Part 1, An Introduction
I’ve spent the better part of the last four years working on approaches to online scenario planning as part of my PhD. During this time I have designed and implemented three systems – each of which explored a different approach to crowd sourcing, engagement and online participation in futures work. I call these experiments in “large-scale participatory futures systems”.
Over the next few weeks I will be writing these up for the blog. If people are interested, I might go back and explore some of the initial inspirations for these systems, as well as speculate on where this area might be going over the next few years.
The Three Systems, an Overview
The first system is called “Futurescaper” and was developed in partnership with the International Futures Forum (IFF), Tony Hodgson and my friend Nathan Koren. This was piloted on a project for the UK Government, exploring secondary and tertiary impacts of climate change.
The second system is called “SenseMaker Scenarios“. This uses a customized version of Cognitive Edge’s SenseMaker Suite to aggregate micro-stories about the future into themes and patterns for scenario generation. This was done with Dave Snowden and Wendy Schultz, and was unveiled at the 2010 RAHS conference in Singapore.
The third system is called “FogCatcher”, and was developed with Anab Jain and Jon Ardern from Superflux. This was based on a modified version of Jerome C. Glenn’s futures wheel, combined with a “hot or not” style cross-impact analysis engine. As before, this approach benefited greatly from previous conversations with my colleagues above, but also from others such as Andrew Curry of the Futures Company, Emile Hooge of Nova7, Indy Johar of 00:/research, Vinay Gupta and others.
All three projects are still in continuous development and available for experimental project use.
Futurescaper uses a Wikipedia-style approach to collect trends and weak signals. It was designed to explore specific questions about the future through a structured, online data entry form. It soon became clear that it was actually more useful at dumping a ton of apparently unrelated data into the system, coding and tagging it, and then seeing what patterns emerged.
How it Works
Futurescaper never quite made it so far as to generate specific scenario themes, but it did help identifying and prioritizing driving forces and uncovering surprising connections between them. The core strength of Futurescaper is how it helps you explore the connections between forces and factors that might influence the future.
The way it works is that individuals come across a trend, a piece of research, or a news item that they think is worth taking note of. Then they use a simple a simple web form to log it and code it in the system. The form has spaces for the title, an abstract, a description or body of text, web-links to the original, subject tags and a few other things. It’s pretty ugly, but functional.
Users can start to map out how each trend of force relates to other forces through linked subject tags. When a user enters a bit of data, they are asked to create a list of things they think influences that trend, as well as the things they think are impacted by it. To make this easier, we used a predictive text engine which draws from a database of futures-related key words that I pilfered from the UK Foresight’s online database of trends and drivers, Sigma Scan.
This is done with a simple 2×2 form, as pictured below.
Once there is enough data in the system, you can start to explore the trends and linkages between them. How does water use relate to deforestation, for example? How strong of a link is there between natural disasters and political change? What impact would the introduction of electric cars have on cities and towns?
There are two ways to do this. The first was is what I call the “bottom-up” way. This way you start with a single data point or subject area, then explore what kinds of other forces and facts are linked to this subject. Starting with a single trend or news item, like, “Mexican Zetas win firefight with State Police forces”, and you can seen what other trends and news items are related. You can also start to see what kind of impacts and influences the trend or news item might have.
Futurescaper does this by letting you click through the network of causation, based on the relationship strength between trends and forces. The system uses a kind of instant run-off approach to evaluate connection strength between nodes, based on how many data points are related using similar key words. The cool thing though is that it explores the connections of those connections, so you actually get a networked effect instead of just immediate influence.
The other way of exploring trends and themes around a particular issue is to start with the connections between areas, the so-called “top-down” approach. This was developed at the request of Tony Hodgson and is based specifically on the IFF World Model.
When using a top-down approach, you can select any number of high level themes such as “water”, “energy” and “governance” and then Futurescaper will generate a ranked list of all the data and trends which influence or connect all three. So instead of going from a specific trend or force and seeing how it links to others (the bottom-up way), you can pick high-level relationships and see what data points and trends are strongly connected across these themes. This turns out to be particularly useful in the context of thematic research, as opposed to more exploratory, unguided research.
What it Does
Futurescaper seems to be most useful at the early phases of scenario creation, specifically in horizon scanning and thematic analysis. Because it requires a bit of data-entry, it seems to work better when you have a dedicated team or group of interns whose job it is to actually scan for relevant news items, data sources and trends, then input them into the system. It takes about 30 seconds or a minute to enter each trend or force into the system, and the experience of doing so isn’t very fun or interesting. But if you have such a team (or a lot of patience), it’s actually a great way to store data you collect during horizon scans and environmental surveys.
Aside from storage, it also provides a huge kick-start to the thematic analysis side of things. You can use the top-down or bottom-up approach to really quickly generate strong arguments (with evidence) for how trends and forces relate together and, with a bit of creativity, begin to piece together emergent themes and issues.
What it’s not so good at though is actual scenario building. There aren’t any functions for taking the connections and themes you uncover using Futurescaper and turning them into scenario logics or scenario narratives. The approach is fundamentally inductive, which means that a talented futurist or scenarios practitioner can easily take what’s in the system and turn them into scenarios. But it doesn’t explicitly draw out uncertainties or scenario grids (for the more classically minded corporate futurists amongst us), so therefore requires a lot more ingenuity and creativity than the standard 2×2, workshop-driven approach.
Finally, it’s also not very good at the social side of things. We have a user log-in system which, in theory, should allow people to start to commenting, tagging, and evaluating trends and forces within the system. But the user interface is clunky, ugly and not very fun, so the chances of the general public actually integrating this into part of their daily activities is practically nil. That means it’s more of a expert-driven analytical tool.
There are two immediate improvements that would make Futurescaper more useful to the professional futurist. First, improving the user interface and cleaning up the back-end code. Right now it works very well, is highly functional, but is kind of slow and cumbersome. It also isn’t pretty or intuitive and would benefit a lot from some very simple UX love.
The second big area of is to integrate it with social media, which would make a huge improvement. The idea is to turn it into a kind of repository for interesting things which people flagged off-system, i.e., through Twitter, through an email, etc. So now in addition to retweeting interesting news items and links, you could add a special hashtag that Futurescaper searched for and would then hoover up into the system. Right now it would be too difficult to actually expect the system to be able to process and make sense of the link you sent. But it could quite easily scan key hashtags into a kind of dedicated “holding area” for future processing by system operators.
This would be a great first step towards the “social scanning” concept pioneered by Alex Pang, Wendy Schultz and others. The idea is that people tag things they find interesting in their day to day research and correspondence, and these things go into a central storage place somewhere which anyone can access. As people start investigate real topics or trends as part of a project, they can draw on this clearing house and bring together the input of hundreds or thousands of different participants and inputs built up over time.
This takes advantage of the geometry of activity on the web – lots of people doing many, many small things in parallel – to create greater scanning breadth, diversity of input, and larger participation, but in a way which isn’t a burden to any individual or group.
Futurescaper reflects a lot of my early ideas and conversations in this area, which I explore in this mock-up video from a few years ago. Although it began as a true, wiki-style project, it evolved into something much more targeted and focused on the early-stage scenario creation. In doing so, it lost a lot of the social value which a wiki has, but gained a lot more interesting functionality and features.
Right now I am working with friends at the IFF to explore how it might be made more useful in practice. I suspect that a system like this one might be useful for large organizations with a dedicated mandate for environmental scanning, and the staff to do it. But it is not particularly social in a way that encourages mass participation yet, so I suspect it wouldn’t get much use as a commercial product or web service. That said, with some of the modifications mentioned above, it could be developed into something more like a social scanning engine with more “ambient functionality”.
In Part 2 of this series, I will present the second system design for crowd sourced scenario planning, called “SenseMaker Scenarios”. This was done with Dave Snowden and Wendy Schultz and focused on both thematic generation but also early stage narrative construction, with great success.
Futurescaper: An Overview
What it is
- A system for collecting and exploring trends and factors and integrating them into emergent themes and issues.
- A structured mechanism for collecting input and data from a distributed, loosely connected groups of participants.
- An inductive way of generating scenario themes and logics that doesn’t require deductive, 2×2 grids, lots of interviews or workshops, or a very big budget.
What it isn’t
- An opinion survey, a prediction market, an ideation tool or an ideas contest.
- A scenario narrative generation engine.
- Very social or fun.
Not sure if it or isn’t yet (but I doubt it)
- A collective debating platform
- A consensus-building platform.