Published

14

Apr

2020

Commentary

The Ways & Means of Industrialization

In 2011, LEF developed Figure 1 to describe a set of significant changes that we anticipated would occur in organizations across many of our client industries.  We characterized the vast majority of organizations as being more traditional and predicted how, over the following decade, many would need to become more next generation: less of a department, more of a cell-based structure; less focused on profit, more on disruption; less inclined towards analysts and more towards using ecosystems as their sensing engines.  Less of a one-size-fits-all model and more equipped with chaos engines, scale-out and appropriate methods. The table was then published in our report Learning from Web 2.0.

Figure 1 – A prediction of future changes to organization, 2011

Almost a decade later, many of us are now living in that ‘next generation’ world, while others are still starting out on the journey.  However, a decade ago concepts such as DevOps were still emerging, and mostly disregarded.  Even the idea of the cloud was still in dispute and described as “irrational exuberance” and “unrealistic expectations”.  According to McKinsey, our future was one of focusing “on the immediate benefits of virtualizing server storage” 1.  Even big data was in its infancy with analysts arguing over what it was.

So how did we get it so right across such a broad spectrum in the face of scepticism and even derision?  Were we just lucky? The truth is that when we look at economic systems, not everything is equally unpredictable.  Not everything is random. Not everything is an unknown unknown. With the aid of context and weak signals, we can identify certain patterns and use them to explain change.  One of those patterns concerns the industrialization of technology.

What is industrialization?

We use the term to mean the shift of a technology from a world of product competition to one more of commodities and operational competition.  This shift has a known set of impacts, including the co-evolution of practice that enables new needs to be met and new forms of industry, organization and behaviour to emerge.

But why is context important?  Without context, without maps, there is little chance of identifying repeating patterns such as industrialization.  Unfortunately, our world mostly relies on story telling. We have the rule of narrative both in the political sphere, where science is sometimes reduced to an inconvenience in the way of a good story, and in business, where huge conglomerates repeatedly incur costly failures in the pursuit of efficiency based upon little more than grand stories and inadequate concepts like process flow.  That the CEO impact is almost indistinguishable from random chance 2 is not a sign of poor leadership but a widespread lack of situational awareness.

We at LEF apply context and use known patterns to identify potential points of change, then discover through interviews behavioural changes that people are mostly unaware of, and follow this up with surveys to identify the wider emergence of those changes,  As a result, the ‘random’ can start to be read and start to have meaning. A little bit of mapping and a little bit of context can work wonders. It can almost seem magical, but it should be stressed, there’s no magic at work here. There are just maps.  

Today, we are facing multiple waves of industrialization along with rapid changes in needs – for example, the need for isolation caused by the covid-19 pandemic.  What new practices and new industry will emerge is harder to predict, but at the very least we know to go and look for them (in interviews); and once we identify candidates, we need to follow them up (with surveys).  We have used this same bag of tricks for the reports on serverless and the growth of China – both changes that are fairly inevitable, although denial remains in most quarters.

It should be said here that there are nuances, there are ways in which individual actors can influence and exploit the space.  Industrialization is triggered by a number of factors: concept, suitability, technology and a change in attitude. For example, cloud – which represented the industrialization of compute from product (i.e. servers) to utility, and the corresponding co-evolution of practice (i.e. DevOps) – needed the concept of utility provision (1960s); compute needed to be widespread and well-defined enough (1990s); the technology (Xen, etc, 2000s) needed to be available; and finally, it needed attitude (i.e. companies fed up with the current state of provision). This last requirement is critical: for industrialization to occur, whatever is currently being provided must somehow not be meeting the consumers’ needs – be they the public, government or enterprise – and someone (the actor) must be willing to act on that.

A time of change

We have a challenging time ahead of us. Weak signals suggested that a number of discrete areas of technology will industrialize over the next decade – identified as points of war in Figure 2 (from our report Of Wonders and Disruption). We again have a rapidly evolving set of needs because of the pandemic.

Figure 2 – Weak signals

Though we might anticipate that change will occur, we have yet to identify the novel and emerging practices for each area of technology affected. The problem with this identification is that large existing players tend to have inertia to change. So, if you ask the 'experts’ – or even worse, if you survey the 'experts’ and aggregate them – you almost always end up with exactly the wrong answer. As the marquis de Condorcet pointed out, the crowd is equally wise and stupid. To find the significant emerging practices and behavioural changes, we need to go outside of accepted wisdom. In the early days of cloud, it was more revealing to talk not to the great and the good (IBM, HP, VMware, Dell, Oracle) but to the upstart book seller (Amazon) which had wandered into the industry.

Also, the impact of these new co-evolved practices will vary according to the different value chains and contexts of different industries. Not all industries will be equally affected by such technology change. To complicate matters further, we have exogenous shocks, such as the new need for isolation that the covid-19 pandemic is bringing us, which not only impacts value chains but also the behaviour, and hence the culture, of many collectives. Fortunately, we can map industries and we can map culture. We have tools for exploring these spaces as well. We have all the methods we need to prepare us for when that pandemic ends; we just need to apply them.

Next steps

So, how do we apply this? How do we sort through the mass of changes approaching us to make some sort of coherent and actionable instructions for the future? How do we build a new version of Figure 1 in time to help our organizations and economies when they start to emerge from the current pandemic?

From weak signals I’ve observed, I’ve chosen five areas of technology industrialization – robotics, sensors, immersion, manufacturing and space – that we should explore. We’re going to need to look for emerging practices related to industrialization. The caveat, as we discussed above, is that most experts in these fields will either be unaware of, or have inertia to, these changes. So, we have to look for the oddities, the start-ups on the edge, the land of the different. This thinking was behind our selection of leading-edge companies – iRobots (robotics), BrainCo (the use of sensors in education), Sixsense (the use of immersion and robotics in healthcare) and Lynk Global (the use of space and satellites for improving connectivity in the world) – for the 2020 study tour. However, in order to have any real chance of identifying practices across so many technology areas, we’re going to be forming five virtual working groups, one for each topic.

Alas, it’s not enough to simply identify practices and behaviour – we also need to understand their impact on different industries. This requires us to map these industries and then apply those changing practices to them. So, we’re going to need a further four working groups – defence, healthcare, automotive and government – to undertake that mapping and to test various scenarios.

Finally, we’re going to need a virtual group to explore the changes in needs – those exogeneous shocks such as covid-19. We will look at this in terms of impact along with the application of the more resilient practices that developed from past industrialization (such as chaos engines, design for failure, distributed systems) to other systems of technology. Design for failure combined with robots might sound like a science fiction horror story but may well emerge as an essential element within defence and emergency planning. As we learned and demonstrated with cloud, we can build highly reliable and secure computing systems out of highly unreliable and insecure components. 3

Two things to note. First, I use resilience in Holling’s sense of engineering versus ecological 4 with life as a resilient system (see Figure 3) rather than any other popular terminology. Second, the change of needs may also have long term cultural and behavioural implications. Whilst this year’s study tour will be postponed until 2021 and will be supplemented by a Study Tour Virtual Elective to build momentum with virtual discussion and interviews, it’s a brave person who can guarantee that our isolation economy will be over in a year. We have already proved that, despite the optimistic prognostications of fanciful story tellers, it was not all over by Easter.

Figure 3 – Resilience

Figure 3 – Resilience

Five virtual working groups on technology industrialization, four working groups on mapping specific industries, one virtual group mapping culture change, and all groups – technology, industry and culture – communicating with maps. It’s a grand undertaking but this is what it will require to produce our next decade’s Figure 1. At the end of this research project, we intend to bring the groups together and test their conclusions with our chosen companies on the study tour before publishing the final report.

A call for action

This pandemic will be over. Competition and evolution will continue. Our world will change. We need to prepare for this and hence we’re inviting LEF clients to join one of those working groups. You’ll need to learn how to map, to use maps and patterns to anticipate economic change, to scenario-plan and communicate between groups with maps; but we won’t throw you in at the deep end. There is a LEF online course on Wardley Mapping and an online book, and we will organize a few virtual sessions for newcomers to this craft who are willing to be involved.


1. Clearing the air on cloud computing, Mckinsey & Co, March 2009.
2. Markus Fitza, The use of variance decomposition in the investigation of CEO effects: How large must the CEO effect be to rule out chance?, John Wiley & Sons, 2013
3. JE Dobson & B Randell, Building Reliable Secure Computing Systems out of Unreliable Insecure Components, IEEE Symposium on Security & Privacy, 1986
4. CS Holling, Engineering Versus Ecological Resilience, in ’Engineering Within Ecological Constraints’, National Academy of Engineering, 1996


Cookies

We use cookies to improve the user experience of our website. If you continue to browse, we'll assume you are happy for your web browser to receive all cookies from our website. See our cookies policy for more information on cookies and how to manage them.

Accept