Scenario building, also known as scenario analysis, is a crucial methodology to anticipate and prepare for the future. This is a method used from risk management to strategic foresight through early warning systems. More broadly, it is a key tool for all anticipation that needs to be actionable.
The higher the uncertainty, the more important it is to be able to mitigate risks to develop winning responses. Thence, it is crucial to use good scenarios to be truly prepared. It is thus key to build valid scenarios.
Good scenarios are sound methodologically and include knowledge and understanding of the issue at hand.
In this article, you will find a list of points – necessary conditions for the validity of scenarios – that you can easily check to verify the scenarios you are about to use are correct methodologically. If they are not, if they include methodological errors, then this means that the scenarios are flawed. Thus, you cannot use them to build robust answers, even if the content of the scenarios shows state of the art knowledge of the issue. For example, you can have invalid scenarios that nonetheless reflect great understanding of China, the U.S., quantum technologies, the Islamic State or the COVID-19 pandemic. Even though the understanding related to the subject matter is good, the scenarios may be useless if they are incorrect methodologically.
- The Red Team Analysis Weekly – 29 April 2021
- Strategic Foresight, Warning and Intelligence Products and Documents
- The Chinese Fishing Fleet, Influence and Hunger Wars
- Why the Messenger Got Shot and how to Avoid this Fate
- Early Warning Systems & Indicators – Training for the ESFSI in Tunisia
- Losing Texas to Climate Change and the COVID-19?
- Communication of Strategic Foresight and Early Warning
- Are your Strategic Foresight Scenarios Valid?
- Take the test
- Check list
- 1- Are the scenarios covering the whole range of possible futures?
- 2- Are the scenarios mutually exclusive?
- 3- Are the scenarios dynamic?
- 4- Are the scenarios on the same time horizon?
- 5- Is there a likelihood estimate provided for each scenario?
- 6- Did I provide the right means for the building of this set of scenarios?
Below you will find first an online test for evaluating the methodological validity of your scenarios. We shall suggest possible ways forward according to the test result.
Second, you will find each point of the check list explained. We shall highlight why each point matters to users. Then, we shall explicate why making sure that each necessary condition is respected is most often a guarantee of methodological soundness, or, alternatively, why not respecting a condition is a warning signal for users.
This check list will be useful to both users and practitioners.
If you are a user of scenario
As user of scenarios, if you did not build the scenarios, and especially if you do not master the intricacies of scenario-building, the check list will help you evaluating easily the methodological validity of the scenarios.
If you are a scenario-builder
As practitioner, this check list will help you verify your work as early as possible in the process of scenario-building. As a result you will make sure you are building proper, state of the art scenarios.
What if you do not use scenarios?
Your strategy, policies, actually the whole array of your responses, depend on scenarios. This is true even if you think you are not using scenarios. When you decide upon something, it is because you already have and use, mentally, a model of how the future will unfold (eg. Epstein, ‘Why Model?‘, 2008). This model is a set of scenarios of a sort. It is, however, implicit and created without any methodology. Notably, being implicit, it may be prey to many biases (see online course on modeling, module 2).
As implicitly you use a kind of set of scenarios for the decisions you are taking, then you may also use the test for your vision of the future.
Test your scenarios
Nota: We only take your email address so that you can receive the test results by email. We do not use them for anything else. If you want to subscribe to know when we publish new article, use this form. If you want to become a member, access is here. We shall use anonymised statistics about results to improve understanding on scenarios and their use.
Once you have completed the test, click on submit and read the answers for each question. We suggest what to do with your set of scenarios if it is not valid, according to the type of problem met. You will also get your overall score.
Check list for valid scenarios
1- Are the scenarios covering the whole range of possible futures?
Scenarios must cover the whole range of possible futures. Rutz, McEldowney and Taylor depicted this very well in the drawing on the right hand side (1986, quoted in Taylor, 1993: chapter 1 & fn 7).
Plausible or possible futures?
Taylor focuses on plausible futures. We prefer looking at possibility rather than plausibility.
Indeed, the idea of plausibility incorporates many biases (see online course on modeling, module 2). In other words, if you focus on what appears to you as plausible, you may very well fall prey, for example, to political correctness, to group think or to normative judgements, among others.
As a result, your scenarios could depict how you would like the future to be rather than consider all futures. This would then increase the likelihood of surprise, when our aim is to reduce the odds of surprise.
Why does it matter to you?
It is truly crucial for you that your scenarios cover the whole range of possible futures, because you do not want to prepare responses that would completely forget one scenario or more.
If we take the example of the Brexit before the vote, imagine that your experts considered that Brexit was impossible. As a result, they gave you a set of scenarios that did not include the possibility for Brexit (e.g. it was thought as implausible). They could have built a couple of scenarios focusing on new rules between Britain and the EU (they could create more than one scenarios), another scenario describing similar rules between Britain and the EU, and finally a last scenario depicting a new love story between Britain and the EU. As a result, you started creating a whole range of answers, made corresponding investments, developed policies etc.
Then, the vote took place, and people chose… Brexit. And you found yourself completely unprepared.
This anecdote is exactly the reverse of what we want to obtain with scenarios.
The brief history of scenarios on the COVID-19 pandemic is also replete with instances of scenarios that did not consider the whole range of possible futures. Indeed, the people and firms creating the scenarios at the time focused wrongly on what they thought was possible. As a result, many times, governments and actors were given invalid scenarios that forbid preparedness.
We want to have scenarios that show us the whole range of possible futures so that we can be ready for any future.
Why is it a methodological guarantee?
As you know, scenarios are built from the combination of the values or attributes of the variables selected to “represent” your initial question. The why and how these variables are selected is outside the scope of this article (see our course on scenario-building).
Mathematically, the attributes or values of a variable must be exhaustive, i.e. cover all the possible values the variable can take in reality, as in the quantitative example on oil prices on the right hand side.We covered this in detail in the online course on modeling, module 4, explaining how to identify factors and drivers for a question and how to make sure these were variables.
Thus, if you build scenarios out of a combination of values (of variables), which are exhaustive, and if you selected properly your variables in a way that is representative, then, as a result, the set of scenarios you obtain covers the whole range of possible futures.
If the experts that built the scenarios use a proper methodology, automatically, the scenarios they obtain covers the whole range of possible futures.
If the scenarios do not cover the range of possible futures, then you may ask them why it is not so. They may have a very sound answer, but it is better to check with them. If they do not have a sound answer, then be wary.
At best this set of scenarios will help you think out of the box, and provide you with new ideas. However, you have to be very careful before using the set of scenarios to develop strategy and policies.
The potential for surprise is not as mitigated as it could have been.
2- Are the scenarios mutually exclusive?
Mutually exclusive scenarios mean scenarios that cannot take place at the same time. You can have one or another, but not both at the same time.
Why does it matter to you?
This condition is necessary to cover the range of possible futures. This is necessary so that you are prepared across uncertainty.
Imagine that you are an emergency NGO preparing your budget and material for the next year for, say Afghanistan. If experts, in the set of scenarios they give you, present you with one scenario for war, one for epidemic, and one for earthquake, but not with one scenario for epidemic and earthquake, one for war and epidemic, one for earthquake and war, and one for war, epidemic and earthquake, then you may run into serious problems. You will not be prepared for complex emergencies. You will have neither the material nor the funding for such cases.
On the contrary, what we want to achieve with scenarios is to be ready for any circumstances, including complex emergencies. We want to be able to develop policies that are robust across all possible futures.
Why is it a methodological guarantee?
As previously, this characteristics of scenarios derives from those of the attributes of the variables. The attributes of a variable must be mutually exclusive. As a result, your scenarios will also be.
If the experts use a proper methodology, then their scenarios will, automatically, be mutually exclusive.
If their scenarios are not mutually exclusive, then it means that there is a serious flaw with their methodology. Actually, I cannot truly imagine a way to salvage such scenarios.
At best this set of scenarios will help you think out of the box, and provide you with new ideas.
It would, however, be dangerous to use exclusively this set of scenarios to guide strategy and policies.
The potential for surprise is not truly as mitigated as it could have been.
3- Are the scenarios dynamic?
Scenarios, to be more easily actionable, i.e. to allow you to develop a proper set of responses and actions to handle coming changes, must respect the previous characteristics – exhaustivity and mutual exclusivity – and ideally should also be dynamic.
Why does it matter to you?
Scenarios are also there to help you identify crucial key points, where decisions are needed. In these cases, scenarios may develop into sub-scenarios.
“A scenario is a story with plausible cause and effect links that connects a future condition with the present, while illustrating key decisions, events, and consequences throughout the narrative”.Glenn, Jerome C. and The Futures Group International, “Scenarios,”
As a result, scenarios unfold as a story about the world, which is often presented as a narrative (see Scenarios: Improving the Impact of Foresight thanks to Biases). Stories and narratives are essentially dynamic.
Why is it a methodological plus?
Here we are less in the realm of a methodological necessity and guarantee. Dynamic scenarios enhance the actionability characteristics of scenarios.
If your scenarios’ expert was able to point out causal dynamics, then this is a guarantee of his or her skill at scenarios building and knowledge and understanding of the issue at hand. It means that those who built the scenarios truly thought through them, stress-tested their understanding and endeavoured to explore as many areas as possible.
If your set of scenarios does not explicitly show dynamics, even though this set of scenario may not be as actionable as hoped for, it may nonetheless, according to other answers, be used to guide strategy and policies.
4- Are the scenarios on the same time horizon?
In a valid set of scenarios, you should get scenarios that each depict the same period of time. The authors of the scenarios may choose to develop more or less this or that part, but, nonetheless, all timeframes must be covered for each scenario of your set of scenarios.
Sometimes, you will be given scenarios that are not on the same “time plane”. This is not right.
Why does it matter to you?
If scenario A, for example, depicts a situation starting from now until the end of year 2, and scenario B depicts what happens between year 2 and year 4, and you are given only these two scenarios, then you have no way to know if you are given two scenarios… or just one.
Scenario B may just be the continuation, in time, of scenario A.
What you should get, in this example, is scenario A and scenario A1 (and probably A2), the last two portraying what is happening in the case of scenario A from year 2 to 4. You should also be given scenario B0 that tells the story of what is happening from now to year 2 and led to scenario B. This would be a proper set of scenarios time-wise.
Why is it a methodological guarantee?
It shows that your experts truly master both methodology and subject matter. They are able to articulate processes and causal links.
Incidentally, this is why having a proper model for your issue is so important for scenario-building (see online course on modeling).
If the scenarios are not all on the same timeframe, you have to be very careful before using this set of scenarios to develop strategy and policies.
This set may be completely useless, or it can help you think out of the box, and provide you with new ideas.
The potential for surprise is not as well mitigated as it could have been.
5- Is there a likelihood estimate provided for each scenario?
This means an evaluation of likelihood accompanies each of the scenarios of the set. As a result, you should know which scenario is more or less likely.
We are here, of course, in the realm of estimates. Furthermore, these estimates will vary with time, with your decisions and your actions. Yet, it does matter that each scenario has an estimate of its likelihood.
Why does it matter to you?
Knowing if scenario A has a 80% chance of being actualised, scenario B has a 19% chance of being actualised and scenario C a 1% chance of occurring is a crucial information for you.
It does not mean that you should disregard scenario C, especially if it is a high impact scenario. In that case, you should make sure you have hedged against this scenario, or developed policies that are robust enough across all scenarios.
One of the rationale behind scenario-building is that it should help actors envisioning the future beyond “business as usual” trends. Thus, presenting all scenarios (including, if relevant, a “business as usual” one) with likelihoods that are spelled out will help actors considering all possibilities. With likelihoods, scenarios-builders do not have to hide a scenario – the “business as usual one” – because they fear that biases will lead users to discard other, less comfortable, scenarios, even though those are more likely. Thus, ethically, this is much better for everyone as scenarios-builders do not end up taking decisions for scenarios-users.
As a user, you should always be aware of what is likely and unlikely. Knowing this will allow you to design proper responses, according to what you want to achieve. It may imply that you would need to deploy an immense amount of power to achieve an unlikely scenario, for example. Ultimately this is up to you to decide, and knowing what to expect is crucial for success.
Finally, with probabilities, if ever you were offered a set of indicators for each of your scenario and sub-scenarios, then you could also use the scenarios for monitoring, for early warning and for steering policy. As a result, your scenarios would be even more useful and they would also last longer.
If you are not given likelihoods for your scenarios, then to be able to use this set of scenarios to develop strategies and policies, you need to make sure these strategies and policies are robust across all scenarios.
Without probabilities, a set of scenarios is less actionable than with probabilities.
6- Did I provide the right means for the building of this set of scenarios?
Here the question is about you and not about the set of scenarios you received.
Scenarios-building is a demanding methodology. It demands “deep understanding and knowledge” of the issue at hand (Mietzner and Reger, 2005: 236). It also asks to master the methodology and to know how to apply it. As a result, it is also considered as “time-consuming” (Ibid.). Actually, obtaining not only valid but also good scenarios is an investment, as scenarios can – and should – be used over time. Scenarios are more than a consumer good.
Thus, if ever you asked an expert or a team of experts, be they external or internal, to develop scenarios without giving them the means to do so in terms of resources (time and money), then it is highly likely that you will get bad scenarios.
The conditions you set frame the outcome you obtain.
Global Result and score
If you have answered yes to each and every question above, your set of scenarios is most probably methodologically sound as well as actionable.
Assuming the knowledge of and understanding on the subject matter is also good, then you can use it to develop strategies and policies.
You should however not forget that foresight, what we do with scenarios, is not prediction. There is always a possibility for a black swan event, for example (see Taleb’s Black Swans: The End of Foresight? and Useful Rules for Strategic Foresight and Risk Management from Taleb’s The Black Swan).
With this set of scenarios, nonetheless, you can be quite confident that, methodologically, you will mitigate as best as possible the risk for surprise.
On the contrary, the lower the score the more wary you should be of your set of scenarios. Questions 1,2,4 (60 points all together) are the most important methodologically and negative answers there should make you very cautious indeed. Questions 3 and 5 (30 points) are necessary for scenarios being truly actionable.
Epstein, Joshua M. (2008). ‘Why Model?‘, Journal of Artificial Societies and Social Simulation 11(4)12.
Glenn, Jerome C. and The Futures Group International, “Scenarios,” The Millennium Project: Futures Research Methodology, Version 3.0, Ed. Jerome C. Glenn and Theodore J. 2009, Ch 19.
Ritchey, Tom “General Morphological Analysis as a
Basic Scientific Modelling Method“, Technological Forecasting & Social Change: Special Issue on General Morphological Analysis, 2018.
Taylor, Charles, Alternative world scenarios for a new order of nations, US Army War College, 1993.