On the Ground With Europe and Asia's Startup Scenes

Over the past three weeks I traveled through Japan and Europe, visiting several software startup accelerators — plenty worth sharing from the trip. Along the way, I made a special stop in Lithuania, a country most people rarely visit, to see its startup scene firsthand.

Mention Lithuania to most people and conversation usually stalls — the only thing generally known is that it has fairly good relations with Taiwan; being able to say anything more about Northeastern Europe is a rare feat. Yet despite having a population of just 2.8 million, Lithuania has produced three unicorn startups. One of them, Nord Security — parent company of Surfshark VPN — is a name almost every Taiwanese internet user has heard of. Surfshark VPN has sponsored practically every influential YouTuber at some point; among my friends who make YouTube content, whenever someone lands a Surfshark sponsorship deal, we joke: "Congratulations — you're officially a certified influencer now!"

If you scale up by population ratio to match Lithuania's, Taiwan should have at least 25 unicorn startups of its own. Somehow, we don't. Next door, Estonia is even more extreme — roughly one unicorn per 150,000 people, which, scaled to Taipei City's population, would translate to about 20 unicorns.

Bolt, the ride-hailing service that recently entered Taiwan, is an Estonian unicorn that goes toe-to-toe with Uber in Europe. Even though we like to think of ourselves as a tech powerhouse, our track record in software and internet businesses still seems to lag well behind these small-population Northeastern European countries. There's a lot Taiwan could still learn from Lithuania and the Nordic states.

Europe: Lean Teams, Engineers Doing the Work of a Hundred

Having toured startup ecosystems in several countries, it's clear that VC and accelerator operating models have largely converged around a single global standard at this point. Venture capital itself has actually been around for a long time — the legendary firm Kleiner Perkins, for instance, was founded all the way back in 1972. But the software-and-internet-driven VC playbook we know today emerged much later, for the obvious reason that you needed an internet industry to exist first before dedicated internet-focused funds could show up.

Dedicated incubation and acceleration programs came along even later still. Even Y Combinator, the most iconic name in the space today, was only founded in 2005. Over the past several years, startup accelerators have gone through several waves of hype, and by now the model has become highly standardized and globalized — program formats and incubation methods are quite mature. Walk into a coworking space in Vilnius, Berlin, or Tokyo, and they all look like pages from the same catalog; the incubation programs run largely the same content too. At a glance, the teams inside all look similar too — assorted nerds of every stripe. So if everyone's using more or less the same playbook, why do some places manage to produce unicorn after unicorn?

One reason Lithuania has been relatively successful is that costs in Northeastern Europe run lower than in Western Europe, which has let Lithuania develop an engineering-outsourcing industry not unlike India's. My hotel, right on a busy downtown street, turned out to house an office for Wix, the Israeli website-building giant. An investor later told me Wix alone employs 400 people in Lithuania — no wonder their office looked like a street-front shop, with a full building and a prominent sign out front. This kind of service outsourcing has let local engineers accumulate substantial international experience, giving them a natural edge when they eventually start their own companies, and making it easier for them to connect with international capital.

On top of that, engineering labor costs across Europe as a whole run relatively high compared to the rest of the world — Germany, for instance, mandates over 40 days of paid leave a year. As a result, European startups tend to run remarkably lean on headcount, and their engineers often really do the work of a hundred. I suspect this owes something to the region's education model too, which does a lot to cultivate independent thinking in engineers.

A few years back, Sweden's Spotify became a case study people in the industry talked about constantly for its unconventional development process. Compared to typical software development, their team structure was full of creative choices you probably wouldn't come up with unless you were Nordic. Lithuanians even joke that maybe it's the terrible weather — Northeastern European engineers, stuck indoors, have nothing to do but focus on writing code.

But the more decisive factor is that these Northeastern European engineers consistently manage to produce extraordinary results with a remarkably small headcount. As a handful of unicorns have succeeded, investors have gotten a taste of the payoff and grown more willing to keep pouring capital into software startups with confidence.

Local investment-advisory bodies and trade-promotion agencies have, as a result, become quite fluent in the internet and software sectors. After all, in a country of just a couple million people, building out manufacturing isn't easy — one of the few industries that can both go global and succeed on a small population base is software. This mirrors the logic behind Taiwan's electronics-manufacturing success — success breeds further success.

Japan: A Buzzing Startup Scene Focused Squarely on AI

Heading back to Asia these past few days, I visited an international startup accelerator in Japan. This one was a bit unusual — it operates entirely in English, since more than half the founders there weren't Japanese.

I happened to arrive during an internal demo-day event and sat in on a session or two. Every single team was working on something AI-related, including teams that had traveled all the way from Central Asia — everyone was buzzing with excitement discussing their technology and startup ideas. I'd always assumed the Bay Area had a monopoly on AI-startup energy, but watching the excitement among investors and founders at this Tokyo accelerator, I think Asia is gradually building up that same heat.

The current mood in the startup world reminds me a lot of the mobile internet wave of 2012, or Web 2.0 back in 2008 — another global surge in software and internet innovation is underway, with an enormous number of opportunities emerging right now.