Juho Risku is the co-founder and partner at Butterfly Ventures, a seed entry deep tech VC with offices across the Nordics.
How did Butterfly Ventures get started?
We were putting together an accelerator around 2011-2012. I had some startup experience, and there was a government financing program to support that accelerator program. At the same time, Nokia was collapsing heavily, and lots of talent was coming out, particularly in my hometown city of Oulu. They had a big Nokia site, and the City of Oulu wanted to put out a bid to manage a small local regional fund, which we won thanks to having a good team.
Now with Fund 3 and Fund 4, we’ve become more traditional. We have a couple of sovereign wealth funds, a few pension funds, some foundations, even a handful of former portfolio founders who have money and still want to stick around with us as investors.
A weekly dispatch featuring exclusive interviews with deep tech founders & a roundup of the most important deep tech news.
What stage are you typically investing at?
Very early. We’ve done a ton of deals on literal day one, or started working with founders before they’ve spun off from the university and invested a few weeks after they’ve spun off. The timing also depends on where the fund is in terms of the first-time investment period. Early in that period, you can take more risk regarding timing; later on, you need to be more cautious about timing-related risk. We still always see entry, but we might be more hesitant about taking time-related risk when the fund is later in the investment period.
Are you mostly focused on spinouts from universities?
I’d say that’s true. I’m pretty sure we’ve done more university spinouts than anybody else in the region. There are other funds operating in similar segments, but typically we move earlier than most and do more university spinouts and research agency spinouts. Players like Voima and Nordic Science Invest have a similar science focus as well, but we’ve been around since 2012, so we’re a bit ahead in terms of case count.
What is unique about the deep tech ecosystem in the Nordic countries?
In the Nordics the societies are functioning quite well, and people are solution-oriented. If there’s a problem, they focus on solving it. I used to have development teams in both the US and Finland, and they had communication troubles because Finns tend to talk about problems while Americans tend to talk about accomplishments.
As for tech sectors, Finland has a long inheritance in communications. For some reason, games are strong here, coming from the demo scene in the ’90s combined with the mobile tech industry. Computing is quite strong. Linux and MySQL come from Finland, for example.
The Nordics have a fairly strong tech ecosystem. The per capita likelihood of a unicorn from Nordics and Baltic countries is about seven times higher than the rest of Europe. They generate 17 times more value per capita than the rest of Europe. Even in absolute terms, it’s about double the United Kingdom. It’s a strong ecosystem that may not be so well known globally, but a lot has come from here: Skype, Spotify, lots of other successful companies.
That’s a pretty high hit rate. Why do you think that is?
It boils down to the overall ecosystem. There is a focus on results and solving real world problems. At the same time, there is a focus on companies being honest about who they are. There are also fundamental advantages that come from societal influences. There is a great higher education system, for example. It’s a safe environment. If you fail at putting together a startup, you’re not immediately destitute. And you still have a safety net since childcare is free and university education is free.
Are you mostly focused on spinouts from universities?
I’d say that’s true. I’m pretty sure we’ve done more university spinouts than anybody else in the region. There are other funds operating in similar segments, but typically we move earlier than most and do more university spinouts and research agency spinouts. Players like Voima and Nordic Science Invest have a similar science focus as well, but we’ve been around since 2012, so we’re a bit ahead in terms of case count.
How do you bring out the founder mentality in researchers?
I think it’s really important that a research team has had some time to form before starting a company. The team needs to own the idea that it’s important to them for their work to become a successful startup. That doesn’t happen overnight and isn’t something you can impose on someone. When we select teams, we look not just at the level of science or tech, but what kind of team has formed around it.
The Finnish approach to the pipeline these days is quite good. It’s one of the best recent innovations—it’s been operating for about five years, and I think it’s one of the best pipeline operations I’ve seen. In Finland there’s a research-to-business program that provides about €2 million for a university research group to package their work from a more commercial point of view, including bringing in a commercial founder.
About 70% of our cases are connected to universities in one way or another. There are differences between Nordic countries though. In Sweden, professors typically own a big portion of the IP created through research, whereas in Finland, the university owns the IP by default.
What sort of technology readiness level (TRL) do you look for in an investment?
Typically, the science risk has been tackled already, so it’s more about the product and getting it to market. It’s hard to specify an exact TRL number—it’s more in the 4-5ish range. It might sometimes be 3, sometimes a bit higher. But the TRL level doesn’t always correspond to time-to-market. We’re more interested in time-to-market and how quickly you can get validation from a commercial point of view. This doesn’t necessarily mean paying customers, but some kind of validation.
What categories or sectors within deep tech are you particularly excited about right now?
We’re always looking at sustainability, either social or environmental. Within those bounds, it’s quite wide at the moment. We see quite a bit of stuff dealing with computing efficiency. We recently invested in a company called Fault Computing that’s a parallel computing architecture solving problems in synchronizing threads and memory. They can achieve huge performance increases. I’ve seen demos showing more than 100-fold improvement in parallel computing performance.
This has an environmental angle in terms of how much compute and energy you use to get certain results, but it’s also a capability issue, specifically, what you can do on certain CPUs. We’ve also seen cases in AI that optimize AI models and match them to specific hardware, again pushing toward higher efficiency but also expanding capabilities. We’ve seen a fair amount of these “future of computing” type cases.
How do you define deep tech?
In our minds, it usually has science backing. Something solid beyond just packaging an idea into another platform. There’s often a hard IP, though not always. It’s more fundamental than someone just getting an idea to create an app. This is a question we ponder ourselves from time to time.