Signal over time - a workshop format (Beta)

This workshop will help you to map out the behavior over time of “big-picture” trends that have a material impact on society or at a smaller scale on your organization.

(We jump right into the details below, and at the end of the post add some theoretical background on why this is crucially important and how this format could provide something useful).

When to use: when there is a need for:

- distinguishing the signal from the noise;

- to try and make assessments about potential future direction

Benefits: listing the confidence level of the trend at hand gives us more useful information. Stakeholders can see variability and also steady direction of a trend over time

Instructions

Set an overarching question/inquiry that serves as a topic for exploration.

Have a group of at least five participants at a table, ideally with different background, cognitive style, area of expertise. Set aside ample time (3 hours for the entire process). Ideally, steps Trends, Sentiments, and VIPPP could have some pre-work to be researched upon; then the group comes together to share and build collective knowledge from a variety of sources and perspectives.

·      Trend over time & Iconic stories

The group would map out the trend over time and populate it with big-picture data and with iconic stories. If the trend has quantitative aspects (price, pollution levels, market share, etc) then they should be visualized as well on a long-enough timescale. Iconic stories provide a narrative depth of major shifts, successes, collapses, etc.

·      People’s sentiments

Mapping out on the wall people’s attitudes towards this trend, both in magnitude and strength. Using an ethnographic approach to see qualitative + quantitative data NOT only from surveys, e.g. micro-narratives, twitter trends, reddit comments, et al. These would be mapped out on the same wall with post its of a different colour code

·      Is it systemic?

• Map the knowledge you have about the trend’s systemic nature (if it has one)
• Despite oscillations, is there a direction that can be clearly seen over time? Discern variability vs long-term trend performance if the two don’t match
• Are there in-built systemic forces for this trend to continue in a certain way?
• Here, use Geels’ Transition Theory to chart what “landscape” and “regime” forces may be driving the trend. Are there natural / physical limits that cap it (say, exploitation of water sources in an area?)

·      Black Swans (as a proxy for fragility to them)

Have the entire group brainstorm a list of potential black swan events. Such are scenarios that involve an extreme risk no matter how small the probability. The outcome of this part is a) to list some of the possible extreme events and b) gauge how “fragile” the trend is to a vast (or small) number of such events. Are there reverse black swans? Events that could be incredibly beneficial to the trend?

•  Part one: Brainstorming will happen individually first, then post its will be shared in plenary.
• Part two: after this, the group will dialogue around the trend’s “fragility” as in the susceptibility to such extreme events (or, exposure to extremely beneficial rare events).

·      VIPPP (Vested Interest’s Past Prediction Performance)

Map out how the trend’s most staunch defenders (likely, those who have a stake in it) have fared in making predictions about how it would go over time. Have the supporters of this trend been right in their past forecasts? Have the critics been more accurate in theirs instead? What does that tell us about the strength of their theory about how it will evolve next?

·      Epistemic base

This can contribute to all the other steps of the process. What confidence level do we have about the knowledge base of the data points? Are they all coming from industry reports? Are they all coming from a source which may not gather complete or accurate information? Are there conflicting sources that disagree with some of the core facts? Here the group will list the known knowns, the known unknowns and the unknown unknowns on a whiteboard in three categories that will have different name tags but will be on a continuum.

Theoretical background, and why it is important:

The FSSD (Framework for Strategic Sustainable Development) provides a theoretical basis for seeing how important it is to work with risks from unsustainable trends (and our reliance on them as an organization).

Using stories and narratives with an ethnographic approach to map people’s sentiments seems a more viable source of data compared to opinion polls (no one saw Trump’s true popularity in opinion polls because, well, we refused to read the signal until election night). Sentiments are important to map because they shape the acceptance and adoption of a particular trend. Cynthia Kurtz’ Working with Stories can be a great resource for that, along with recent approaches to capture weak signals in culture. They are made of a) attitude in favour / against the trend, which you can read in a poll and also b) strength of that feeling. It’s an important distinction because only a minor percentage of Americans strongly oppose gun laws, but they are incredibly passionate about their opposition.

To test whether a trend is systemic, simply charting behaviour over time could suffice to find out its course in spite of oscillations. With regards to its societal demands, Geels idea of landscape pressures in Transition Theory could help gauge that.

VIPPP could be sobering, and stems from the idea that a theory is stronger than another insofar as it is better able to predict events. While this gives no guarantee about the capacity to predict future events, it could say something about how strong of an understanding we have about the trend if predictions are consistently off / or consistently accurate. Such a track record is informative.

The idea of mapping out Black Swan events comes from Taleb’s books, and in general his critique of how much we underestimate extreme events until after they occur (the so-called hindsight bias, aka retrospective coherence). There is no point in trying to predict them, the best that we can do is to guess how fragile we are to such extreme events (e.g. we know that the internet is more robust than the air-traffic system of a major airport because its web is more distributed, redundant and able to absorb shocks). More work will be published soon on specific formats to gauge our organizations fragility in the face of potential extreme events. What could the system do in response to such events? How vulnerable does it appear to a variety of these events? 

Fundamental caveats.

• There is value in charting a direction over time and there is a risk in believing that things will unfold in a similar trajectory in the future. Be aware of the desire to forecast and of the limited realm of applicability of this format.
• Picking one trend at a time is a way to isolate one variable, which inevitably leaves out co-occurring trends. Better yet to chart multiple trends.
• Most productive of all would be to chart a trend over time where you can severely critique the data and potentially falsify the initial hypothesis that it is going in that direction (if you can't argue for the opposite to be true even in principle, you are falling for what I call a cheap "inductivist trap")

I hope this can help and inspire some of your thinking and practice.