11 minute read
Secrets of the unicorns
What it takes to build a $1 billion dollar company.
According to Statista*, as of April 2021 there were 591 unicorn companies in the world. The large majority of these are located in the US (288), followed by China (133), India (32) and the UK (27). But there’s one region of the world that’s seeing more unicorns galloping into existence than ever before – Europe.
Advertisement
In data provided by venture capital firm PitchBook, 23 companies in Europe and Israel became unicorns between 1 January and 10 June 2021 alone, compared with only eight across the entirety of 2020. The current Statista European unicorn count stands at 69*.
During 2020, the region recorded unicorn growth in the fields of fintech (Sweden’s Klarna), banking (Germany’s N26) and automotive (Germany’s Auto 1 Group), but 2021 has seen unicorns form in new sectors entirely – with many companies receiving a clear COVID-19 lockdown boost.
Newcomer industries include takeaway food delivery (Finland’s Wolt), digital telehealth (Sweden’s Kry), online meal kit delivery (UK’s Gousto), mental health treatments (Germany’s atai Life Sciences), grocery delivery (Germany’s Gorillas), and digital-first pet insurance (UK’s Bought By Many).
With the European unicorn stampede showing no signs of slowing down, and hundreds more companies being labelled as ‘soonicorns’, the remainder of 2021 looks set to be truly legendary for entrepreneurs across Europe.
Did you know?
The collective noun for a group of unicorns is called a ‘blessing’. How fitting!
Number of unicorns in Europe* 69
in total
11
1 5
6
12
8 2
3 4
10 13
9
1. United Kingdom: 27 2. Germany: 15 3. France: 8 4. Switzerland: 5 5. Sweden: 3 6. Netherlands: 3 7. Spain: 2 8. Luxembourg: 1 9. Lithuania: 1 10. Croatia: 1 11. Ireland: 1 12. Belgium: 1 13. Estonia: 1
7 What is a unicorn company?
A unicorn is a privately owned startup valued at US$1 billion or over. Once a company has gone public or been acquired, it’s no longer considered a unicorn.
The term was coined by venture capitalist Aileen Lee in 2013, using the mythical creature to describe the rarity of this level of success occurring.
An Imperial unicorn
In June 2021, Imperial alumnus Alex Dalyac’s startup Tractable was valued at over US$1 billion, making Alex a very proud unicorn founder.
If you’d asked Alex back in 2013 whether the group project he was working on for his Imperial master’s degree in computer science – which involved him taking photos of daffodils – would lead to the birth of a US$1 billion company, he would have told you the likelihood was as rare as … well … a unicorn. But that’s exactly what happened. Tractable was the world’s first computer vision unicorn for financial services and is one of only 20 computer vision unicorns in the world. We spoke with Alex about the path his business journey has taken and his advice for fellow entrepreneurs.
Can you tell us about computer vision?
Computer vision revolves around algorithms that can make sense of imagery and video like we do as humans, so recognising objects and understanding actions being taken by the various entities. More specifically, we’re focused on putting a ‘visual expert’ in people’s pocket that can help them sort out insurance claims for their car or home.
What exactly is a ‘visual expert’?
Imagine you’re driving your car and you get into an accident – instead of dealing with a month-long horrible process with your insurer, you simply take out your phone and Tractable’s AI guides you around your car to capture video and imagery. It then figures out what it costs to repair your car and takes care of the insurance claim for you, directly with the insurer.
We’re expanding this to homes as well. If floods damage your home, instead of waiting two months for the insurance company to send somebody out to look at it, you just take out the visual expert in your pocket and let that take over the appraisal so you can receive a payout from your insurance company sooner.
Amazing! How did you come up with the idea? Were you trying to solve a problem?
It was all thanks to Imperial College London actually! I was doing a master’s in computing at Imperial in 2013 to 2014 and got into image classification with deep learning, which was a groundbreaking technology at the time. This was just before AI company DeepMind was acquired by Google. In my second term, there was a group project where we built a plant
recognition app which was almost like a visual expert in your pocket for plants.
I remember going to Hyde Park and taking photos of daffodils in the wild, and the AI just working. This was a first – it had previously been impossible for computer vision to recognise objects and cluttered backgrounds in broad daylight because those would confuse the computer vision algorithms.
Then in my third term, my supervisor Jack Kelly put us in touch with a plastic pipe inspection company. The pipes under the ground in London get old and can crack and break. This can lead to contamination of water or gas pipe explosions, so it’s quite important to upgrade them. There’s a key step in the construction process, which is welding two pieces of pipe together – you need to heat up the pipes, connect them and let them settle. If you don’t set up your weld well, there’s going to be a fault area where water and earth can get in and contaminate what’s being sent to people’s households.
It was then that I thought: “What if construction workers had a visual expert in their pockets that they could pull out to make sure the weld set-up is correct before proceeding with it?”. That ended up being my master’s project. The results for the pipe inspection company were great and they ended up being Tractable’s firstever customer!
Right after that, I realised that Tractable was going to be the visual expert in people’s pocket and we could use it for a wide variety of use cases. Pipes were clearly fine, but it wasn’t a big enough market. We realised that accident and disaster recovery was bigger, so we ended up focusing on cars and homes.
The second organisation to be credited for Tractable is Entrepreneur First (EF). They are a startup accelerator programme that’s now global, but it started in London. It helps individuals who want to start a company, but who don’t yet have a co-founder – they’re very unique in that way. I applied while I was at Imperial, and that’s where I met my first co-founder, Razvan Ranca, who was a researcher in machine learning at Cambridge. Thanks to EF, we raised our first round of funding, which was US$1.9 million, and that’s when our third cofounder joined us – Adrian Cohen. He was a more experienced businessman. I had met him previously in Vietnam when he was bringing online shopping to Vietnam for the first time.
You’ve mentioned that Imperial gave you support on your entrepreneurial journey. How did that start?
Imperial gave an incredible amount of academic freedom and a lot of bandwidth to projects that let you build products that are actually useful in the real world, as opposed to coursework and exercises. Imperial encouraged you to do research at the global forefront so your academic work could make a difference. It’s not imposed on you – you get to choose. If you have your own project, you’re encouraged to pitch it and find an academic who wants to support it. That was incredible. It actually culminated in a PhD offer with Imperial, too. I found it really amazing that Imperial believed in me so much!
What did you want to do as a career when you were younger?
As a teenager when I was graduating from high school, I remember thinking that going into science and engineering, rather than a social science, was going to be boring, isolated and dry. I couldn’t have been more wrong! Going into science and engineering allows you to build new things that didn’t exist before and help people live better lives … and to me, that’s unbelievable – it’s almost like magic!
How long did it take you to go from your initial idea to your billion-dollar valuation?
It’s taken 6.5 years. It was extremely hard getting to our first million dollars of revenue. Raising our first round of funding – which is seed – just required an exciting idea, but the next stage was hard and required a lot of work.
What are the next steps for Tractable?
One thing we’ve been working on over the past year and a half that we’re really excited about and want to expand globally is recycling cars. When cars get damaged, a quarter of the time they’re too expensive to repair, so they end up going to the scrapyard. But it turns out that you can often upcycle much of the car. If the parts are not too damaged, you can reuse the door as a replacement part so when another car gets damaged and needs a new door, you don’t need to buy a brand-new manufactured door – you just upcycle the other one. This has major CO2 emission avoidance impact. For every car you upcycle like this, it’s about half a metric tonne of carbon dioxide emissions avoided. We’ve already started doing this in North America.
The other thing we’re doing is natural disaster recovery. When you think of climate change, one of the worst impacts is the increase in frequency and severity of natural disasters – whether that’s bushfires and wildfires in Australia and California, hurricanes in America, typhoons in Japan or severe storms in Europe. These can damage our homes on a really large scale and leave people without a proper roof over their heads for months on end.
As insurance companies are totally understaffed to deal with this, wouldn’t it be good if there was a completely scalable solution? People could rely entirely on software – with the visual expert on their phone – to get an appraisal on their home and a payout from their insurance company. That way, you’re giving people certainty and peace of mind to fix their home and get back to a normal life quickly. We announced in July that we’ll be doing this in Japan with one of the country’s biggest insurance companies – Mitsui Sumitomo. We hope to be live this autumn and help at least 1,000 families to start rebuilding their homes much faster.
What advice would you have for any Imperial students thinking about commercialising their research?
First of all, well done – you’re in a great place to get started! Before commercialising your work, you’ve got to create something that doesn’t exist yet but that solves a problem much more effectively than anything else out there. That’s hard to do if you’re a student, but the key is being aware of scientific and engineering breakthroughs, because when they happen, they totally change the rules of the game.
In our case, this was moving from traditional computer vision algorithms to deep learning. It meant that the years of work people had put into developing traditional computer vision techniques were no longer useful, so it totally reset the playing field for us. By diving into deep learning and getting straight into this new paradigm, we ended up among some of the best skilled people out there – and that’s ultimately what gives you the edge.
The stats of the super founders
We were proud to welcome Ali to the Imperial Enterprise Lab earlier this year when he hosted our ‘How to … be a superfounder’ talk and discussed his new book. Ali’s Super Founders: What Data Reveals About Billion-Dollar Startups is out now and can be purchased at superfoundersbook.com
Imperial Biomedical Engineering alumnus, Ali Tamaseb, was so keen to discover the secrets of unicorn founders that he interviewed and analysed more than 200 of them. In his book Super Founders, Ali gives us the stats behind the success …
As a partner at Silicon Valley venture capitalist firm Data Collective, Ali Tamaseb was often asked about unicorns. What was it that made them reach that magical US$1 billion valuation figure when so many others failed? Was it luck? Or something else, like market or team or fundraising? Was it their founders? What did they do differently? Ali set about finding out.
He collected 65 different data elements from unicorn companies in the US that had formed during the previous 15 years. He collected the same 65 data elements from every other US company that had raised a minimum of US$3 million in investment but had not become a unicorn later on – and used these as his baseline.
Comparing and contrasting over 40,000 data points, he discovered some truly interesting insights about what made unicorn founders so special. His findings became the book Super Founders: What Data Reveals About Billion-Dollar Startups. This is what he discovered …
Do you have what it takes to be a unicorn founder? 34
The average age of a unicorn founder is 34.
20%
of unicorn companies had solo founders.
45%
of co-founders had either attended the same school or worked at the same company.
11 years
Founders had an average of 11 years’ experience before their companies became unicorns.
49.5%
of unicorn founders had a technical background.
5.5
It took on average 5.5 years for a company to go from launch to unicorn status.
4%
of unicorn founders were university dropouts.
36%
of unicorn founders went to a top-ten-ranked school (great news for Imperial alumni!).
85%
of unicorn companies had competition from day one.
70%
of unicorns were expanding the market, not creating it from scratch.
60%
of unicorn companies were started by repeat entrepreneurs, and many already have at least one US$50 million+ exit under their belt.