Back in the mid 1990s when Sky TV was in its relative infancy a clear emblem of middle class achievement in UK homes was the Sky set-top box in a living room. The single biggest target of Sky executives at the time was to own the piece...
After years of preparation, discussion, scandals and graft the Rio Olympics are now consigned to the history books. For me, the coverage has been inspirational, with glimpses into the real nature of what it takes to be 'Olympic standard’. Aside of the glories of the GB position in the medal table there are moments where you understand the sacrifice and sheer bloody-mindedness that are required to even be there.
Mary Meeker’s annual report on Internet Trends always includes a slide on “time spent in media” versus “proportion of spend”, and the “laggard" that is mobile ad spend in this context (see pic). This is knowingly an incomplete approach to attribution modelling, as ultimately time spent doesn’t entirely correlate to “sales made”. At the end of the day as metrics go this is the only one that matters.
The context, receptiveness, timing, creative capabilities of each media type play their own critical parts in the sales and marketing funnel. A sum of the parts analysis is insufficient. If Jaguar sells a car and attributes the sales value to Google as the last referral point to the location of the Jaguar showroom and reallocates their ad budgets to the detriment of TV and Out-of-Home advertising “because the model says so" they will sell less cars going forward. Yet a search marketer will make that argument. Attribution is a complex quasi-science, and there is a requisite level of cynicism required to any "econometric” modelling of advertising performance.
As a tech company who entered the media market, one serious critique is of the structural silos that promote particular channels as opposed to overall campaign effectiveness (which is what the end customer cares about). Our world is outdoor media. Attribution has been, and remains a major challenge. In the UK, Route, the standard for audience measurement has helped take some big steps forward.
The future of attribution may well lie in the use of aggregated mobile-location data, where demographic movements in the real world are mapped in real time and digital screens serve content to crowds based on real time inputs, but you need some heavy tech infrastructure to power that data at scale.
In the short run the major benefit our technology provides is in automating the processes of buying and selling this media, freeing up smart people to present scientifically defensible reasons to spend their precious resources on advertising.
Somewhere in the region of $30bn of advertising accounts are up for review currently, an unprecedented amount. I’m sure the main question CFOs and CMOs are asking as part of these reviews is why do we do this? Automation tech frees up agencies to engage with answering this question; the best answers will be transparent, honest and not silo-driven.
In 1953 an IBM salesman boarded an American Airlines flight from Los Angeles to New York. In the pre business class days, Blair Smith sat at the back of the plane, where he found himself seated next to a dishevelled unshaven man, who over the next couple of hours would establish himself as both a master conversationalist and the current President of American Airlines, C.R. Smith.
When the two arrived in New York the seeds had been sown for a 6 year R&D project between IBM and American Airlines called SABRE (or its less catchy name: "Semi Automated Business Research Environment") a system which ultimately reinvented the way we travel. As Wired magazine also points out, “it shook up computer science too”. Today in the world of self service, online reservations, SABRE is still used to book flights on hundreds of airlines. 50 years on there are lessons for other industries, not least the the outdoor advertising industry where BitPoster is focused, and where many of the pre SABRE practices continue.
Before SABRE, American Airlines reservationists kept track of travellers in much the same way that restaurant waiters keep track of dinner orders. The used ‘lazy susans’ like the one shown below.
This was a clumsy, ineffective system that was costing American Airlines money. Sales cost per ticket were very high, demand based pricing was impossible and there was significant 'inventory' wastage - airplanes were flying with empty seats and about 80% of them were because of bookkeeping errors from paper-based processing. With unsold outdoor media often running at 50%+, and the lazy susan equivalent of telephone, email and excel spreadsheet the weapons of current choice for the market, the scope to "SABRE-ise" is clear: It's shocking in fact that it is 2015 and it hasn't yet happened in every applicable industry.
SABRE’s first system was made up of two IBM mainframe computers, based in New York. 8 years after the 2 Smiths had shared a flight together the 1st live system connected 1,500 terminals across the US and Canada. The hardware alone cost about $30m, or around $250m in today’s money. The hardware order from American was at the time the biggest ever received by IBM.
Being able to have instant updates to its seat inventory and passenger information gave American a huge competitive advantage - they could price and sell in real time through a distributed network. American’s reservationists connected remotely via custom terminals to the IBM mainframes in New York, with data exchanged via an IBM created "frequency modulation data transmission” that worked over AT&T’s telephone lines. There really was a time ‘pre-internet’.
By 1965 Sabre was handling 7500 reservations per hour; which given that it took American 90 minutes to process each reservation under the lazy susan system, this was an epic efficiency improvement. In our market the current average processing time for transactions is 55 man hours with the bottlenecks around manual checking of availability and matching advertiser briefs to that inventory.
The market consequences are the same - human errors, undersold inventory and limited dynamic pricing. As the network of outdoor screens digitises and the scope develops for literally infinite numbers of bookings the lost opportunity also grows exponentially. BitPoster is taking the lessons from Sabre to solve this problem, and we're helped by where the tech ecosystem is at versus when C.R. and Blair Smith started talking.
The entry ticket price for SABRE was $30m ($250m in todays money) on hardware alone, covering the mainframe computers, the terminals, the AT&T line access and the engineering, all of which was bespoke, closed loop and requiring highly specialised operators.
In 2015, cloud based servers that are distributed, shared and standardised for access from any device, combined with open source software that is flexible, scalable and very cheap mean the creation of a SABRE-like system comes with a tiny entry ticket by comparison, and ongoing operations requirements can be made for customer self-service as the likes of Expedia, Opodo and BA.com all provide.
One of Ireland’s great business success stories of the past 50 years has to be a company called Glen Dimplex. The company began life manufacturing oil-filled radiators, in Northern Ireland, with seven employees. 40 years later the company is the a global powerhouse in manufacturing domestic appliances (like kettles and radiators), cooling and ventilation. Glen Dimplex is also a world leader in high tech intelligent electric heating and renewable energy solutions, but at it’s core the business dominates markets where perceived scope for innovation is lost on the masses. How do you make a better kettle? Surely kettle innovation peaked sometime early in the life of Glen Dimplex? I’ve often asked those who might know how a company like this can be so successful. “It’s a big world out there and everybody wants a kettle”.
Many start-up businesses focus on solving the complex, apparently unobtainable solutions to abstract problems. How else could you have a competitive advantage against the status quo? Many of these start-ups are before their time and never get traction or discover product/ market fit. Often overlooked are the opportunities to do something that already gets done significantly better. Like manufacturing kettles better than anyone else.
A simple and i think great example of an efficiency improving start-up is Naked Wines. They sell wine. Not new. Their model is different though. They have signed up 200,000 “angels” who commit £20 per month to their wine buying kitty. In return angels get 25-50% off the retail price of wine on every order. The £4m monthly working capital commitment from the angels solves a financing problem, and is invested in independent wine makers who give their wines to Naked at wholesale prices, which they pass back to their angels. The independent wine makers get solid volume order commitments, solving their cash-flow concerns, Naked has negative working capital as they have £4m each month to buy stock with, and the angels benefit from the wholesale prices. The social element to the site makes it feel like a cheap entry wine club. And the wine is good.
Making better kettles and improving the cash-flow for independent wine makers are not new ideas. They are simple ideas executed to perfection. There will always be a lot of value in that.
The FT recently produced a calculator on its online site that calculates the value of your individual personal data to marketers. The average person’s data often retails for less than a dollar. I am clearly an average person, with the my data worth the princely sum of $0.4639. General information about a person, such as age, gender, location is worth a minimal amount. More specific information, such as if you are shopping for a car, a house or travel is slightly more, and milestone events such as having a first child (which prompts a dramatic change in buying patterns) drives up the value significantly.
I love the story that I heard in a marketing course at business school about Tesco’s monitoring of post birth shopping patterns. Their ClubCard data appeared to pickup a high correlation between nappies being added to a shopping basket and beer sales. As it turned out in the weeks post childbirth it is often the male who is dispatched to the shops, which results in an element of randomness to shopping baskets. The increase in beer sales was a temporary phenomenon as after a few weeks the shopper reverted to her former self. What was apparent to the shopper marketing folk at Tesco was that as beer is bought quite generically most of the time…i.e. it’s always Carlsberg, or always Heineken (as to be quite honest as long as there is beer in the fridge at home, men tend not to be that picky), there was a major opportunity for brand switching during the 2-3 period after nappies started appearing in shopping baskets. Apparently, the smart folks who sell beer in Tesco started running beer promotions in the nappy aisle, resulting in switching patterns that were of significant value to the beer retailers who’s product they were trafficking.
The opportunity that “big data” presents to marketers is profound, and there is clearly a fine line between personalising a marketing approach and breaching common decency and privacy laws. The opportunity insightful data mining presents is in mass personalisation - customising advertising to limit the waste of spend and to maximise the likelihood of encouraging certain behaviours.
Google manages this at scale - if you search for “hairdressers in Clapham” it probably means that you are interested in the services of a hairdresser in Clapham. The search itself pre-selects the searcher into a category of buyer whos likelihood of purchase is exponentially higher. That is incredibly valuable, as Google’s market cap would attest to. As the FT calculator shows, you can acquire this data from 3rd party sources and often relatively cheaply – but Google owns the bazooka; both the data generator and the advertising platform.
Paul Graham, the essayist, venture capitalist and founder of Y-Combinator wrote in 2009 about whether it was possible to "buy a silicon valley". Clearly if you could "buy a Silicon Valley" most countries would write the cheque for the economic results that follow - tax revenue, job creation, inward investment, perpetual innovation etc. As Graham points out - a lot of cities look at Silicon Valley and ask "How could we make something like that happen here?".
The organic way to do it is to establish a 1st rate university in a place with a lot of rich people, which is how Silicon Valley came to be. Clearly that in itself is not sufficient - there are particular cultural characteristics that are unique to Silicon Valley that make that world work - hippy, laissez-faire attitudes fostered in San Francisco in the 60s and 70s are part of the story.
While it's possible to create an environment that fosters entrepreneurship, the real goal is to tie the culture of entrepreneurship to a particular place, so that the economic benefits are captured by that place. What ties founders to Silicon Valley for the long-term is often different to what originally fostered the entrepreneurial mindset. As Graham points out encouraging startups is different from encouraging startups in a particular city. "The latter is much more expensive".
This week in the UK a report from the Centre for Entrepreneurs and Duedil, showed that 1 in 7 companies in the UK is founded by an immigrant, with 14% of the workforce employed by companies started by immigrant founders. 17.4% of migrants start a business versus an average of 10.4% for UK nationals. Top of the pile of immigrant founders, are the Irish with a staggering 48,854 founders of UK companies registered as Irish by Companies House. I am one of those 48,854. Not so unique anymore.
Most of these 48,854 founders run SMEs. In the real world (as opposed to the popular press) start-ups don't typically reach the multinational status of the iconic tech names that have emerged from Silicon Valley. Most are small to medium sized local businesses. Nonetheless, these businesses are the lifeblood of the economy (40% of the value of the economy in the UK and significantly higher in other countries).
A major concern for Ireland from Duedil report is that the culture of entrepreneurship clearly exists, but swaths of the value of the entrepreneurial culture are being captured elsewhere. The report just covers the UK and not other popular spots for the Irish diaspora. 48,854 Irish founders in the UK is a staggering number in the context where the total number of registered businesses actually in Ireland is around 190,000.
The Irish economy is recovering from a truly awful economic crisis, but getting a percentage of the pioneering foreign founders to build their business "at home" has to be key to safeguarding the future. (The next blog might be on why I am in London as opposed to Dublin)
We moved the business into new digs recently, kindly provided by one of our investors on a good deal. The office move made me think about the difference between running our business now and what it would have been like even 10-15 years ago. Our team is made up of 5 people in London and 7 developers in New Delhi. The London move involved shifting a few Apple macs and a few bodies. A single London taxi was sufficient to move home from Shoreditch to Bloomsbury. The move took about an hour, including packing. We plugged our Macs in at Gresse street, typed in WIFI access codes, installed a printer driver and the business was back to work. The last bit wasn't even necessary.
The assets of our business were instantaneously accessible in the cloud - the Amazon web servers that host our databases that our developers work with are probably located somewhere near the Arctic circle amongst the polar bears to take advantage of the natural cooling. They are certainly physically a long way away from London or New Delhi. The communication between the London team and the New Delhi team runs via daily Google hangout or Slack "stand-ups" and a several cloud based project management and messaging apps, accessible anywhere with an internet connection. We store our knowledge in a cloud based Wiki, and use Dropbox and Google Drive to share and store simple information. We run the management accounts in Xero, where paperless invoices are stored and are callable to generate reports in a few clicks. All of our business systems are off-the-shelf software as a service applications that we pay a few hundred pounds a month for in total.
In fact, the only key physical assets beyond the humans (the computer hardware itself isn't critical), are the Nespresso coffee machine and our fridge that holds the Thursday team beers. If Warren Buffett was looking at us (which I don't believe has happened yet) on the face of it he might be hard pressed to identify what our "moat" is; what's the protective insulating layer that keeps our competitors at bay.
15 years ago when Brent Hoberman and Martha Lane Fox were setting up lastminute.com, they needed £6m of investment to get their platform to a point of reasonable viability. Much of the cost was driven by the need to build each piece of the product from scratch - there were limited off-the-shelf bolt-on features that could be employed. Not knowing what user traffic they would get they had to physically install substantial server capacity in their office to protect against a situation where they were popular. The office was probably quite warm as a result. The entry ticket was the £6m of venture investment. That in itself was a sufficient "moat" to keep many would-be competitors at bay. Today the same platform technology could be put together from a number of generic pieces of SaaS (Software-as-a-Service) technologies, at a fraction of the cost for lastminute. Their "moat" is their partnerships, their database and their customer loyalty, not their technology.
So the barriers to starting have all but disappeared. What's critical to our success is the speed at which we can unlock key relationships with customers, investors and data providers, our ability to interpret their needs and to turn them into "product" that solve big problems quickly. Last week I wrote about the difference between fostering entrepreneurship and fostering entrepreneurship in a particular place. As an Irish founder, based in London, could we run our office from Dublin and pay our taxes there? Of course - the plug sockets are the same - and our Macs would be WIFI connected just as quickly.
So what's key about being in London?
We're building a media tech business. London is a global leader in digital and creative media. The key relationships for our business are all within 25mins walk. Like it or not, the financial services industry has created a lot of discretionary wealth that allows money to flow towards early stage ventures. The enterprise investment scheme tax advantages make the money flow with less risk. Serendipitous meetings can happen because we are close-by. London is a melting-pot of people - our London team is made up of 5 people with 5 different nationalities; French, US, British, New Zealand, Irish. It's ironic that our business truly could be anywhere with an internet connection, but the importance of having a physical presence at the center of the ecosystem that allows us to join the dots quickly is critical.
The Facebook acquisition of WhatsApp last week for $19bn shows just what a disruptive world we are in. 4 and a bit years, just over $8m of capital, a handful of people and business of truly global reach is the result. WhatsApp had 450m users on the announcement date, and has added 15m more in the week since - they are on target for 1bn users by early 2016.
Exponential is a hard concept for the brain to grasp…the stadium analogy builds the picture. Imagine you are in the Nou Camp in Barcelona sitting up "in the gods". Lionel Messi on his day off is standing in the center spot and has a pipette of water…he puts one drop of water on the grass and walks off. Being Messi, this drop of water is magic, and it's going to grow exponentially - doubling every minute. How long do you have before the stadium is completely full and even you up in row Z are engulfed? Human thinking tends to be linear and wildly over-estimate the time…days, weeks or months.
The answer is 49 minutes.
What's really surprising and hardest to fathom is that at 45 minutes the stadium is 93% empty. That's 4 minutes from an apparently safe situation to a flooded stadium.
Exponential is WhatsApp - from a standing start a few years ago to carrying more SMS messages than the entire SMS of the telecoms world combined. The network connected world is exponential. Moores law is exponential. Exponential is why leaders in industries such as music, publishing, retail, travel and advertising constantly get it wrong - they're aware of the danger to their businesses, but aren't aware that they are in the last 4 minutes.
If you are an investor in global multibillion dollar businesses like the telecoms companies that WhatsApp disrupts, then risks are higher than ever, and in many cases underestimated. The McKinsey topple rate - the % probability that any industry leader will lose their market leadership position in any year - is at its highest ever, and is 40% higher since the 1980s. Even for Facebook, who have a market cap of $180bn - how can a P/E ratio of more than 100x be justified if its business is already so at risk that it is obliged to pay $19bn for a scrappy start-up that has beaten them in the mobile messaging race.
It’s a great time to be a tech start up and a scary time to be big business with hardwired processes.