Christian Delbert - Fotolia
In the first of our extracts from Simon Wardley’s work, we learned how as a new CEO in 2004 he realised there was no business equivalent of a map to guide strategy decisions, which was like playing chess without being able to see the board. In the second, he gave instructions for drawing a simple map of your business along the lines of his own first map from 2005. In this final extract, we will see how to apply climatic patterns to extend the map, anticipate the future and inform your strategy.
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
The purpose of producing a map is to help us to learn and then apply basic climatic patterns, doctrine and context-specific forms of gameplay. Maps are our learning and communication tool for discovering these things and enabling us to make better decisions before acting.
However, the strategy cycle is iterative and we’re not going to learn all the patterns the first time we use a map any more than we learn everything about chess in our first game. Instead, like a game of chess, play by play, move by move, we get a little bit better. This is what happened to me starting in 2005 and I’m still learning.
In this article, I’m going to look at the effect of climate – the third of Sun Tzu’s five factors that matter in competition. I’ll go through a number of climatic patterns relevant to business and apply them to our first map.
Climatic patterns are those things that change the map regardless of your actions – for example, common economic patterns and competitor actions. Understanding climatic patterns is important when anticipating change. You cannot stop climatic patterns from happening, however, as you’ll discover, you can influence, use and exploit them.
Climatic pattern 1: Everything evolves
Organisations consist of value chains made up of components that are all evolving from genesis towards commodity under the influence of supply and demand competition. So, all components on the map are moving from left to right. This includes every activity (what we do), every practice (how we do it) and every mental model (how we make sense of it). Everything has a past and a future.
For example, as shown in Figure 1 (see below), in the map of our online photo service in 2005, the component labelled “platform” was positioned at the “product” stage of evolution. In our case, it was provided by the Lamp (Linux, Apache, MySQL and Perl) stack, but other competing product sets with different features were also available.
Things had already moved beyond the stage of building our own operating system, computer language and web server software – although a few years before, that’s what we would have needed to do.
Although I would have loved to metaphorically flip a switch and start coding on some form of utility platform, we still had to manage our own installation, configuration, setup, networks and the many underlying components that had to fit together to provide a working stack. However, the platform was evolving and would at some point in the future become a commodity, even a utility.
Figure 1: Everything evolves
The same principle applies to the component labelled “compute”, shown further to the right. In our company we had already created a system known as the Borg which provided virtual machines on demand. It would only take a small leap from that to the compute utilities described by Douglas Parkhill in his 1966 book, The Challenge of the Computer Utility.
Climatic pattern 2: Characteristics change
The journey of evolution from genesis towards commodity has profound effects because the basic characteristics of the component have to change. For example, take the genesis of computer infrastructure from the first digital computer in the 1940s.
Back then, the activity was scarce, it was poorly understood and we were still in the process of discovering what a digital computer could do. Computing was uncertain, unpredictable and rapidly changing. There was no firm market to speak of, and customers were on as much of a journey of exploration as the suppliers. Yet it had the potential to make a difference and be a source of differential value and competitive advantage.
Computing infrastructure did prove to be useful and the idea started to spread. Custom-built systems such as LEO (Lyons Electronic Office) emerged and later standard products were launched (such as the IBM 650) followed by ever-more functionally complete systems.
By 2005, computing infrastructure was starting to be treated as a commodity, with racks of fairly standardised servers. Computers had become commonplace and their purpose and use were well understood by a large number of people.
We were already starting to think less about what a digital computer could do and instead about what we could do with vast numbers of standard units. In our Borg system, we had even abstracted away the concept of the physical machine to virtual ones that we created and discarded with abandon.
I had seen the same changes in the behaviour of the users of our online photo service as the industry evolved from photo film to digital images. In the past, every single photo taken on film was precious; as the format became digital and more of a commodity, users increasingly took many shots. Throwing away unwanted pictures was no longer seen as waste but an expected consequence of taking thousands of them.
Everything evolves from the uncharted and unexplored space of being rare, constantly changing and poorly understood, to eventually industrialised forms that are commonplace, standardised and a cost of doing business. What happened with computers and images had also happened with electricity and Penicillin, the once marvel drug that became a generic.
This progression and change of characteristics is shown in Figure 2, on which I’ve also listed the characteristics of the uncharted (genesis) and the industrialised (commodity/utility) domains.
Figure 2: Characteristics change
You cannot stop the components on your map evolving if there is competition around them and, as they do so, their characteristics change. Since this change is common for all components, I was able to draw up a list of characteristics at different stages of evolution to produce the table or “cheat sheet” shown as Figure 7 in the second article of this series.
Climatic pattern 3: No one size fits all
Given that every large system, whether a line of business or a specific IT project, contains multiple components that have relationships with each other but are also evolving – and as they evolve, their characteristics are changing from the uncharted to the industrialised domain – you need to manage both extremes to survive and compete against others. You cannot afford to be hand-carving nut-and-bolt pairs when a commodity form exists.
When dealing with the industrialised domain, you need to encourage coherence, co-ordination, efficiency and stability. However, the exploration and discovery of new capabilities in the uncharted domain requires a different approach. Any structure, whether a company or a team, needs to manage both of these polar opposites. This is known as the innovation paradox.
The story is even more complex because there will also be components on the journey between the extremes. These transitional components have a different set of characteristics and require a third mechanism of management.
For example, in the uncharted (genesis) space, no one knows what is wanted and exploration and experimentation are required. Change is the norm and the management method must enable and reduce the cost of change. In this part of the map, I tend to use an agile approach that has been cut right back to core principles – a very lightweight version of extreme programming (XP) or Scrum.
As a component evolves and we start to understand it more, our focus changes. Sometime during the custom-built stage, we start to think about creating a product. While we may continue to use underlying techniques such as XP or Scrum, our focus is now on reducing waste, improving measurements, learning and creating that first minimal viable product.
Today, lean is a popular approach for this stage. The component continues to evolve, becoming more widespread and defined as it approaches the domain of industrialised volume operations.
Our focus again switches, this time to mass production of “good enough” products, which means reducing deviation. At this point, Six Sigma along with formalised frameworks such as ITIL are appropriate.
This also applies in other fields. Purchasing requires a venture capital-based approach in genesis, switching to outcome- and off-the-shelf-based approaches during transition and then more unit-based approaches in the industrialised stage. Hence any large system, whether a company or even a government, needs to use multiple purchasing methods.
Equally, genesis is more suited to in-house development whereas the industrialised can be safely outsourced. Even the best approaches to budgeting are vastly different, moving from investment accounting to product profit-and-loss to activity-based cost control.
Any significant system will have components at different stages of evolution. There is no single method that will fit all. In finance or IT or marketing, there is no single magic method. Unfortunately, most companies have no map of their environment to understand the differing needs and plump for a one-size-fits-all solution – but it doesn’t.
Climatic pattern 4: Efficiency enables innovation
The story of evolution is further complicated because components not only evolve but also enable new higher-order systems to appear. A standard electricity supply paved the way for all manner of things from television to computing. Genesis begets evolution begets genesis.
In his theory of hierarchy, Herbert Simon showed how the creation of a system is dependent on the organisation of its subsystems. As an activity becomes industrialised and provided as ever-more standardised and commodity components, it not only enables faster implementation, but also rapid change, diversity and agility of the systems that are built on it.
In other words, it’s quicker to build a house with commodity components such as bricks, wooden planks and plastic pipes than it is to start from a clay pit, a clump of trees and an oil well.
This doesn’t mean that change stops when components are standardised. Significant improvements continue, but the standard acts as an abstraction layer. When an electricity supplier introduces sources of power generation such as renewables, we don’t have to rewire our houses. But if the constant operational improvement in electricity generation were not hidden behind the interface, all the consumer electronics built on it would also need to continuously change.
Instead, as a component evolves to a standard commodity and then to a consumer product, improvements are increasingly hidden. Changes are reflected as greater efficiency or a better price or quality of service, but the activity itself to all intents and purposes remains as it is. Exceptions to this create significant upheaval because of all the higher-order systems that need to change, and government involvement is often required – for example, when changing the electricity standard or the currency.
The increase in efficiency that results from the provision of more industrialised components enables innovation. For example, electricity generation evolved through competition to become more industrialised, which in turn through componentisation effects enabled higher-order systems such as computing, which enabled new industries serving new user needs.
Computing in its turn evolved through competition, enabling the creation of novel higher-order systems such as databases, which again enabled new industries. And so the process continued, until we have today’s intelligent machine agents.
This is summarised in Figure 3, where the black line shows the present and the red lines both the past and the enabling effect of evolution, producing new capabilities. The figure adds the fairly obvious anticipation – where we will be – that intelligent agents will themselves become commodity-like.
Figure 3: Efficiency enables innovation
In the above map, I’ve reduced the actual number of the components for simplicity. Obviously, not everything becomes a component of something else, but mechanical, electrical and even IT systems commonly do.
Without a long history of more industrialised forms offering highly efficient components for once-magical wonders, I would never have had the ability nor the capital to write this in a Microsoft Word processor on a digital computer.
Modern-day cloud computing represents the evolution of many IT activities from product to utility services, and the provision of standard components is fuelling the rapid rate of development of higher-order systems and activities. Many services we consume, from Netflix to Dropbox, are unlikely to have been practical without commodity and utility computing infrastructure.
Climatic pattern 5: Higher-order systems create new sources of worth
An idea has social value; the implementation of that idea as a new activity can create economic value when that activity is useful. This process of transformation from social to economic value is known as commodification.
As a new activity evolves, various iterations of it will spread throughout society until the activity becomes commonplace and its differential benefit reduces close to zero. What started with a high differential benefit due to its scarcity ultimately evolves to have little or none because it is commonplace – a commodity. This is the process of commoditisation.
At the same time that the differential benefit of a component declines, it also becomes more of a necessity and a cost of doing business, such as the telephone.
This creates a situation where the unit value of something may be declining while the total revenue generated is increasing due to volume. Alongside this is the cost of production of each unit changes as it evolves. For example, the unit cost of production of a landline phone is vastly less today than when phones were first invented.
As a result, the transitional domain – the time of products on the evolution axis of our maps – between the extremes of the uncharted and the industrialised tends to be the most profitable in an industry.
This wealth generation is due to a combination of high unit value, increasing volume and declining production costs. As a rule of thumb:
- The uncharted domain is associated with high production costs and high levels of uncertainty but potentially very high future opportunity. Being first is not always the best course due to the burden and risks of research and development.
- The transitional domain is associated with reducing uncertainty, declining production costs, increasing volumes and highest profitability. However, while the environment has become more predictable, the future opportunity is also in decline as the act is becoming more widespread, understood and defined. So as we reach the zenith of wealth creation, the future of the industry is looking decidedly less rosy.
- The industrialised domain is associated with high certainty, high levels of predictability, high volumes, low production costs and low unit margin. The activity is not seen as differential but an expected norm – it has become commonplace. Activities that have evolved to this state – such as nuts and bolts – have a minimal differential effect. They are not associated with high future opportunity except in early stage replacement of any existing product industry. Their future is one of stable and increasingly low-margin revenues that may nevertheless be significant due to volume.
However, as we have already seen, the more industrialised components enable new higher-order systems that are future sources of worth and wealth generation, such as electricity-enabled radio, television and computing. This is shown in Figure 4.
The downside is that the higher order systems are uncertain and you do not know which will be successful – electricity also enabled Edison’s electric pen and Gaugler’s refrigerating blanket.
Figure 4 - Higher order systems create new sources of worth
Climatic pattern 6: There is no choice on evolution
As components in your value chain evolve then some competitors will adapt to use it, unless you can form some sort of cartel and prevent any new entrants. The benefits of efficiency, faster creation of higher-order systems and new potential sources of wealth create pressure on others to adapt.
As more adopt the evolved components, the pressure on those who remain in the old world increases until it is overwhelming. This effect is known as Van Valen’s Red Queen hypothesis and it is the reason why companies don’t build their own generators from scratch to supply their own electricity.
A secondary impact of the Red Queen hypothesis is that it limits one organisation – in biology, one organism – from taking over the entire environment in a runaway process.
If, for example, only Ford had ever adopted mass production and every other good was entirely handmade, then not only would every car today would be a Ford but so would every radio, every TV and every computer. In practice, new practices spread and other industries adapt. Hence the advantage that Ford created was diminished.
Climatic pattern 7: Past success breeds inertia
The Red Queen hypothesis might force organisations to adapt, but this process is rarely smooth. The problem is past success creating inertia.
If you make a component that is evolving from product towards utility, you are likely to resist its industrialisation because you are benefitting from the time of highest profitability and wealth creation in the transitional domain. You want things to stay exactly as they are. This resistance to movement is inertia, shown in Figure 5. Both consumers and suppliers exhibit various forms of inertia due to past success in either supplying or using a product.
Figure 5: Past success breeds inertia
It is almost always new entrants unencumbered by past success that initiate this change. While VMware’s CEO Pat Gelsinger felt that Amazon as a “company that sells books” shouldn’t beat VMware and its partners in infrastructure provision, it is precisely because Amazon was not encumbered by an existing business model that it could so easily industrialise the computing infrastructure space.
The natural initial reaction to the change is scepticism, despite any latent frustrations of consumers with the costs associated with past models.
However, some consumers – usually new entrants themselves entering other industries – start to adopt the more evolved components because of the benefits of efficiency, agility and ability to build higher-order systems of value.
The Red Queen kicks in, pressure mounts for others to adopt, and what started with a trickle suddenly becomes a raging flood.
Existing suppliers’ resistance to change will continue until it has become abundantly clear that the past model is going to decline. Unfortunately for those suppliers, by then it is often too late as the new entrants already dominate the future market. Many past giants don’t survive.
This sequence of new entrants, a trickle of adoption becoming a flood and slow-moving past giants due to inertia is common throughout history.
Using climatic patterns
We’ve quickly covered some basic climatic patterns:
- Everything evolves.
- Characteristics change
- No one size fits all
- Efficiency enables innovation
- Higher order systems create sources of worth
- There is no choice on evolution
- Past success breeds inertia
Now we’ll take the same step that I did back in 2005 and apply some of these basic patterns to my first map of our online photo business, highlighted in red in Figure 6.
Figure 6: First map with patterns
Through this process, I was able to anticipate in 2005 that:
Point 1: Our online photo service was moving into the product stage of maximum wealth generation.
This meant it was going to become much easier for others to create a competing service and big players were likely to move into the space. This had already started and, though our diversified focus might have enabled us to survive, we were rapidly falling behind our competitors.
We were doing well because everyone was doing well, but we were relatively small fry and that wasn’t going to improve unless we refocused. We needed either to invest or to find a new angle and some new differentiator. However, I had to be mindful of the fact that we lacked the financial muscle of others and any investment in something novel was a gamble.
Point 2: Compute was likely to become more of a utility. I didn’t know quite when but I had signals that this transformation was going to happen soon – for example, that a company like ours could create its own internal private utility, or what is now called a private cloud.
Compute was a massive industry with huge profitability and revenues; someone was likely to attack it, and that someone would not be encumbered by an existing product or rental model. I expected that Google would be first.
Point 3: There would be resistance to the change – inertia – of compute becoming a utility. That inertia would exist in suppliers of both hardware and rental services, along with their customers. Compute was going to evolve regardless, and companies would be under pressure to adapt. The first movers would likely be unencumbered companies, such as startups.
Point 4: What was going to happen to compute was also going to happen to coding platforms. This was another area where there was considerable revenue and profitability to attack. Tedious tasks such as configuration, setup and installation would disappear. We were going to enter a future world where I could just code and deploy.
Point 5: These utility coding platforms would eventually run on utility compute environments. We could anticipate a “line of the future” where the relationships between components remained the same but the manner in which they were provided differed.
Point 6: The transition from product to utility for both compute and platform was going to enable all sorts of novel higher-order systems to be created rapidly. I had no idea what they would be but among them there would be many new sources of worth along with many more failed efforts. Again – everything novel is a gamble.
I sat in the boardroom looking at the huge map that I had created with my CIO’s help. In reality, it was far more complex than the simplified version above and used slightly different terms. But for the first time in my business life, I was able to have a serious conversation about what we thought was going to change.
Had you been in that room, you might have disagreed with how we had positioned the pieces or the patterns we saw, but at least we had a basis for discussion.
It felt exciting but also intimidating. We were talking about fundamental changes to the computing industry staring us in the face with what seemed like blinding obviousness.
I realised I had a visual means of demonstrating what Nicholas Carr had described in his exceptional 2003 paper IT Doesn’t Matter.
I was a huge fan of that paper and his subsequent prophetic book and had got into many arguments over it. When I mentioned what I thought were amazing ideas to my peers, they had roundly ridiculed them. Compute, they said, was a relationship business, all about trust. I disagreed, and now with the map I had the evidence to back my position up.
Read all three extracts from Simon Wardley’s book on value chain mapping
- Making sense of executive strategy: In this first excerpt from his book, Wardley explains why business leaders need to understand the importance of maps for corporate strategy.
- Finding a path: In the second extract from his forthcoming book, Wardley explains how to draw a map to describe the changing nature of your business.
- Exploring the map: In the third and final excerpt from his book, Wardley explains how maps can determine future business strategy.