Over the last 200 years there have been significant changes in technology and the pace of change over the last 10-20 years is now accelerating
The industrial revolution began around 1760, a little over 200 years ago with the commencement of the use of steam for industrial processes.
The electric motor was invented 200 years ago, electric light a little less than 150 years ago, radio 100 years ago, the transistor just over 50 years ago, and the personal computer less than 40 years ago. The industrial revolution caused major disruption, with large increases in industrial and food production, as well as rapidly changing labour markets as society moved from an agrarian centered lifestyle. The invention of personnel computer and the internet less than 40 years ago have initiated the next industrial revolution.
Figure 1a shows the last 4,000 years of history with the X-axis timescale in log scale. Figure 1b shows the last 200 years with the X-axis timescale set normally.
Over the last 10-20 year we have seen rapid technological and business environment change. These changes have created many new businesses and market, as well as opportunities for traditional businesses. But, they have also led to the disruption of old markets and businesses, many of which either no longer exist or are a shadow of their former self.
Many factors contribute to the advancements over the last 20 years including the globalisation of business, the rise of China and now India. But arguably two of the most important changes have been:
- the ongoing performance improvement, miniaturization and cost reduction of micro-processors; and
- the improvement of communications and interconnections brought about by the internet
Both of these factors are covered by “laws”, namely Moore’s law for processors and Metcalf’s law for network effects.
Moore’s law is the observation that the number of transistors on an integrated circuit doubles approximately every two years.
What this has mean’t in reality is an ongoing reduction in micro-processor size and power consumption, and increase in microprocessor performance. This trend, which commenced around 1970, has continued to date. At the time of writing, 2016, that is 46 years or 23 doubling periods. Doesn’t sound like a lot until you realise that doubling something 23 times (e.g. 2^23) results in a 16,777,216 times increase.
While recent improvements in processor performance are starting to fall behind Moore’s law for computer processors, they are still improving. And, new technologies for massive parallel computations such as graphical processing units (GPU) are picking up the slack so that Moore’s law is effectively continuing.
Figures 2a and 2b show this growth in logarithmic and normal scale
Metcalf’s law, or the network effect, is the observation that the value of a network is proportional to the square of the number of connections.
If only two people have telephones, they can only make one connection. Five people with telephones can make 10 connections, 12 telephones can make 66 connections. But, the cost of increasing connections is largely proportional to the number or connections, or less. Ignoring economies of scale, and higher costs associated with incremental infrastructure, the cost of installing 10 telephones would be expected to be roughly double the cost of installing five.
In recent years, with the development of initially the internet, and subsequently mobile communications, we’ve seen the cost of adding connections reduce by several orders of magnitude, while the benefit of new connections continues to increase. This can be seen with markets like social media, web search, knowledge sharing etc.
We’re now at somewhat of a tipping point.
The marginal cost of processing power and storage is approaching zero. That’s not to say that it will become free, but it is now trivial in terms of effort and cost to essentially rent whatever processing power or storage you want through cloud computing. This is a relatively new phenomenon, and has only really become available in the last few years.
Research, modelling or analysis which use to require massive hardware and cost or rare compute time on a supercomputer, and which may have taken days to perform, can now be either run on or from a laptop in seconds or minutes. And its only getting faster.
At the same time, the amount of data being collected is rapidly increasing, providing more information from which patterns can be inferred, and the internet of things (IoT) is rapidly increasing the number of network connections.
Cisco is predicting a 77 per cent CAGR in mobile traffic from 2014 to 2019, from 2.6 to 46 exabytes per month, and an increase in machine to machine communications of 71 per cent CAGR over the same period, from 0.3 to 4.6 petabytes per month.
All of this is facilitating the growth of machine learning / artificial intelligence, and leading to the next potential wave of disruption of business, such as that currently challenging GE.
But more on GE later.
Innovation, disruption and other buzz words
Regardless of what you call it, technological changes have had a significant impact on the business world in the last 20 years, and are continuing to do so.
Figure 3 shows the change in market capitalisation of the top 10 companies in 2016 over the last eight years. Solid lines are tech companies.
Even if you combine the two Alphabet businesses (Google), still half of the top 10 companies are now technology companies, and are all growing fast.
Again, more on GE later.
While tech companies are doing well, business in general is finding it more difficult. According to recent studies by innosight and the Boston Consulting Group (BCG), the average life of a company on the USA S&P is continuing to decline. See figures 4a and 4b.
There have also been some spectacular failures of major companies. Companies that had everything they needed to succeed, but failed to understand the changing market. Two classic examples are Kodak and Nokia.
Kodak dominate the photography market for over 100 years but went from success to bankruptcy in the space of fourteen years, filing for bankruptcy in 2012, largely as a result of the growth in digital photography (figure 5)
In 1996 Kodak had over 80 per cent market share. It invented digital photography, owned many of the patents related to it and, in 1986, was the first company to release a digital camera into the market.
So what went wrong?
Here is a recent article about the failure of Kodak by Harvard Business Review (HBR).
Ultimately Kodak had all the elements to succeed in digital photography, but it was fixated on printing. Kodak even purchased a photo sharing website before social media existed, but tried to use it encourage people to print photos. The company structure, incentive schemes, development paths etc were also still based on the old business model where the development, manufacture and sale of chemicals and other photographic equipment was the norm. Chemical engineers were the golden children of the old Kodak, they didn’t help the new one.
From the HBR article;
Kodak created a digital camera, invested in the technology, and even understood that photos would be shared online. Where they failed was in realizing that online photo sharing was the new business, not just a way to expand the printing business.
Like Kodak, Nokia was doing everything right and was flying high one minute, the next it was gone (figure 6).
Like Kodak, Nokia owned many of the key patents associated with mobile telephones and mobile phone networks, having essentially started the worldwide development of mobile communications.
Ask anyone older than thirty if they ever owned a Nokia phone, chances are they did. But how many do today? None.
Nokia dominated the mobile phone industry for many years, was the first to release a touch screen smart phone to market, in 2007 had 67 per cent market share of mobile operating systems and was the first to have an app store (figure 7)
Nokia, like the rest of the market, was focused on making feature phones smaller, lighter and cheaper, and on growing its sales in terms of number of handsets. Nokia was the first to offer phone customisation options, recognising that customers were treating their phones as fashion accessories, but it missed initially the flip phone revolution, dominated by Motorola, and later the smart phone revolution.
The introduction of the iPhone in 2007, essentially a pocket computer which also acted as a phone, and which was offered at a significantly higher price, started the smartphone revolution and the downfall of Nokia.
It is estimated that Apple invested approximately $150m in R&D for the iPhone during a time when Nokia was spending $3-4 billion pa on R&D. The iPhone didn’t invent new technology in developing the iPhone, but it did introduce new ways to use it. For example, a large storage capacity, for its day, an advanced tough screen, and ultimately an ecosystem which attracted developers to invent new ways by which customers could use their iPhones.
Nokia’s operating system (Symbian) wasn’t a full featured computer system and was difficult to develop for. It had the opportunity to be involved in Android development and Microsoft’s phone OS, but decided not to.
As a counter example, Samsung, one of Nokia’s main competitors just prior to its demise, at least saw the start of the smartphone revolution increasing its number of Android developers from 50 to 2,000 between 2008 and 2009. This potentially saved Samsung from a similar fate although it doesn’t control the Android ecosystem, Google does.
Apples ownership of the iOS ecosystem enables it to make its profits from hardware (90 per cent of Apple profits come from hardware sales), and ‘lock’ customers in through the ecosystem.
So what does all of this have to do with non-tech businesses?
Back to GE.
GE is a prime example of a traditional, non-tech, business which is innovating its products and business models in-order to survive and thrive into the future.
GE has traditionally been a products and services business, developing, manufacture and selling equipment for a profit, and offering maintenance services for a profit. GE’s key operating capabilities were:
- Quality, reliability and reputation
- Product features and performance
- Service network and plans
- Performance guarantees
In the last few years various analytics focused companies have started to offer services to GE product customers to monitor their equipment and provide information on life-cycle extension, operating cost reduction and efficiency improvements, monogamist other things. These companies include IBM, SAP, accenture and Google.
The growth of these service offerings has several potential negative impact on GE including:
- reducing GE’s potential service revenue
- enabling someone other than GE to influence equipment purchase decisions and advise customers of the best solution for their needs
- the potential loss of information beneficial to GE.
If this pattern continued, GE’s products, as well as those of other manufactures could become largely commoditised
In order to fight this trend, and to capture additional or different value in the future, since 2012 GE has been working to adjust its business and build its solution, the industrial internet of things.
GE’s industrial internet enables sensors and data from all GE equipment to operate on a common platform, and for an analytics ecosystem to provide monitoring and information on equipment globally. As more equipment is networked, value continues to increase in accordance with Metcalf’s law. For example, there are millions of GE locomotive engines operating globally, information learnt and insights attained from analysing these in various operating conditions will not only help to improve their performance overall but will inform the development of new components and products. Figure 8 illustrates the value of a one per cent saving in various markets over fifteen years which may be delivered through this initiative.
But, this is an outcomes based business model, which is fundamentally different from the product/service based model GE has historically operated under. GE’s new key operating capabilities will need to be:
- customer cost reductions and efficiency improvements
- complex and bespoke, outcomes based sales contracts
- revenue and cost reduction sharing
- data analytics
To facilitate the development and implementation of this new business model, GE has been pushing two key initiatives:
- developing the innovation and change preparedness of existing BUs through incubators; and
- the establishment of a software and analytics organisation to develop the required technology and systems, and to work with the BUs to deliver change.
Within BUs, incubators have been facilitated by Harvard Business School to encourage the development and rapid introduction of new product offers and contracts as prototypes of how to work with customers using the new business model. BUs have also teamed up with external incubators in order to inject new ideas, see figure 10.
The software and analytics organisation was established in Silicon Valley and operates as a cost center. The organisation worked to determine the proposed structure of GE’s industrial internet systems and working with BUs to help them:
- standardise monitoring, data collection and other software systems
- rapidly deliver technological initiatives; and
- develop analytic tools and products.
GE now offers these services though its Predix platform (Figure 11).
These initiatives and changes have all been driven by GE’s CEO Jeff Immelt since inception. After initially being identified as a key strategic issue by GE’s former CMO Beth Comstock, Immelt spent six months getting a better understanding of developing technology and its potential strategic impact on GE, before commencing implementation of the current initiatives.
Industrial internet services are currently responsible for $5bn of GE earnings and growing.
What does this mean for enterprise?
Technology is advancing rapidly delivering numerous changes:
- the growing collection of data
- the power and value of analytics (historical and predictive)
- the benefits and learnings available from networked sensors and equipment
- improvements in automation
- machine learning and artificial intelligence
- improvements in resource use which could significantly reduce the size of markets
- new technological advancements and their potential impact on industry
- renewable energy
- energy storage
- sensor miniaturization
- new materials (e.g. graphene)
These changes will continue to impact all businesses and markets, regardless of whether they rely on technology or not. GE is a prime example of this.
Businesses which are not actively watching these developments, facilitating and recognising the value of innovation, and watching developments in current and potential markets are at best missing opportunities, and at worst sailing into oblivion, just like Kodak and Nokia.