I love Google Translate.
It was in August 2016, in Rio de Janiero, when our eyes first truly met. I picked up Google Translate on my iPhone at the 2016 Olympics, and it turned out to be a lifesaver. Alone in the city, it allowed me to get around the city without having to depend on anyone for help or wave my arms around helplessly. At Rio 2016, all the infrastructure and transport links were new, completed just days before the event. The venues were spread out, separated by vast distances in remote parts of the city. While the infrastructure in Rio was fantastic, the volunteers lacked the information beyond their immediate zones of operation. And the taxi drivers, a tourist’s fount of all knowledge worldwide, had very little spoken English. Google Translate solved all the communication barriers.
Google Maps was a reliable guide to public transport links and travel schedules. When public transport links (with walking distance, travel times and changeovers) were inconvenient, low-cost Uber stepped in to make the journey simpler.
But in December 2016, I was jilted in Tokyo. Although electronic language translators have abounded in Japan for a long time, tourists and locals were hardly using them in December 2016. You would normally only see them in more formal settings such as offices.
Fast forward one year however, and everything has changed.
It’s December 2017 in Nagasaki, Japan. On a holiday trip to the Shinkamigoto remote set of islands, 100km off Nagasaki, our taxi “concierge” host hands us a Huawei D-Tab tablet with a VoiceTra app to let us talk to our Japanese speaking driver for our sightseeing trip. By the end of the 3-hour tour, both the driver and I were using Google Translate text to speech natural language voice rendition on our phones. The Huawei D-Tab tablet was cast aside, no longer required. It was like swimming with the Babel-fish from the Hitchhiker’s guide to the Galaxy (by Douglas Adams, as any 80s kids out there will surely remember). Later that evening at our Ryokan hotel, both the staff and I again resorted to our phones to negotiate our late checkout, use additional hotel facilities and book our return taxi.
Although I speak a little (survival level) Japanese, I have been travelling to and in Japan every year since 2007, and it’s only in 2017 that I noticed the widespread use of these translation applications, even in the remotest parts of the country.
Google Translate as “an innovation” has become mainstream.
A project that started in 2001, in 2013 Google Translate served 200m people per day By 2016, it was serving 500m.
What has their journey to tell us about how innovation and entrepreneurial ventures succeed in the near future economy? I am first going to describe what these journeys now look like, and then get into what to look for as an investor, entrepreneur, in company innovator, corporate finance officer, or just interested in how the next big thing develops.
Some innovations get adopted and make the mainstream. But even for successfully commercialised products i.e. the ones that we are likely to see in the market, the failure rate can be as high as 90% . And sometimes success can take decades…
You may be familiar with the ‘S curve’ that follows the diffusion of innovation. This isn’t the right model for this type of new technology though. Adoption doesn’t slow down after a while, it accelerates, and looks more like this.
Salim Ismail in the introduction chapter of his book Exponential Organisations, writes “we are experiencing a new breed of organisation that is scaling and generating value at a pace never before seen in business”. The chart below shows the accelerating metabolism of the economy. The time taken to reach a billion-dollar valuation has decreased EIGHTFOLD in the last 20 years. The graph below completes the picture.
This level of accelerating growth is only possible if the products and services these companies launch are adopted at an accelerating rate. A corollary to the rate of adoption is the reduction of price, (discussed later) and sure enough costs of some of these technologies have plummeted at an accelerating rate. For example
• 3D Printer costs from $ 40,000 (2007) to $ 100 (2014) Scale: 400x in 7 years;
• Industrial robots cost: $ 500,000 (2008) to $ 22,000 (2013) Scale: 23x in 5 years
• Drones cost: $ 100,000 (2007) to $ 700 (2013) Scale: 142x in 6 years
• Solar cost: $ 30 per kWh (1984) to $ 0.16 per kWh (2014) Scale: 200x in 20 years.
“The diffusion and adoption rates for new technologies have risen over the years. The graph shows the number of years it took technologies such as electricity, television, and the Internet to be adopted by at least 25 percent of the U.S. population.” As shown in the graph below from the US census, wall street journal, reproduced by Marketrealist and the Startup Way by Eric Reis.
Faster and ever faster
For innovations that do become mainstream, the time for this adoption has been dropping, because companies can scale much faster than ever they could. as the scaling capacity of companies is no longer constrained by having to create and own the means of production. Entire value chains can be rented on demand. Global communications, 3D printing, elastic value chains, scalable cloud technologies and AI driven insights make all of this possible.
The term used to describe the complexity of our environments, which impacts our ability to predict the future reliably (or at a minimum our success in it) is VUCA. VUCA stands for volatility uncertainty complexity and ambiguity and is used widely in strategy to set the context for scenario planning. Because we can’t predict the future and are less and less able to day by day as the world becomes more complex, business strategists work on many possible future scenarios as once as a set of experiments. Making one big bet on one possible future just means you are guaranteed to be wrong.
This new technology ecosystem in this complex environment creates opportunities for well-funded start-ups and incumbent firms to innovate and create new services. In the beginning, a large number of firms (with their propositions) enter the market, and in this primordial soup, Darwinian evolution takes over. Only a few survive to create a dominant design. This phenomenon is not new; it happened in the 1900s in the automobile industry where at the peak in 1910, 275 automakers were producing and competing . Over the next 20 years there was a massive exit by failed firms and the big three emerged, getting up to 88% market share by the 70’s. What is new however is the speed of this selection survival cycle. It is like an Elon Musk rocket.
Product, pricing and product ecosystems
Google Translate is free. As a tourist, with the use of offline dictionaries, one does not need roaming data or Wi-Fi to be able to use the service. It doesn’t cost to download, it doesn’t cost to update, and it doesn’t force any other additional or hidden costs on you.
Google translate as a product may not make money, its company Google (or parent Alphabet) makes truckloads, which brings us to our next point on ecosystems.
An innovation is an invention (or doing something different), creating value (leading to adoption and scale) and a means of commercialising the success. Companies don’t survive by single products and services alone; they thrive on product and service ecosystems, where some free products are necessary to create the hook for paid services. The Google AdWords business would not have been the success it is today if Google search was a paid-for service. That is why Google search, with all its power of bringing the world’s information to your fingertips, is free to use. We see this practice in other industries. In education, business schools provide free online content. In consulting, the good companies publish reports for free, some better ones even share their “IP” .
So, what of the other businesses that won’t survive the next few years?
Obsolescence. Cigar butt businesses
A Cigar butt business was a term coined by Warren Buffet as he describes businesses that have just one puff left. This is probably why his investments go to companies that he believes will exist for another 100 years. It is also why some of my close friends will lose large sums of money they have invested in AI-driven eLearning systems, where the number of survivors that eventually become dominant from the present primordial soup of intense competition will be a single digit fraction.
Part 2: So what can we learn from this?
How can we evaluate a new opportunity or an existing business idea? How are early stage investors guided?
Innovation reinvented and adoption: We need to get beyond the mind-set that R&D or invention = innovation. Innovation is doing things differently that creates value, it’s the commercialisation, scale and adoption. Without adoption, it is a failed innovation, and there is a palette of innovation opportunity beyond product innovation (or it’s narrower definition of disruptive innovation). The type of “innovation” a firm can deploy should vary based on whether it’s an incumbent or an entrant and whether the market is developing or mature. Work by Costas Markides and Geoffrey Moore in Inside the Tornado expands the palette of these innovation options. Each has a significant impact on adoption.
Acceleration: The markets are driving the Darwinian process of tinkering, selection, adoption and survival of the fitter few. A large number of new firms enter these new market opportunities (AI-driven recruitment and candidate selection is a current example of the number of companies with offerings), and only a few remain as a dominant design, technology or player emerges. Firms wired for exponential growth are structurally different. They leverage the power of the crowd, algorithms, platforms, automation and AI at unprecedented levels. The strategic assets that companies have used in the past to protect their competitive advantage are the very assets that are becoming liabilities. New competition is increasingly asset light and uses collaboration and external ecosystems to compete.
Salim Ismail and his team conducted a post facto analysis of the structural characteristics of firms who grow at 10x the industry average and developed a series of questions. An adapted version of a firm’s future potential for exponential growth is available here.
Product advocacy and virality is a significant driver. I see it happening with Zoom, a web conferencing alternative to WebEx that is taking the world by storm and we have used it successfully with our clients and partners in Myanmar, India, Sri Lanka, and Oman (where most Voice over IP technologies are banned). We are big fans and have introduced it a dozen organisations including the World Bank.
Being saved from Obsolescence: A good model is a seven-domain framework developed by John Mullins, in his book The New Business Road Test. We have used this model widely in our international faculty development programmes when partnering with LBS and NUS faculty. The model not only looks at the opportunity with its current potential but also asks vital questions about the business’s sustainable advantage from IP, the creation of strategic assets, on-going innovation activities and process. It’s a pre-emptive look at the future competition you can’t spot today. Perhaps the most critical aspect is the evaluation of underlying economic models that will give the business the oxygen to reinvent itself over years to come. A healthier business has a higher chance to reinvest in innovation, experiment and create new markets. The latter is covered in Getting to Plan B by John Mullins. You can find links to resources including downloadable book chapters at the end of this piece.
It is no wonder, why even ten years ago the overarching question on the future sustainability of the venture seven domain framework was “Why won’t Google eat you?”. It was valid the and still valid today. That’s where the title of this blog came from.
Freakonomics: Our prediction for the future.
Most experts are wrong at predicting the future . McKinsey had famously estimated that market for mobile phones for AT&T was only going to be 900,000 units (AT&T subsequently lost out on a billion-dollar opportunity by not entering the mobile telecommunications space). And on the other hand, even a stopped clock is correct twice a day. Here is a small risk I will take in making this prediction. It is based on the work of Steven Levitt , an American economist known for his work in the field of crime, in particular on the link between legalised abortion and crime rates. His hypothesis (Donohue-Levitt hypothesis) is a hypothesised reduction in crime as a result of the legalisation of abortion (that happened a few decades previously) and not as a result of government intervention on managing crime in the present day. Even an economist of Steven’s stature (voted in the top 4 living economists under the age of 60) has had his work come under criticism that his study was flawed .
Here is my prediction. Japan will see a massive surge of tourism travel in the next 3-4 years. The administration will point to the Tokyo 2020 Olympics and associated promotional activity. I predict the not only a surge in tourism but also an entirely different enabling factor. The ubiquitous, free usage of Google Translate. Perhaps the only way to test that out is to observable adoption and use over the next five years.
By Viren Lall, Managing Director, ChangeSchool