, so we couldn't hire experienced developers.
The PAW Patrol books kept selling very well, and France and the US were our biggest markets. The games on the other hand didn't sell well at all.
Our customer acquisition was still too expensive and we were growing too fast to focus on recurring revenue. I think you can say that we partly found product market fit with our PAW Patrol book. If we just had a few more of those, we would be profitable in no time.
Next to product development, we also focused on global expansion. We launched in a lot of countries across Europe.
We also decided to try personalizing other products: t-shirts, drinking bottles, blankets, mugs, etc. Simple designs, no translations needed, easy to try out. Over 2 years, we launched about 100 of them, trying to find that one product that would actually get profitable. They didn't reach enough scale to have a large effect on our business.
Our licenses worked well in more countries, and we tripled: 3 million in revenue.
2024: Trying and failing to grow
In 2024 we set out to triple again and achieve better margins, which would push us well into profitability. More products, more countries, more licenses and a better website was going to get us there.
We hired more people, created more products, designed more ads, made more website improvements and created a lot of dashboards to analyze everything.
It didn't work and we didn't grow. We tried pretty much everything, but we ran into a wall. Our licenses worked but they didn't grow in popularity. Our new licenses didn't do what PAW Patrol did. Our t-shirts and sweaters were not attractive enough to overcome the lower margins. Some countries started performing worse than before. We had theories about what was going on and we tried to counter it, but there was not so much we could do.
We launched A/B tests left and right and analyzed our website over and over again, but we didn't find the problems.
What we did well is that we started to focus on more events than Christmas: our father's day book worked well.
Frustrations in the team
The team is a really important part of any company. We spent lots of time on hiring. We had a fun team, an open culture and an enormous Slack emoji library. They were great people, but our inexperience in management unfortunately created problems.
We hired a lot of people. We always tried to make it clear that we expected more from them than other companies would. But after a while, when people didn't see their expectations come true, it turned into frustrations for some.
We gave stock options to all employees to align incentives. But our compensation was unstructured, and people didn't understand why salaries were different.
The four founders also had different beliefs on how to run a team. I believed more that "if people enjoy what they're doing they will do it well", while others were more demanding and set ambitious deadlines. It didn't help that most people had multiple managers.
Our employees had very different jobs. Some did very manual stuff that would get boring, while others had to keep concentrated throughout their day. But everyone was together in one office, and this led to tensions.
Each team member was also very different from each other. No one had a similar background and everyone had different views on the world, on the role of companies in a society, and on how managers in a company should be. There was always a lot to discuss, but it probably distracted from what we really needed to do: grow the company, because if not, problems were inevitable.
We always heard culture was important, but then we realized how important it actually is.
This would have been prevented if we would have had more management experience. Now we all realize that we should have communicated better that we can be friendly to one another, but that in the end, the employee is a part of a team that should perform well, and if that's no longer possible, they should not be part of the team. That doesn't depend on their own work but also their influence on the work of other people. We called it being a "sports team" instead of a "family", where we hire and fire based on performance, regardless of how nice people are or how hard they are trying.
And we definitely did try to give performance reviews and raises accordingly, but it was very difficult to do this objectively. When programming something totally new, it is normal things don't work out and are delayed. It was hard to judge if people tried their best. We hired junior programmers in the end. But I gave the benefit of the doubt too many times.
At the same time I was also too lenient in accepting and setting unrealistic deadlines and projects. If every project gets delayed time after time, employees also won't take deadlines seriously anymore.
As the growth didn't come, management got stressed and the pressure turned into frustrations for the team. A lot of the blame for the stress turned to management, who the employees started to distrust for poor planning or being unrealistic with regards to targets. From our side, we felt like we were doing everything to keep the company alive.
By the end of the year, all these frustrations made a lot of people dislike their jobs and made our company much less efficient.
I listened to these frustrations for hours each month. I wanted to do something about it, but I think it was a matter of incompatible beliefs, styles of management and personalities. It's unrealistic to expect someone will change. The family feeling kept the team together, otherwise many would have quit.
First half of 2025: Downsizing
We didn't raise enough capital to sustain a team of 13 people into the next year. In April, while trying to raise more funding to survive, we reduced our team to 6 people. This was very tough to do, because we had never fired anyone until then, and we kind of felt like a family.
In some kind of weird coincidence the four of us were also selected for
that month. A nice recognition, but bitter for our team.
Second half of 2025 & 2026: Back to basics
We decided to go back to the basics and try to grow with a smaller team, but also not to make the same mistakes again when it comes to managing the team. We were going to be more direct, and feel less like a family and more like a sports team. We would be more strict to culture, and we wouldn't tolerate disrespect or cynicism anymore.
We focused on launching books and on nothing else.
I didn't want to go back to a team where I was the only experienced developer (more about that
), so I decided not to hire junior developers anymore for now.
Laura and I also decided to fully split the developers and product development team, in order to prevent the multiple manager problem and improve specialization. Instead of developers working on products, there would be one supporting developer and then a team of less technical people who we called product implementers, who were hired specifically for that role and not "a bit of everything" like before.
This was a really good choice. The supporting developer could focus on doing the more challenging parts of product development, and the product implementers got satisfaction from being a smooth and efficient team that would launch more products in more languages than we ever did.
I focused on improving the tools everyone needed, so that people could work independently without running into obscure bugs and needing a programmer's help. Next to that, we also structured the process very well in Notion, striking the right balance of details and overhead. In a few months, the product development team became very reliable and with a high output, and I didn't need to be involved anymore.
We introduced a project manager to take over some of Laura's management tasks. There were so many products, languages and translators that this was really needed. This worked very well.
On the technical side, the goal was simple: make the product team independent by building reliable tools. We refocused efforts and used agents to optimize or even recreate parts of our technology we hadn't touched in years.
Even though this fresh start was forced, our team dynamic was healthier and the company got much more efficient than before.
The business recovered thanks to new licensing deals, and we closed the year with a record revenue of €5 million.
Being a founder means everything is your problem
I loved being a founder, but when you divide the competencies between your team, everything that doesn't fit will need to go to someone. GDPR and the cookie banner? Doing customer support? Taking care of the sales invoices in accounting? Creating hundreds of graphs and charts of sales or conversion data? Hiring people although you don't know anything about it? Writing an IT strategy for the funding round that is preferably full of buzzwords that you personally hate but that no one else knows enough about to actually do it? Those are all things I had to figure out completely on my own even though I would have preferred it was someone else's job. I definitely learned better what I don't like, but I also learned hundreds of things I wouldn't have known in any normal job.
Deciding to leave
Before the downsizing, I already decided for myself that I would leave at some point. Being the CTO of a startup is great, and there are new challenges every day. But I didn't love all of them. I wanted a new perspective. I didn't want to solve every problem anymore. I wanted to work in a team where I'm the least experienced developer and where I can learn from them, which doesn't make financial sense for a startup. We decided we would look for a new CTO, and that I would leave in June 2026.
We found a new CTO in September 2025 and he learned from me for the rest of the time being there.
Being default dead
When you're in a company that is not profitable yet, you're bankrupt unless you raise money or grow fast. By default, you're dead. Every year brings a huge amount of pressure when the runway clock starts ticking too close to the end. You cannot think long term because you need to increase profitability right now. Expenses need to be delayed until after the funding round. You're in a constant battle with yourself to figure out if you're doing the most important thing possible or if you're wasting time on something you find interesting and that might be useful. This constant inability to plan ahead is tough.
Realize what kind of business you are in
I think we saw ourselves for too long as a Silicon Valley software startup instead of as an e-commerce business. It's definitely true that we're not just an e-commerce business, as a big part of our edge and how we operate relies heavily on technology. But we're still an e-commerce company, and that means marketing, sales, and physical products are more important than the technology. The technology is a means to an end, not a product on itself. It took me a few years to realize that. It's an important realization. It means you can't hire a team of expensive developers to create features that will be worth their salary, because such software-only features don't exist.
It also means I started focusing less on moonshot projects. I began optimizing processes in the company with technology. I started relying more on external services rather than building everything myself without time to maintain it. It was fun to think software could do anything in our company, but it was good to put it in its right place.
Finding balances
In management and business, everyone likes a simple solution. You don't do X, you will go bankrupt. If you do Y, your team will get better.
What I've realized is that all of those stories are relative. If you're a hierarchical company and you don't understand your employees, you might need to make the structure more flat. If your employees get too familiar, you might want to get more hierarchical. That's what management is about for me: a balancing act where you constantly have to recalibrate based on your reading of people, your business, your self-reflection. Not an objective list of rules you have to apply.
I found this especially well expressed on a wall of the Lego House in Billund, of all places:
The 11 Paradoxes of Management
- To be able to establish a close relationship with your employees — and to keep proper distance.
- To take the lead — and to recede into the background.
- To show the employee confidence — and to be aware of their doings.
- To be tolerant — and to know how you want things done.
- To be concerned about your own field of responsibility — and at the same time to be loyal to the overall goals of the company.
- To plan your work day carefully — and to be flexible to your planning.
- To express your opinion — and to be diplomatic.
- To be visionary — and to keep both feet firmly on the ground.
- To aim at consensus — and to be able to make a decision.
- To be dynamic — but also thoughtful.
- To be self-confident — and humble.
Thanks to all
I'm very grateful for my 7 years here. I believe I'm incredibly lucky this opportunity came my way. And while we may have had quite some disagreements, I'm also grateful for learning so much from working with the people that crossed our paths.
Thanks to all the developers that ever worked in my team: Oliver,
,
,
, Krystian, Jérémy, Leo.
To all the people that joined BubblyDoo: Lara, Emma, Iryna, Renato, Fer, Cate, Jennifer, Michelle, Pedro, Dora, Venita, Raitan, Afiful, Shabrina, Amelinda, Tabas.
To the students and interns in my team: William, Salah Eddine, Min Jie, Senne, Jörg, Jeremie, Dirk, Kyllian.
And to my co-founders for the adventure and the support: Toine, Grégory and Laura.
