Accel was already one of Silicon Valley’s most successful venture capital groups when it decided in 2000 to open its first overseas office in London.
Since then, Accel has become a prolific backer of European start-ups, including consumer tech companies Spotify, Vinted and Monzo, games groups Rovio and Supercell, and corporate software groups such as Snyk and UiPath. It raised a new $650mn early-stage fund last year, after a study by Dealroom and Sifted found Accel had backed more European “unicorns” (private companies valued at more than $1bn) than any other VC between 2005 and 2023.
Harry Nelis is one of Accel’s longest-serving partners in Europe, joining the firm in 2004. A former start-up founder, engineer and investment banker, he led early investments in enterprise software groups Celonis and Personio, international payments group Zepz (formerly WorldRemit), and travel site Kayak.
In this conversation with Tim Bradshaw, the FT’s global technology correspondent, Nelis reflects on how the European start-up industry has changed over the past two decades. He also discusses the boom in artificial intelligence.
Tim Bradshaw: Accel is marking 25 years in Europe and you were among the first members of its London office. How have you seen the market change since then?
Harry Nelis: The key thing that I remember from the very beginning was, in raising our funds, investors asking, “Can Europe generate billion-dollar outcomes?” Up to that date, in the early 2000s, most exits in Europe were multihundred-million dollar exits. They weren’t billion-dollar exits. And successful venture capital funds can only be created if you have billion-dollar outcomes — outcomes that really move the needle on a fund.
To be honest with you, it was a while before the first billion-dollar outcome happened, which was 2005 in the form of Skype. And then there were one or two more in 2010. So imagine: we raised our first [Accel Europe] fund in 2000. And the first billion-dollar outcome took five years to materialise. And the next two took another five years. So, initially, it wasn’t that clear that Europe was going to develop into a vibrant ecosystem, the kind of ecosystem that it is today.
So if you ask yourself, “What really made the ecosystem work?” it is the following. In the beginning, there really weren’t any repeat entrepreneurs yet. So everybody who started a company did it for the first time. Then when they assembled their team, they assembled people who hadn’t been in start-ups before either: typically, the teams were first-time teams as well. Role models were missing. There were no successful European entrepreneurs yet. There were a few, but not the big outcomes that we’d seen in the US.
But over time, when you go through enough cycles, there’s a successful entrepreneur, they exit a company, their team wants to do it again, then a flywheel starts to happen. That flywheel today is happening on a really big scale. Unicorns in Europe — and we have around 350 of them — have generated 1,500 new start-ups. So there’s a flywheel going where employees from unicorns here in Europe start new companies here in Europe. And that flywheel is going with people who have done it before. And that’s the key thing that has changed.
TB: For a long time it felt less like a flywheel and more like a conveyor belt: if people got successful in Europe, they just moved to the US. When did you see that start to change and people start to realise they could make it big as a European company by staying here?
HN: I think that happened when there was enough capital here to support them. So in the very beginning, there were only one or two firms supplying capital to entrepreneurs here and so some entrepreneurs decided to go to America. But with the steady supply of more early-stage capital and follow-on capital, many entrepreneurs have realised that Europe is a great place to start a business, especially when it comes to engineering. We’ve got great universities. Our employee base for engineers actually is more stable here than it is in the US. In the US, there are so many opportunities that people end up leaving very quickly, whereas here in Europe, people tend to stay a bit longer in their engineering jobs.
TB: In that quiet first decade, did you or Accel’s investors ever think, “Maybe we should just go back to the US?” Because there have been several instances over the past 25 years of US investors coming to Europe and then pulling back again.
[Andreessen Horowitz, for example, announced it was closing its less-than-two-year-old London office after this interview was conducted.]
HN: We were lucky in the sense that we had a set of partners here in Europe and a set of partners in the US who really wanted to make this work in the long run. Typically Accel thinks long term and knows that it takes a long time to build a valuable business. It involved a lot of travelling back and forth. It involved spending a lot of time together. There are several firms where it ended up not working and they ended up kind of breaking apart. But, in our case, through determination and the investment in relationships between the partners here in Europe and in the US, it ended up working really well. In fact, it worked so well that, in 2008, we decided to open an office in Bengaluru, India, using the same kind of model.
TB: And was there one investment or deal in Europe that either crystallised it for Accel that this was going to work, or can you show at your own next fundraising that Europe had turned the corner?
HN: I think for us, the really big one was Supercell [which was acquired by Chinese tech group Tencent in 2016 in a deal valuing the online games company at $10bn]. We did the series A investment [with] a great Finnish entrepreneur, Ilkka [Paananen, Supercell chief executive].
The interesting bit is that Supercell’s first game was not successful. Their second game was not successful. And then their third or fourth game [Hay Day and then Clash of Clans] was a big success and it really put Supercell on the map. So it shows you that great companies take a very long time to build and that having patient investors who keep supporting you is the key. Spotify [which was founded in 2006] was another breakout company.
TB: If we fast forward to today, if Europe now has the kind of talent and ambition that it lacked when you started, where do you see the bottlenecks now? Is it Europe’s regulation, fragmentation, access to compute capacity?
HN: One of the things that we’ve learnt over the years, is that venture is a people business. And what makes great companies is the entrepreneurs behind them. I think we continue to have a really steady flow of great entrepreneurs emerging. I don’t think much is holding them back. We have a pretty wide and long runway ahead of us. We have a fully functioning venture ecosystem here. We have universities that are educating business leaders and engineers at world-class levels. Great companies can come from anywhere. That’s the other key thing that we’ve realised. We’ve realised that Supercell came out of Finland, which was not a top five market. UiPath came out of Bucharest, Romania. Celonis came out of Munich. So great companies can come from anywhere. And there’s a steady supply of them.
TB: What are the biggest mistakes that European entrepreneurs make that their peers in Silicon Valley don’t? Do you feel like there are still differences in how we approach building our businesses here?
HN: Traditionally, entrepreneurs sold too early. And that generated the $100mn outcomes. But that has changed dramatically over the years. And that has to do with ambition. A lot of it has to do with role models. And having the role models of Supercell, of UiPath, of Spotify, where you can build global category-defining businesses, has dramatically changed that. And that’s why we’re seeing these big outcomes.
What has changed is that becoming an entrepreneur is an acceptable career path. When I was an entrepreneur and I came back home from Silicon Valley, people would ask me, “Why? Why aren’t you going to work for Shell? Why aren’t you going to work for Philips? Why do you have to do this?” Whereas if I had to tell people the same thing in America, they’d say, “That’s great. Good luck to you.” So that mindset has completely changed.
TB: Do you think there is still a valuation discrepancy between European and US start-ups? Are investors here getting a better deal?
HN: No. I wish we were.
TB: When did that even out? Because that is something I heard people talk about, even in the most recent wave of American firms landing here.
HN: I think the real motivation for more US firms setting up shop here is the realisation that great companies can come from anywhere. And therefore [those firms] had to be here because they will come from Europe too. I don’t think there was an incentive or a thought that, “OK, we’re going to get cheaper deals here than we are going to get in America.”
TB: How does that influx change the market here for you?
HN: Oh, the market has become much more competitive. In the beginning, essentially you had Index Ventures and Accel and a few regional firms. But now all the big firms are here as well. But not only is the market more competitive, it has also grown a lot. So we still feel that we get more than our fair share of the great deals out of Europe. It’s roughly 40 to 50 per cent the size of the US, in terms of venture funding.
TB: What does that competition and extra capital mean on the ground? Does it mean that prices get bid up on the most exciting companies? Does it mean that the wrong companies get funded because people just have to deploy their capital?
HN: No. What it means is that we need to be running even harder than we did in the past. That we need to spend even more time on aeroplanes flying across Europe to speak to an entrepreneur. It means that deals happen faster. It means that sometimes in the case of a great company, the valuation becomes very high because there are multiple people going after it — in the same way as it happens in the US. It’s a faster market.
TB: Did it get too fast in the last boom, during Covid? There were deals getting done in days in that period. Have we all learnt our lessons from that, as we go into AI’s [fear of missing out] moment?
HN: The Covid-19 pandemic was a really distorting event. I think some people learnt their lesson and others did not. There was a lot of capital that became available to the market. Enterprises, companies bought software as if their life depended on it. And in many instances, their lives did depend on it, right? Because you had to work remotely. So there was a real reason to build our start-ups fast, because companies were buying. So we had to build sales forces that became bigger and bigger. And then of course, things went in reverse because interest rates went up, less capital became available and in many ways companies overbought software. And the logical thing that you do after you have overbought software, is you end up buying less and rationalising what you’ve bought. So it’s been a very distorting period for start-ups.
TB: Isn’t AI an equally distorting period for start-ups?
HN: It’s a different one. I see it more similar to the internet boom that happened in 1999. People realise that this is a fundamental shift in technology and a shift that will cause the entire tech stack to be reinvented. And there are many, many opportunities in that. People want to be on the ground floor. And with many technology adoption cycles, people tend to overestimate the speed with which it happens and then underestimate the eventual scale that that trend becomes.
TB: With that in mind, do you feel like people are behaving rationally at the moment in AI?
HN: I tend to think that there aren’t that many irrational people. Most people think rationally but from their [own] perspective. And those perspectives sometimes are very different. But it is very possible that AI start-ups will have much larger outcomes than we’ve seen in the past. If the outcomes are much larger, it’s possible that you can afford to pay a bit more in the beginning. We will only learn the answer 10 years from now.
TB: The richest companies in the world are more valuable than they’ve ever been and they’re all tech companies. So, is the argument that suddenly we’re talking about trillion-dollar outcomes rather than just billions? Why could the outcomes be that much larger?
HN: Because technology is a much bigger part of our life now than it was in the past. Technology impacts a much larger part of society, of GDP, if you will, than it did in the past. As a result, the kind of companies that you can build on that basis are bigger companies.
TB: So how’s Europe doing on AI?
HN: I think where Europe has done well on AI is on research. So we have a number of leading universities in the UK and in France that have a disproportionate impact on the industry.
That caused companies such as Facebook to set up [its AI research lab] in Paris, Google to buy DeepMind in London. So Paris and London have become the centres of AI in Europe. There’s been a number of interesting companies that have been formed here, whether it’s Mistral in Paris, a number of application companies, whether it’s Synthesia in London and others. The US has been ahead in terms of infrastructure investment in AI, but we’re at the very beginning of this long cycle. Our expectation is that, over time, the playing field will even out.
TB: Does this tell us anything about our ability in this part of the world to commercialise research or to spin out start-ups from universities? Is that where the UK or other parts of Europe have still not quite got such a well-oiled machine as Stanford in Silicon Valley, Stanford to [the VC firm cluster on] Sand Hill Road?
HN: No, I think that that is working quite well. We did some interesting research on AI and 38 per cent of founders of [European] AI companies came from positions in universities, as professors or as researchers, which was an astoundingly big number in my mind.
[The figures are based on an Accel study of 221 start-ups across Europe and Israel.] And then 25 per cent of co-founders come from DeepMind, Amazon, Google, etc. So that path is working.
TB: To your point on outcomes, the most recent big European AI takeover was US chipmaker AMD buying Finland’s Silo AI for $665mn. That was the biggest European AI M&A deal since DeepMind in 2014. It amazes me there haven’t been more of that size in the past decade.
HN: We’re still at the very beginning of the AI cycle, maybe we’re two years into it. But this cycle can run for 10 to 15 years. We’re really at the very, very beginning.
TB: What gives you the faith that AI will improve, either as a business or as a technology over that period?
HN: Currently companies are experimenting with AI. Everybody is experimenting with AI. But there are very, very few real applications in operation today. So it’s all experimental budgets and no real applications yet. Partially that is because of the practicalities of applying AI in a business: they need to be secure, they need to deal with private information properly. So there’s always a period in which things need to go from an experimental lab within a company to being in operation. And that takes time.
TB: So what gives you the faith that it will come?
HN: Within our start-ups, we’re seeing interesting areas [of AI’s application] — whether it’s customer service, whether it’s engineering productivity, whether it’s underwriting in loans — where it has an impact already. But then investors and venture capitalists are optimistic and we tend to extrapolate those kinds of things to grander things. But of course, it’s also informed by 25 years of being on the ground. It’s my conviction — and this is not an outlier opinion by the way — that this is a big cycle and this is really going to work.
With every company, entrepreneurs and investors should ask themselves: is AI helping this company or is AI threatening this company? Most of the [start-up] companies that we’re involved in, AI is a help. And part of that is the question of who owns the information, who has access to data? The more companies that we invest in where they are close to the data and have access to information, the more AI is an opportunity rather than a threat.
TB: In a market where there is more competition among VCs and a larger share of biggest AI deals are going to Big Tech companies or sovereign wealth funds, is there an argument that VC returns have peaked? Are the VC industry’s best years behind it?
HN: I don’t think so. It is very possible that the kind of companies that we invest in today end up becoming bigger companies than we’ve seen in the past. So if prices are somewhat higher now, but if the outcomes are also commensurately higher, it’s very possible that the returns stay the same or even increase over time. Returns, in my mind, are driven by the scale of the disruption and the quality of the entrepreneurs. The quality of the entrepreneurs is higher than it’s ever been and the scale of the disruption arguably with AI is of a really high scale as well. It’s comparable to the internet, mobile and cloud.
TB: What are the biggest lessons you’ve learnt over Accel’s 25 years in Europe?
HN: One thing is venture is a people business. In the end, whether it’s the entrepreneurs, whether it’s the investors on our team, venture is a people business. Number two is it’s all about the outliers, in terms of a successful venture fund. And then the third one is, great companies can come from anywhere.
TB: Has what you look for in an entrepreneur changed, compared with 25 years ago?
HN: Yes, it has. Ambition has become less of a filter, because people are pretty ambitious today. So really, the thing that we end up looking for is whether they are able to lead and whether they are able to tell a story that will get co-founders on board, that will get employees on board, that will get investors on board, that will get customers on board.
TB: Some of the entrepreneurs who were most successful at fundraising over the last [Covid-era] cycle did not turn out to be the most successful business leaders. [Accel previously backed Hopin, a virtual events start-up that raised almost $1bn in 2020-21 but was broken up and sold for a tiny fraction of that sum in 2023-24.] How do you see through the pitch to see what’s going to work?
HN: You look for substance. You end up trusting your 25 years of experience. By the way, the biggest mistake in venture is not losing your investment. The biggest mistake in venture is missing the outlier. Being wrong sometimes is OK, but you had better not miss the outlier.
