All About Pivoting Your Startup
Startup pivots, Greece mints another unicorn, InAccel joins Intel, cybersecurity, jobs, events, and more
Startup Pirate is a writing about technology, entrepreneurship, and startups every two weeks. Made in Greece. If you haven’t subscribed, join 5,825 readers by clicking below:
All About Pivoting Your Startup
Pivots are common in startups. On the one hand, you have Slack, which had raised money, had 100 employees, and developed a video game called Glitch before shutting it down to incubate the chat features between Glitch players and staff. This turned into Slack, the communication platform we know today. On the other side, there are teams at the earliest stages of company building, prelaunch or very near after-launch, where changing ideas generally feels more lightweight. In all cases, pivoting is one of the most important decisions a founder can make.
A few months into building Moonshot, having graduated from Y Combinator with funding and plenty of users, Dimitris Nikolaou (a second-time guest on this newsletter) and his co-founder decided to shut things down and focus on an entirely different problem and industry. Their new venture, Wondercraft, is now used by over 30,000 creators and hundreds of enterprises worldwide. This is their story pivoting from an angel investing platform for young athletes to an audio studio powered by AI.
As Paul Graham said, “almost all founders learn brutal lessons during the first year […] The most common one is that customers don’t want what you’re making”.
Before we get to it, here’s an audio version of our conversation powered by Wondecraft’s AI engine.
Dimitri, thanks for coming over. We don't often hear stories behind startup pivots, less so for teams that went the extra mile, had users, and raised funding. What made you realise that something was not really working with Moonshot?
DN: Alex, thank you for having me. Always a pleasure to be part of your efforts. I'm not an expert in pivots, by all means. Pivots are generally hard and circumstantial. Hence, this trade-off between exploring and exploiting the particular thing you are working on can only be resolved personally. So, take what I will suggest today with a pinch of salt.
Moonshot was a marketplace where we enabled anyone to support young athletes; you can think of it as an angel investing platform for rising athletes. My co-founder Youssef and I were involved in competitive sports growing up and were aware of the young generation's troubles in pursuing their dreams in sports. It was deeply personal, and we also believed there was a decent market opportunity. However, a few months into our journey with Moonshot, plenty of our initial assumptions proved wrong.
Even though we still think there is space for such a business to exist, this is likely not high-growth, venture-backed. That's often the case with human marketplaces, where you need to convince users to join, onboard, and serve them one by one to a large extent, while margins don't justify such an effort. For instance, I had to travel to San Francisco to convince just one athlete to join the platform. Growth would always be slow, and we didn't have an existing network of athletes to kickstart things. Most importantly, it felt like this would never become a software-first business, which is what we both set up to do when we left Palantir.
It's harder if you have little traction vs. 0 traction, right? A dangerous (and common) scenario is a startup with just enough traction that they aren't willing to change ideas/markets and not enough traction to know they have product market fit. And then you also hear about the anecdotes of people who kept doing what they were doing, and it didn't work, and then five years later, it was great. How was that debate with your co-founder and investors?
DN: It was in March 2023 when we sat down with Youssef to evaluate progress. It's easy to get lost in the weeds of building daily and never assess your long-term plan. You have a strategy and try to execute it, but it is essential to sometimes check against the overall goals. At that point, it was clear Moonshot would not turn into the success we dreamed of. Slow growth was the peak of the iceberg, but digging deeper, we realised market dynamics were not in our favour, and we were building towards a local maximum.
There was never a moment when we doubted ourselves as a team. We wanted to continue together, and this wasn't even a topic of discussion. It was a matter of finding something else to work on. Also never thought of going back to corporate life. We just understood that we were not facing the right direction and had to give ourselves a better opportunity to succeed.
We went through YC in the summer of 2022, and after graduating, we raised funding from Will Ventures, Y Combinator, and angels. Therefore, we had to return all that capital or convince them we were onto something new, and they kept supporting us. We hadn't spent too much money on Moonshot, and I recommend not doing so until you find product market fit. Only then start growing at scale. We had been transparent and shared progress with our investors since day one, so as soon as we decided to shut down Moonshot and pursue Wondercraft, we discussed possible options. "Bad news, Moonshot is dead. Good news, there's something bigger and better coming." Showing early traction with Wondercraft definitely helped get them onboard and continue the ride with us.
When did Wondercraft come into the picture, and what early signs pushed you to go all-in? I remember a bunch of viral posts online from early users during the first few months.
DN: There was so much PTSD for being slow with Moonshot's progress. That, along with the need to present possible options to our investors, led us to act fast as soon as we decided to shut down Moonshot. We brought another friend, Julian, now a founding engineer at Wondercraft, into the conversation. Julian is a fantastic engineer and product-led growth oriented, and we knew our next venture would be on the opposite spectrum of what Moonshot was — self-serve and software-first.
We went back to the drawing board, discussing different ideas based on our interests and new technologies we could leverage. The "Why now?" had to be clear. Sounds simplistic, but at the end of the day, ideas don't just fall on your head from nowhere. Wondercraft was at the top of the shortlisted ideas. Back then, we were thinking of leveraging text-to-speech and LLMs to create podcasts. We ran an ideation phase in 24 hours. Essentially, we woke up on a Tuesday building Moonshot. Then, by Wednesday night, we had decided to pivot to Wondercraft.
We hacked an MVP together in 3 days, and I remember messaging you and others to get early feedback. We had over 100 subscribers and $3k MRR in the first three days, which allowed us to build confidence that this could turn into a high-growth business. Obviously, there is no way you can build a vision for a new business in just seven days, but it did set our minds that it's worth pursuing further. It took us about six months iterating on that MVP.
There was great traction on Twitter, viral posts on Hacker News and Reddit from podcast creators using our product, "Product of the Day" spot on Product Hunt, and more. We even created our own AI-generated podcasts. For instance, Hacker News Recap, a daily summary of the most important articles on Hacker News, and PG Essays, Paul Graham's essays in audio format. We could feel a clear contrast between building something when there's not so much market pull versus when the market is pulling you.
We have been discussing Wondercraft for a while but realised you haven't really explained what it is.
DN: Our goal is to lower the barrier for creators to produce audio content. Typically, you need a microphone, a proper recording setting, and then spend time editing. The production bar is high. With Wondercraft, any creator, whether it's b2c or b2b, can write content and turn this into studio-quality audio with their own tone of voice or predefined ones, whether it's podcast, audiobook, meditation, ads, company comms, and a lot more, and effortlessly translate for a global audience. We have over 30,000 b2c users and hundreds of enterprises, including podcast networks, audiobook platforms, etc.
Behind the scenes, we have different generative AI tools, the most important of which are text-to-speech conversions. The AI community has made fantastic progress over the past few years, enabling all sorts of innovative use cases to be built on top of new models. In fact, a key piece of our stack is text-to-speech created by Eleven Labs, a pioneer in AI voice synthesis founded by Mati Staniszewski, a close friend and ex-colleague at Palantir. To an extent, watching their company take off in early 2023 (a unicorn in two years) opened my eyes to what exponential growth means and the leverage these models can give you.
There's also a lot of OpenAI for script generation, as we have a script assistant that helps users craft content before it can be voiced. In addition, there is an AI dubbing engine, a post-production process in which additional or supplementary recordings (doubles) are "mixed" with original production sound to create the final audio used for language translations. These are a few different components of our technology, including models that make sure emotions are incorporated and translations are accurate.
A few weeks ago, you announced a $3m Seed to build the “Canva of Audio”. What are the next steps?
DN: We are just starting this journey to help creators give voice to their stories. We see a lot of demand for what we are building, and customers come up with new use cases almost weekly. The current focus is on iterating the platform while keeping a product-led growth mentality. It's eight of us now, and we have built a team combining industry experience with engineering expertise. Our third co-founder, who joined after the pivot, Oskar Serrander, is a former COO at Acast and executive at Spotify and iHeartmedia, so he really brings insights from the audio creator space.
Before we wrap things up, if you could rate from 1 to 10: early market feedback, founder/market fit, and broader market opportunity, how would do for Moonshot and Wondercraft?
DN: Early market feedback? Moonshot was a 2/10. Not much was going on, and we were trying to force things. It's vital to differentiate between people saying they're excited about what you build and buying it. We had a ton of the former with Moonshot, and as a first-time founder and young kid, you get excited, but if no one's taking their wallet out, then the early market feedback is not there. Wondercraft feels in the opposite direction. Blockers from our side, such as integrations, keep people from buying it, so there's clearly market pull. I'd say it's an 8/10.
For founder/market fit, Moonshot was a 3/10. It was a passion of ours. At the same time, we are engineers and best at building a business that grows through the success of the software. And that's not what Moonshot was. So, even though we were deeply interested in the concept, what you're interested in doesn't necessarily reflect what you're good at. We had to learn this the hard way. On the other hand, we had no idea how to control decibels or crop audio before Wondercraft. Neither Youssef nor I come from the media, audio, or text-to-speech spaces. Nevertheless, we're now building SaaS, iterating through customer feedback and removing friction points to build a product that users will love. With the addition of Oscar as a third co-founder, I believe we have now nailed founder/market fit, so probably a 10/10.
Market opportunity? This goes back to answering "Why now?" and what market or technology waves you are riding. For Moonshot, looking back, nothing was pressing at that moment. There was no good answer to "Why now?". Perhaps a 2/10. For Wondercraft, we are clearly riding the AI boom. New use cases are discovered daily, models’ performance improves, and incredible talent joins the space. We just don't have the opportunity to address everything, mainly because we want to focus on what we can deliver. Moreover, companies in the AI speech-to-text space, such as Eleven Labs, are growing exponentially, so using that as a proxy, it must be a 10/10.
We covered a lot of ground, but if I could ask one more thing, what would be the main takeaway from your journey?
DN: If I had to pick one, it's patiently building a great network of people I could build stuff together. Without knowing it, this really prepared the ground for starting a company. I'm just connecting the dots here, but throughout my time at Imperial, Palantir, and then Y Combinator, I was essentially placing myself in environments with people I really believed in. Good universities are a great example, not so much for the studies but for the network you can build and the opportunities they give you further on. So put yourself in a position you're proud of at every moment, which will always lead to the next success, wherever it might be.
Loved it, Dimitri; thank you so much!
DN: Thanks Alex.
Jobs
Check out job openings here from startups hiring in Greece.
News
Greece mints another unicorn! Founded in 1999 by an AUTH team, R&D in Thessaloniki, and industry leaders across aerospace, automotive, healthcare, and more as customers, BETA CAE Systems was acquired for $1.24 billion. (link)
Marathon-backed InAccel joined Intel. Chris, Yiannis, and Elias built world-class FPGA technology for computationally intensive applications. They partnered with leading industry players and are now part of Intel’s FPGA division. (link)
Insurtech Hellas Direct raised €30m led by ETF Partners for climate-focused insurance solutions. (link)
Olympia Group acquired 53% of IT software and services provider Entersoft. (link)
Language model for healthcare ScienceIO was acquired by Veradigm for $140m. (link)
Laser bioprinter developer Phos Print raised an angel round by Google Maps founder Lars Rasmussen. (link)
Resources
Haris Pylarinos, founder & CEO of Hack The Box, on cybersecurity, hacking training, building a startup in the space, and more. (link)
The future of battery technology with Nick Kateris, Postdoctoral Researcher at Stanford University. (link)
Scaling blockchain-based identity with Kostas Chalkias, co-founder & Chief Cryptographer at Mysten Labs. (link)
The Murakami approach to beta customer feedback by Anthoula Poniraki, Senior Product Manager at Workable. (link)
Concise unit tests in Python by Maximos Nikiforakis, Software Engineer at Skroutz. (link)
Intentional work framework and aligning teams towards goal setting with Konstantinos Mouzakis, co-CEO at Balena. (link)
Costas Mantziaris, co-founder & Managing Partner at REBORRN, on company culture, entrepreneurship, and more. (link)
Events
Talent Days by CollegeLink on Mar 9 - 10
Athens #4 by AI and Beers on Mar 13
Athens Python Meetup on Mar 13
Kubernetes Athens vol23 on Mar 13
1st Athens eCommerce MeetUp on Mar 14
Mastering Prompt Engineering by Thessaloniki .NET on Mar 15
Data Storytelling: Key to Insightful Reporting by Athens Power BI on Mar 18
M&A Episode 1: Founder's take by Tech Finance Network on Mar 20
Devoxx Greece 2024 on Apr 18 - 20
That’s all for this week. Tap the heart ❤️ below if you liked this piece — it helps me understand which themes you like best and what I should do more.
Find me on LinkedIn or Twitter.
Thanks for reading,
Alex
Loved reading through this. Great read.