From an idea scrawled on the back of a pub receipt, to the fastest-growing tech start-up in the UK, now with ambitions of being listed on the Nasdaq, Black Swan – the business that predicts the future to help others boost revenues – has truly spread its wings. Director meets chief executive Steve King to learn how the company is now using its capabilities to tackle diseases like Alzheimer’s
Inside a picture frame on the fourth-floor wall of an office next door to London’s Waterloo station is a torn receipt from a bar in Toronto. Scrawled on the back of that receipt is an idea hatched together over several pints by two Brits that would eventually become Black Swan. Within four years of its launch in 2011, Black Swan was named the UK’s fastest-growing tech start-up. Last year, Wired hailed it as the company that “can predict the future”. It all looks mightily impressive on paper. But is it just another case of the emperor’s new tech or a real harbinger of the future?
Black Swan’s founders are chief executive Steve King and chief strategy officer Hugo Amos. King, who had a job creating websites and mobile apps, was working on a campaign for PepsiCo in Canada when he met Amos, a marketing manager for the firm. After that extraordinarily productive night at the bar in Toronto, King sold his business and eventually persuaded Amos to come on board and back to Britain. Speaking
to Director, King explains how the business works: “Essentially, Black Swan is a prediction company. We deal in giving people a better view of what is going to happen in the future.
“The way we do that is by analysing information on the internet. That can be open data, social data, government data, weather data… We bring that to a company and we overlay that information with their own data. So that might be information on what they are selling on a weekend or a certain day of the week. And then you can see the patterns of how things behave in the real world and how they reflect on things like sales. And once you’ve created a pattern you can go back three, or even five, years and you create more patterns and more comparisons of internet behaviour against what happened in real life.”
By bringing all this data together, King says, “we are able to make predictions and feed back to the companies and show them things to help them to innovate, or just to understand what people were having conversations about in the past and what they will be having conversations about in six months’ time. Or we can look at things in the short term. A simple example would be: this Wednesday, the forecast says the temperature will increase by several degrees, which makes it a really good day to stock and sell more ice cream.”
Black Swan also uses language-processing tools to deconstruct and make sense of all the different conversations that could be happening online around a particular product or event, allowing them to build up a picture of what is likely to happen next. Sticking with ice cream, this could mean two people talking about it. One might be highly educated, the other with a more limited vocabulary, but they are both telling you they love ice cream.
What Black Swan looks for, however, are the moments we haven’t anticipated rather than ones confirming what we already know. Or, to paraphrase the former US secretary of defence, Donald Rumsfeld, Black Swan is looking for ‘the unknown knowns’. It also sheds some light on the name behind the business.
King explains: “It dates back to 16th-century London when people used to say ‘that’s about as likely as a black swan’. All they had ever seen were white swans. That was the only information they had. Then an expedition sailed from Europe [in 1696, led by Dutch sea captain Willem de Vlamingh] to Australia and that was the first time Europeans saw a black swan. We really liked that as a concept – to see things differently by seeking them out through data.” This notion was perfectly illustrated in the company’s fledgling days, when King and Amos were trying to convince businesses that they could indeed predict trends.
King laughs as he recalls, somewhat ruefully: “I had a bet with my counterpart at Disney that we would prove that getting people to go to the cinema was easier when it was raining. If we could prove the correlation between rain and going to the cinema then we could automatically change ad servers to put out more ads for their films. I was so confident that we were right I said, ‘If I’m wrong, we’ll give you your money back that it would cost you to do this.’ So we did all the work, we bought all the data – remember, we were still a really small company, we couldn’t afford to pay ourselves properly – and we were wrong. There was no correlation whatsoever. Yeah, well done, Steve!
“It’s actually low pressure that drives people to the cinema, when you get three or four days of cloudy weather. You can see a bell curve starting on that first day. They want to get out, but because the weather isn’t great they decide to do something indoors – but eventually they will go out. Thankfully Disney paid us anyway even though they didn’t have to. But that, for me, was a breakthrough moment because it made us realise that data isn’t always going to prove your point and that’s what we look for.”
King says: “I am constantly surprised by what can be done with data. I used to get really upset with how people weren’t using it properly. Age and experience has taught me that most companies are busy focusing on what they do well. If it’s your business to make crisps, you make great crisps. You don’t necessarily think about that data that could be important to boost your sales five per cent higher. To me, that was an important reassessment of what we do; namely, we make value out of data and we do that really well, but we shouldn’t be surprised that people or businesses don’t do that, because they are better at their core business.”
In the case of Tesco it meant predicting what would be the first big weekend of the year for barbecues. It was calculated that a 10°C rise in temperature, as occurs in spring, would lead to Tesco selling three times as much meat. Posts on social media along with historical trends supported this. Successfully predicting the peaks and troughs for barbecue season also means that a supermarket won’t be left with a lot of unsold stock that will simply go to waste. “One of the things we are proud of is the fact that optimisation isn’t always just about saving money,” King says. “That’s nice, of course. But if you’re not wasting product you are using your supply chain more efficiently and not damaging the environment with unnecessary waste. I think that’s one of the more important things that we do.”
This may sound like another bunch of techies and corporations profiting from the public’s information. We’re creating the data for them, after all. But King is also justifiably proud of a spin-off project known as White Swan, for which Black Swan staff provide free data services in their spare time to not-for-profit organisations such as the NHS. Indeed, it arguably stands out as his most important achievement: “The inspiration for this was my own sister. Around 10 years ago she began to get very sick. A few years ago she became seriously ill. But we just couldn’t get a diagnosis.
“So we used some of the technology we use at Black Swan to analyse blogs and posts, discussing similar symptoms. That helped us shortlist three conditions she had never been tested for previously. It turns out she had an early onset of a form of Parkinson’s known as dystonia. She was able to get the right medication and in a couple of months she will be running a triathlon. So we made a pledge to ourselves that we would help others by using our technology.”
He is hopeful that White Swan will be given charitable status in the summer and eventually hopes to create a foundation that “can really help not-for-profits use that technology to make a meaningful difference”. He is also focused on what he hopes will be the next great leap forward for the business and for data analysis. “Most of our work at the moment is really centred around one thing. We recently managed to prove our theory of trendspotting, to predict a big trend before it happens. This could give a business six months’ to two years’ advantage over its competitors.
“We only got the results of that analysis in April. Through the summer we will be working on how we can use technology to look further forward than we could ever have thought possible. So it’s noses to the grindstone!” But what does that mean in reality? “We went back over the past five years and the results of our analysis showed we would have predicted something like quinoa becoming a popular healthfood ingredient. So now we’re looking at what ingredients could make up the food products of the future.
“We’re also looking at in-flight entertainment – the way that people travel is changing – as well as digital products, such as mobile phones, and the way we think they are going to be used in future. There are individual pieces of artificial intelligence (AI) functions that will allow us to do things a bit better. So, for example, it may be able to detect a pattern in a weather system. We may be able to see the start of a trend a lot earlier than before by using AI techniques rather than mathematical ones. But we’re not going to use it to build the robots that will take over the world! With White Swan we want to do more research into Alzheimer’s and back problems.”
Anywhere and everywhere
The rapid expansion of Black Swan means it now has offices in four continents. In the UK, in addition to its Waterloo HQ, there are small teams working in Bristol, Exeter, Manchester and Reading. King went to university in Exeter and feels an affiliation to the city; White Swan has collaborated with his alma mater on developing internet-delivered preventative therapies. He says creating these smaller units simply reflects the way that work is changing. Black Swan can source the best talent and bring the work to them rather than having them move to London. “I can’t claim this was the product of some amazing strategy. The technology and the systems people are using to communicate have really modernised in the past few years. Systems such as Slack and even Skype have come on so far that it really doesn’t make much difference where you are.”
Looking to the future, King sees the US as the land of opportunity for Black Swan: “We would like to start our listing on the Nasdaq in 2020. We feel there is space there for a British company helping other businesses that hasn’t been filled yet. So that’s very much our ambition.”
Reflecting on Black Swan’s seemingly inexorable rise, King says: “We’ve seen a lot more acceptance. A lot of work in the first couple of years was in persuading people that you could use the internet as a data source. Now it’s a lot easier to have these conversations.” King may have lost that wager with Disney but you won’t find too many people betting against Black Swan creating its own fairy-tale.
Black Swan vital stats
HQ The Tower Building, Waterloo, London
Turnover £12.5m in 2016
Major clients Disney, Tesco, Samsung, PepsiCo, UKTV
Did you know? Steve King has a music studio and used to be a nightclub DJ