or region of the world.
We are faced therefore with a conundrum. On the one hand we have to account for the human craving for certainty about the future. On the other hand we have to work within the constraints imposed by the emergent property of complex systems. The solution lies in scenario planning.
Many clients are surprised to hear that scenario planning is not forecasting. The two methods are quite distinct. Whereas the forecaster seeks to define precisely a single future point in space and time, the scenario planner is seeking to identify a number of equally plausible points in future space and time. The distinction is that to the forecaster the future is a singular concept – there can be only one future and his or her job is to identify it. For the scenario planner, the future is a plural concept. There will always be more than one plausible future, and each of these must be respected as having a roughly equal degree of plausibility. It is only by accepting this that it becomes possible to overcome the crippling effects of the emergent property of complex systems.
At this point more than one client has suggested that it all sounds a bit hopeless. What is the point of identifying a series of roughly equally plausible futures? How can a decision be taken? What should she or he tell the shareholders and the board? That client has just run into reality and is now in danger of succumbing to the temptation to start forecasting.
Fortunately there is an acceptable compromise.
Our first answer to such a client will be that scenario planning projects are not going to throw up hundreds of different futures. Most scenario projects will deliver between two and four different future worlds. Our second answer will be for the client to develop a strategy for each world and then adopt the one that seems to align most closely with the current environment. Not for a moment, however, should that strategic decision be taken at the cost of jettisoning the other scenarios. The client should develop a series of indicators indicative of the emergence of each of the other scenarios and be prepared to turn on a dime the moment it seems that a new scenario has become most probable.
Take a practical example. Many clients ask advice on how ANC economic policy will change. A forecaster could provide an answer, and the client might use that answer to build a strategy for his operations in South Africa. A scenario planner would provide two or three different sets of answers, probably going as far as compelling the client to face up to the question of what economic policy would look like if the ANC were no longer around. A decision could then be made about which of the answers the current climate seems to suggest as the most probable, and the strategy that applies to that scenario can be put into operation. But when the climate changes – and it will – the client already knows exactly what to do, while competitors are scrambling to figure out what just happened and how to respond.
Does the singling out of one scenario not contradict the complex systems basis of scenario planning? To an extent it does, but reality leaves no alternative and as long as it is not done, with the result of writing off the other scenarios, it offers the best solution to the conundrum that exists between the prescripts of complexity theory and emergence, on the one hand, and the neurological craving to know what will happen tomorrow, on the other.
The differences between forecasts and scenarios are therefore the following:
Forecasts develop a single certain future around which a strategy must be built. There is not much early warning that the forecast may be wrong. There is no fall-back position or Plan B if it is wrong. New strategy must be developed in the midst of the chaos of change.
Scenarios usually develop two to four varied futures. A strategy is built around each future, and indicators are available to show which future is most likely to materialise. If the markers change it is easy to change strategy. In a well-built set of scenarios nothing should be able to occur that takes the company, country or government that commissioned the scenarios by surprise. They have a contingency plan for each eventuality. Somewhat counterintuitively, by agreeing to work within the constraints imposed by the emergent property of complex systems, the client who has agreed to accept the plural nature of the future actually has far more certainty about the future than the client who chooses to rely on a single forecast.
How do you create a set of scenarios? The methodology most commonly employed is tried and tested and has been used with variations by consultancies around the world for over 30 years.
The first step is to identify what the client actually wants to know, and over what time frame he or she wants to know it. In other words, what is the focal question and over what time horizon must the question be answered? The question can be very broad or extremely narrow. One request may be to know how banks might be exposed to land reform policy over the next three to four years. Another may be what the long-term (ten- to twenty-year) implications of current mining policy might be for greenfield mining exploration in South Africa. An activist group may want to know what the worst-case outcome for civil rights might be in order to test the likely efficacy of a contingency plan it had developed. A media company may want to know what the effect of ‘view on demand’ technology might be for traditional radio and television stations. In the case of this book we want to know what life will be like in the South Africa of 2030.
The second step is to identify every economic, social and political force that might have an impact on that decision. The net must be cast very wide, and several hundred indicators or pieces of information may be gathered. In the case of this book, four chapters will be devoted to explaining current economic, social and political trends.
The third step is to gather those trends into a series of groups or families of related major trends. Ideally we want to get down to 40 or 50 major trends, each of which will have a definitive influence on the question the scenarios seek to address.
The fourth step is to rank those trends according to the impact they are likely to have on the core question that the scenarios are trying to answer, and the uncertainty associated with that impact. This is done on a graph with two axes such as the one set out below. The left axis measures relative uncertainty and the bottom axis measures relative impact. Trends that are grouped towards the lower left corner of the graphic will have a relatively limited impact on the scenarios, and there is relatively little uncertainty about what that impact will be. Trends grouped towards the top right corner will have a relatively high degree of impact on the question the scenarios are trying to answer, and there is great uncertainty about what that impact will be. It is these types of trends against which business and political strategists need to test their contingency plans if they are to be confident that those contingencies can anticipate and respond effectively to sudden and dramatic shifts in the environments they operate in.
The fifth step is to determine what those sudden shifts are likely to be. This is done by taking the trend of greatest impact and plotting it against that of greatest uncertainty on a matrix such as that set out below. The matrix in turn delivers four quadrants, and each of these will become one scenario. The robustness of this methodology is that it takes the greatest uncertainties faced by an organisation and multiplies these by the trends that will have the greatest impact on that organisation. To demonstrate how this works, the matrix provides an example of a set of mining and natural gas scenarios. That hypothetical study suggested that geological conditions would have the greatest impact on the future of the mining industry in South Africa, while mining policy was the greatest uncertainty faced by the industry. The matrix suggested that the likely best-case scenario for mining was Scenario 1 in which generous geological conditions coexisted with enabling mining policy. The worst case was Scenario 3 in which increasingly difficult geological conditions (very deep gold seams and limited natural gas reserves, for example) coexisted with a hostile mining investment climate.
With the matrix plotted, we now have direction. We know that the future will fall within one of the four quadrants of the matrix. But we
