If you were to conjure up a stereotypical mental image of a typical computer science professor, Daniel Kroening might well be it. Tall, angular, with wire-rimmed specs and a high-domed forehead framed by a halo of wild, curly hair, his fingers trace an ergonomic keyboard unrecognisable to most of us, while his English has soft traces of his native German accent. But Kroening is not a typical computer science professor. Because he is that rare thing in Britain: both an academic and entrepreneur.
As well as being Professor of Computer Science at Magdalen, Oxford, he is the founder of Diffblue, a start-up that wants to develop artificial intelligence that can write its own computer code. In his office, a stone’s throw from Oxford’s station, he asks a simple question: “Do you know the average number of spin-outs each Professor of Computer Science at Stanford has?” Spin-outs are commercial companies founded on the back of academic research. Stanford, based in the heart of Silicon Valley, is recognised as the world leader in the art of the spin-out. “The average number of Stanford spin-outs is 3.8 – per professor.”
In Britain the picture is rather different. In 2015 Oxford, the UK’s number one university for research, produced four spin-outs. Not per professor. That was for the whole university. The situation was not better elsewhere. Data on British university spin-outs is not in any publicly available league table. But it exists, via what’s called the HE-BCI survey (it stands for Higher Education – Business and Community Interaction). For 2015-16, Cambridge University recorded a total of two spinouts in the HE-BCI survey. Imperial College London, another of this country’s most vaunted research universities, listed three. Of 160 institutions, 59 officially produced no spinouts at all.
Of course, it not easy to be perfectly precise about numbers of spin-outs. Just when does a new company cease to be related to the academic institution from which its founders emerged? But no one is in any doubt that Britain’s pipeline of new companies founded on the pioneering technology developed at our world-leading universities is blocked. “It hasn’t really worked half the time between academia and industry,” says James Dyson. “They are connected but not that well.” One of major tasks for Rachid Hourizi at the new Institute of Coding will be unblocking the pipe. “Everyone wants the link between universities and businesses to work better,” he says. “We’re trying to jump that forward.” At the moment, however, it is so hard to take an idea from a British university and turn it into a successful business that the gap between academia and profitable company is known as The Valley of Death.
The benefits of crossing the valley are obvious. David Cheriton is a professor of Computer Science at Stanford. He has a long history of mentoring academics and students who want to commercialise their ideas. In 1998 a pair of Stanford PhD students were looking to start up a new company to make sense of the mass of information on the internet. Struggling to raise money, they were after advice and turned to him. Yahoo – remember them – had turned down the opportunity to license their search algorithm. But Cheriton saw the potential in the concept, and even invested $100,000 of his own money. The two students were Larry Page and Sergey Brin. Their company was called Google. Barely 15 years after making his investment, Cheriton’s stake was worth more than a billion dollars. The message is clear: University research has the capacity to build, in double-quick time, the world’s most valuable firms.
Larry Page and Sergey Brin
Google’s co-founders
Google’s co-founders
Google headquarters
Mountainview, California
Mountainview, California
Among Britain’s dreaming spires, however, money has too often been a dirty word. “Cash is grubby in our universities,” says Harry Briggs. “Ask people at Oxford about Niall Ferguson. He was a hero; now they say ‘He’s a sell out because he writes popular books! Books people read!’” Indeed, in Oxford the situation got so bad that, in 2015, Oxford Sciences Innovation (OSI) was founded specifically to fund spin-outs from university research. The results have been astonishing. Within three years, OSI has become the biggest fund dedicated to a single university in the world. “We have £606 million to invest,” says its CFO, Jim Wilkinson. “Nothing like this exists anywhere.”
The model is clear. OSI scours 29 university departments for prospects (“from Advance Engineering to Zoology” as they like to say), then mentors and funds the budding academics/entrepreneurs it finds. In the three years since it was founded, the number of spinouts from the university has risen from an average of 4.5 a year, to 17. It now has a stake (on average about 30-40 per cent) in a total of 51, from 1-person ventures worth as little as £1m, to those with a dozen or more staff with an estimated value of £50m-plus. They range from AI-based ideas, like DiffBlue, to Oxonomy, which simulates maritime trade. “For 900 years Oxford University has been an academic superpower that hasn’t been capitalised upon,” says Riwa Harfoush, Principal at OSI. “Today we will sit down with academics who might not even see the commercial potential of their ideas themselves and build a company around them. We will have companies that will change the world, and for the better.”
For example one of OSI’s spinouts, Ultromics, uses machine learning to spot flaws in cardiostress echo-tests, which currently have only a 75-80 per cent accuracy rate. “At the moment, says Harfoush, “One in five people are either sent to have open heart surgery they don’t need, or are sent home with a deadly disease.” Ultromics aims to improve that to 95 per cent, potentially saving the NHS £2bn a year. “We’ll give it to them for free,” says Wilkinson. “But sell it to everyone else.”
“In the UK, the role ‘product manager’ doesn’t even exist. If I was a Chancellor at Oxbridge or Imperial, that’s what I’d really focus on changing.”
Some other universities are following OSI’s example, opening “accelerators” and “incubators” to fund and nurture their own researchers’ ideas. But a broad cultural shift is still required in academia, where founders are all-too-often seen as shysters, unwanted Del Boy Trotters of the laboratory or department. “I don’t see why every university doesn’t have a business accelerator,” says Stefan Allesch-Taylor, KCL’s Professor of Entrepreneurship. “The problem is they still do look down on entrepreneurship as a little bit ‘Cor Blimey Guv.’” As a result, the field of product development – turning an idea into something customers actually want to buy – barely exists in British academia. “In America you can really study it,” says Fergal Mullen, a venture capitalist who focuses on early stage software and internet companies. “The usability of your product, the software you are designing, say, and the estimated size of its market. In America, all the best universities have courses in ‘product dev’ often linked to engineering or computer science. In the UK, the role ‘product manager’ doesn’t even exist. If I was a Chancellor at Oxbridge or Imperial, that’s what I’d really focus on changing.” Even the Government’s own digital strategy casts jealous glances at: “Stanford University’s 145 education courses on enterprise which feed into Silicon Valley employment opportunities”.
Some other universities are following OSI’s example, opening “accelerators” and “incubators” to fund and nurture their own researchers’ ideas. But a broad cultural shift is still required in academia, where founders are all-too-often seen as shysters, unwanted Del Boy Trotters of the laboratory or department. “I don’t see why every university doesn’t have a business accelerator,” says Stefan Allesch-Taylor, KCL’s Professor of Entrepreneurship. “The problem is they still do look down on entrepreneurship as a little bit ‘Cor Blimey Guv.’” As a result, the field of product development – turning an idea into something customers actually want to buy – barely exists in British academia. “In America you can really study it,” says Fergal Mullen, a venture capitalist who focuses on early stage software and internet companies. “The usability of your product, the software you are designing, say, and the estimated size of its market. In America, all the best universities have courses in ‘product dev’ often linked to engineering or computer science. In the UK, the role ‘product manager’ doesn’t even exist. If I was a Chancellor at Oxbridge or Imperial, that’s what I’d really focus on changing.” Even the Government’s own digital strategy casts jealous glances at: “Stanford University’s 145 education courses on enterprise which feed into Silicon Valley employment opportunities”.
“If we have a more global immigration policy after Brexit that would be a good thing.”There’s another problem, too. Many embryonic university businesses are founded by foreign students, who flock to Britain for the quality of the teaching. But having taught them all they know, they are allowed four months to leave the country at the end of their course by the Home Office. This is stringently enforced because students currently are counted in the UK net immigration number, which the Government has promised to reduce “to the tens of thousands”. In British technology this cycle – inviting, teaching then expelling – the brightest talent from around the world, attracts near universal eye-rolling. “60 per cent of people reading engineering at British universities are from overseas and 90 per cent of people doing research at universities are from overseas and we should encourage them to stay,” says James Dyson. “We shouldn’t kick them out at the end of it, we desperately need them.” He doesn’t see Brexit as a problem, arguing that EU membership merely skews the balance of those we do let in away from high growth areas like China and India, towards EU nations. “If we have a more global immigration policy after Brexit that would be a good thing.”
“Foreign students are a huge pipeline of talent”
The frustration with counting students as immigrants, which Theresa May, herself a former Home Secretary, has refused to budge on, is ubiquitous. “Foreign students are a huge pipeline of talent,” says Tech City’s Gerard Grech. “Many of the successful businesses I’ve seen here have been started by immigrant students.” Even at DCMS, Matt Hancock seems embarrassed by its illogicality. Normally loquacious, he clams up, pursing his lips and repeating: “We are determined to attract the brightest and best people from around the world.” Asked if he would like to see students removed from the immigration stats, he stonewalls: “I am happy to support Government policy on this.” Many who talk to him suspect that, privately, he thinks it is bananas.
The frustration with counting students as immigrants, which Theresa May, herself a former Home Secretary, has refused to budge on, is ubiquitous. “Foreign students are a huge pipeline of talent,” says Tech City’s Gerard Grech. “Many of the successful businesses I’ve seen here have been started by immigrant students.” Even at DCMS, Matt Hancock seems embarrassed by its illogicality. Normally loquacious, he clams up, pursing his lips and repeating: “We are determined to attract the brightest and best people from around the world.” Asked if he would like to see students removed from the immigration stats, he stonewalls: “I am happy to support Government policy on this.” Many who talk to him suspect that, privately, he thinks it is bananas.
So if you have a good idea, how do you found a start-up?
Like many people, Alice Bentinck was a little nervous about taking her boyfriend home to meet her parents. But she was not nervous because she didn’t think they would approve of him. She was nervous because she was worried they wouldn’t approve of the way they had met: online. “So she told him to tell them that they’d met in a bar,” says Matt Clifford, with whom Bentinck co-founded Entrepreneur First, a hothouse six-month course which spawns some of Britain’s best new tech companies.
Alice Bentinck
Entrepeneurs First
Entrepeneurs First
Essentially EF, as it is known, is a start-up hatchery which aims to force the conception of businesses too valuable to be left to the messy, chancy, often unproductive method of natural, organic entrepreneurial birth. Every six months it picks 100 people from 1000s of applications. Then it puts them together in combinations to maximise the chances of start-up fertilisation. “Starting a company is so important you want the right partner,” says Clifford. “Why put this to chance?”
Why indeed, when the rewards from getting it right can be immense. After the first three months, which are “about finding a co-founder and a fabulous idea”, EF discards half of the initial 100. It then takes a 5 per cent stake in the 20 or so remaining partnerships, in return for an investment of £80,000, immediately valuing each at £1.6 million (EF claims a further 5 per cent as a “finders fee” for matching up co-founders). The next three months are all about getting the hatchling companies ready for investment, developing some proof that their new product or technology might actually work, for example, and providing evidence that consumers will buy it. Then, at the end of the six months comes “Demo Day”, when Britain’s most important venture capitalists turn up to listen to three-minute pitches. Increasingly they come from America and Europe too. “Every single person who might invest in these companies is in the room,” says Clifford. “It is the most highly-leveraged three minutes of our founders’ lives.”
The average valuation of each partnership at the end of the day is £4.5 million. “Not bad for a six-month-old company,” says Clifford. It’s not a foolproof process, of course. One EF alumni, Magic Pony, only managed to sell a 30 per cent stake on Demo Day, for £1.5m. A mere 15 months later, however, Magic Pony was acquired by Twitter for $150m. The investors who did have faith made about $45m – $3m for each month they held their stake.
EF may be taking the mystery out of entrepreneurialism, but it appears to be working. “Picking from a curated group of people is the smarter way of doing things,” says Clifford. “It’s like online dating. 40 per cent of weddings today are of couples who met online.” Indeed, the reason I am speaking to him, and not him and Bentinck, is that she is planning her own nuptials – to the man she met “in a bar”.
Most people are not lucky enough to get taken on by EF. Their start-ups are founded the good old-fashioned way, by people meeting randomly, starting a conversation, realising they share an idea, and then plucking up the courage to give it a go. Herman Narula, founder of Improbable, one of Britain’s billion-dollar “unicorns” met his co-founder Rob Whitehead while they were queuing to have their dissertations reviewed at Cambridge. “We started talking and within literally 30 seconds we had planned to build Improbable,” he says. “I’ll never forget that conversation. The next day we coded for 12 hours, starting to put it together.” Narula and Whitehead soon retreated to a barn on the grounds of Hyver Hall, his parents’ mid-19th century home near Barnet. But other smitten partners usually have to get a room – and pay for it.
Hence the popularity of a host of “shared” or “co-working spaces” in London. They have names like WeWork, RocketSpace, Plexal and Second Home. If you have an idea, and the cash to fund it for six or 12 months, the chances are you’ll end up in something like this. Why? It’s easy to expand – essential for fast-growing start-ups. And contracts are often month-by-month, rather than standard 5-year commercial terms, useful it all goes down the tubes, as it so often does. Meeting rooms, phones and internet connections are there on tap, and are host of extras of often provided too: lectures by successful founders or funders; and of course the chance to network – and perhaps team up with – a host of like-minded souls.
Such spaces are a neat barometer of a city’s entrepreneurial health, and these days London is the global capital for co-working spaces, ahead of New York. Research shows that in central London, 2.5m sq ft was leased to flexible workspace providers in 2017, a 190 per cent increase on the previous year. In a decade, the total stock of flexible workspace in London has risen from less than 4m sq ft to more than 10m. Such is growth that, in the last five years, WeWork has added more office space in London (2,578,000 sq ft since 2012) than any other company; more than Google (1,344,000 sq ft), and Amazon (1,013,000 sq ft). In fact, only the Government now has more office space in London now than WeWork, a staggering sign of the development and scale of the new flexible, start-up, gig economy largely driven by digital technology.
It is in a WeWork office that George Davis, the 19 year-old UEA dropout, now runs Senlab, an archetypal start-up story of endless cash crises, pitching to investors, developing the product – a constant battle between setbacks and self-belief. Senlab’s software, Prosper, uses machine learning to help small businesses to predict the financial bottom line. Most such companies fear running out of money above all things, and George Davis has created a program that alerts companies to the cash crunch they can’t see coming – showing them what Davis calls “their graph of death” and identifying the very moment when the Grim Reaper might come calling. The hope is that his bony fingers can be dodged as a result.
George Davis
Senlab CEO
Senlab CEO
Martina King
FeatureSpace CEO
FeatureSpace CEO
The Senlab story began just 18 months ago, in November 2016, at a conference centre in Norwich, where contestants vied to put develop a business idea and put together a new company in just 48 hours. Davis, then still a student, didn’t win. But he did develop a taste for the process. And at a follow up event he saw a company propose a business analytics idea which he thought he could do better.
Coding furiously and “missing about half of my lectures” he had a product ready by February. He applied to UEA’s Enterprise Fund for £50,000 financing, only to be knocked back because his business proposal was aimed squarely at capitalists and didn’t have a groovy “social impact”. It was a make or break moment. Time away from study meant Davis had failed a module and been asked to resit. Instead he decided to quit, and scraped together money to pursue Senlab from friends and family. “One week,” he says, “we were thinking the worst, that we simply wouldn’t make it.”
“People used to laugh in my face, tell me to go back to Mummy”
Over the summer holidays last year, he and a team of fellow students managed to assemble a version of Prosper to demonstrate to potential customers. But by September, when the academic year began, his colleagues returned to their studies and Davis was left alone, Senlab’s sole employee. No matter. In November, this astonishingly fresh-faced entrepreneur (“people used to laugh in my face, tell me to go back to Mummy”) secured a further £350,000 investment, allowing him to recruit. Today Senlab has 15 employees, and has accepted another half-million pounds from a Middle Eastern backer. But if the development of one student’s idea into reality has been astonishingly rapid, it is nothing as astonishing as Davis’s ambition. He aims to have 100 employees by the end of 2018, and reads management books on how to allow his employees to blossom (“my job now is to hire the best people, not to do all the best work”). He now wants Senlab to get as big as possible as fast as possible. “I am in the business,” says this beardless youth, “of empire building.” And who would bet against it happening? This vulnerable hatchling already looks set to fly.
Over the summer holidays last year, he and a team of fellow students managed to assemble a version of Prosper to demonstrate to potential customers. But by September, when the academic year began, his colleagues returned to their studies and Davis was left alone, Senlab’s sole employee. No matter. In November, this astonishingly fresh-faced entrepreneur (“people used to laugh in my face, tell me to go back to Mummy”) secured a further £350,000 investment, allowing him to recruit. Today Senlab has 15 employees, and has accepted another half-million pounds from a Middle Eastern backer. But if the development of one student’s idea into reality has been astonishingly rapid, it is nothing as astonishing as Davis’s ambition. He aims to have 100 employees by the end of 2018, and reads management books on how to allow his employees to blossom (“my job now is to hire the best people, not to do all the best work”). He now wants Senlab to get as big as possible as fast as possible. “I am in the business,” says this beardless youth, “of empire building.” And who would bet against it happening? This vulnerable hatchling already looks set to fly.
If it all works out for Davis, his company will soon move from the realm of “start-up”, to that of “scale-up” – where a proven idea is expanded into a major concern. Scale-ups are defined as companies whose staff or turnover have grown by 20 per cent for the last three years, a criterion which only 1 in 200 meets.
One such is Featurespace, a machine-learning analytics company rather like Senlab, only a little further down the track. Founded in 2008, it raised just £4.5m in investment for the first eight years. But in the last few years it has expanded dramatically, raising £23m. Now, in its offices which overlook Cambridge’s Cavendish laboratory, which played a key part in Britain’s Second World War development of nuclear atomic weapons, it is a shining example of a relationship between university and private company that really works.
Indeed, Featurespace started as a way for its founder, the Professor of Signal Processing, Bill Fitzgerald, to generate profits to fund his PhD students. Before his death in 2014, Fitzgerald was a key figure in statistics, a man whose work on sonic wave data in seabeds spawned the phrase “separating the signal from the noise”, now associated with AI algorithms which see patterns in vast, formless troves of numbers. “Without Bill, machine learning would not be what it is today,” says Martina King, Featurespace CEO.
The company used that machine learning to develop anti-fraud software for the betting platform, Betfair, detecting unusual patterns of behaviour which less sophisticated vetting could not pick up. The predictive technology was so good that when Martina King arrived in 2012, it seemed like Featurespace had an incredibly powerful tool at its disposal, one which could almost see into the future. The question then was what do with it? The possibilities seemed endless.
“We had access to a data set of skin lesions, so we wondered if we could predict the moment that a patient would have to go to the doctor and have it checked out,” says King. “For a few minutes we got very excited that we might be able to do something really important in the world. But at the same time we were running out of cash, so we had to go back to basics. That was the real pressure.”
Today Featurespace has 140 employees, and has returned to its core anti-fraud business. It is expanding again, but this time not in its range of products, but rather in its footprint, launching offices around the world, notably in America. There the Cambridge University connection opens many doors. It also helps Featurespace, providing a conveyor belt of programmers to its door. But for would-be empire builders, the company, and Martina King, provide a salutary lesson: “There are so many people we meet who have a great idea. And they try to build the idea, but it’s decoupled from whether it actually will work commercially or not. You really have to understand the problem in the market that you’re trying to solve and know that there’s a commercial opportunity if you get it right.” King says she has even had to lay down this law to some of her own “young guys” at Featurespace.
“I think commercialisation is a challenge for Britain on the whole,” she says. “Nobody here wants to be in sales. In America it’s a noble profession”
“I think commercialisation is a challenge for Britain on the whole,” she says. “Nobody here wants to be in sales. In America it’s a noble profession and there are huge amounts of training courses. If you look at the MBA program in most UK universities they don’t have negotiation skills, they don’t teach how to win contracts, they don’t have selling as a module. I think that is something really important for us to change, if in Britain we want to monetise our inventions.”
“I think commercialisation is a challenge for Britain on the whole,” she says. “Nobody here wants to be in sales. In America it’s a noble profession and there are huge amounts of training courses. If you look at the MBA program in most UK universities they don’t have negotiation skills, they don’t teach how to win contracts, they don’t have selling as a module. I think that is something really important for us to change, if in Britain we want to monetise our inventions.”
Cambridge, in fact, offers one of history’s best examples of the cost of not doing so. In 1975 César Milstein developed a technique, with Georges Kohler, for producing monoclonal antibodies, in the city’s MRC Laboratory of Molecular Biology. Milstein deliberately did not patent his discoveries, nobly believing they should be for the benefit of mankind. But if he did not profit financially (he did win the Nobel), others certainly did. A little more than three decades later, six of the top ten best selling drugs in the world were monoclonal antibody therapies, an annual market worth $55bn.
Some 20 years later, in 1997, over drinks at the Panton Arms pub in Cambridge, four chemists developed an idea of how to sequence DNA – effectively a way of transcribing the recipe of all life. Within a decade, they had sold their company, Solexa, to the American rival Illumina for $650 million. At the time a press release described it as “one of the greatest commercialisation success stories to emerge from the University of Cambridge”. But was it? Illumina is now worth more that $35bn, all based, as one green-eyed entrepreneur told me, “on British technology”.
“The Brits are the great inventors, and the Yanks are the guys who are great at putting it in a box and selling it,” John Milton, who played a critical role at Solexa, said after the takeover.
Learning both to invent and to sell, and then resisting the temptation to take the money and run when commercial suitors come calling, is the key to building billion-dollar businesses. And it is Britain’s billion dollar businesses, known as unicorns, which are the focus of the next chapter in this series.
No comments:
Post a Comment