I have an article in the ‘Scenarios Symposium’ published in the latest edition of the Journal of Futures Studies, which has just been published online. The Symposium asks whether scenarios are worth using, and why. The starting point was an article by Graham Molitor – the distinguished futurist who invented the trends ‘S-curve’ – in which he wondered why he’d bothered to use scenarios at all. I argue that the strength of scenarios – when compared to other futures tools – is that they help organisations reduce the ‘variety’ of uncertainty about the future to a point where they can understand it, process it, and respond.
My contribution (opens in pdf) draws on the early work of Ross Ashby, the cybernetician, whose ‘law of requisite variety‘ states that organisations need to match the variety in their external environment if they are to function – either by increasing their internal variety or reducing external variety. Which relates to futures like this:
Scanning, that essential tool for futures work, is a case in point; it produces so much data, much of it new, that organisations are overwhelmed by it. One of the reasons we create scenarios, therefore, is to help people interpret data in a way which allows them both to manage variety and to comprehend it.
There is a number of responses in the Symposium, and of the rest I’ll pick out a couple quickly here. Angela Wikinson, of Oxford Institute for Science, Innovation and Society, points out that scenarios work is ‘under-theorised’ and the evidence base remains thinner than it ought to be. This is a recurrung criticism of the field; she observes that it is much easier to gain funding to use futures methods to research an issue than it is to research the methods themselves (which could be interpreted as evidence of efficacy from a funders’ perspectives). Marcus Barber picks up Graham Molitor’s evidence challenge by providing a case study from his work (with the Australian beer producer Fosters). He also observes that scenarios are a method to get to better questions:
The universal futures tool is the Question and Scenarios are nothing more than an elaborate questioning technique. They exist to assist us to discover ‘doubt’ in our own thinking.
And Elena Hiltunen reminds us that one of the most useful aspects of scenarios is that they can help us test strategy: “A colleague of mine that has worked with scenario technique for years commented that scenarios are like wind tunnels for strategy. Another car related metaphor could be test driving.” It’s a simple point, but it resonated for me because some of the more successful scenarios projects I’ve run recently have enabled clients to test strategy when conventional methods would not have been as successful.
I think that improving the theoretical base of scenarios work is going to be necessary if it is to develop and mature as a discipline, but this will take time, especially since it is still largely dominated by practitioners. But improving the evidence base (also necessary) could be done quickly and should be done as a matter of urgency. The evidence is out there, waiting to be gathered. As a futures practitioner I’m tired of reading criticism of scenarios work which assumes that the only evidence of success is some 40-year old work done by Shell and some 20-year old work (good though it was) done in South Africa.
One criticism of the journal; many of the contributors are writing in English as a second (or third or fourth) language, and their contributions haven’t been edited as well as they might have been to allow for this. But in general it’s a useful snapshot of some ideas about current scenarios theory and practice. A smart piece of editing by Sohail Inayatullah to have picked up Graham Molitor’s original article and made this opportunity from it.
The whole thing can be found in the February 2009 edition of the Journal of Futures Studies. Update Or the Symposium can be downloaded as a standalone supplement by clicking on the pdf below.
I’d only take issue with one point in your argument, Andrew. That is that ‘scanning produces too much data’. I’d argue, only if you let it.
First scanning should be focused on the vital few, second electronic tools can increae the power of data and amplify to information and knowledge. Scenarios for me can be improved by applying more robust scanning and asociated evidence bases. There are emerging tools which we are working with that allow this process to be seamless.
Andrew, thanks for v.useful guide and summary and your own piece which is fabulous. Anyway, my comment: this idea that “the only evidence of success is some 40-year old work done by Shell and some 20-year old work done in South Africa…” is the view propagated by the folks who did that work. It’s just GBN/Generon marketing. (Btw, the Mont Fleur scenarios were, at best, a restatement of the obvious – speaking as one who lived in SA through the 1980s and early 1990s – and made absolutely zero difference to govt, civic, party, or corporate initiatives, or real outcomes.)