Predicting the future is impossible, but that doesn't mean you should never think about it.

Even widespread techniques such as threat modeling are fundamentally about looking into the future and preparing for what might happen. To make the most of the time and energy spent trying to anticipate the future - and be ready for it - there are a variety of structured forecasting techniques meant to help you through the process. In this series, we'll walk through several widely-applicable techniques and show you when and how to use them.

This is part one of a forthcoming multi-part series covering different forecasting methods and techniques.


Before we start, let's go over a few of the concepts and terms we'll be looking at throughout this series.

Futures - forecasting isn't about predicting The Future - that's simply not really possible to achieve. Instead, it's about considering a wide variety of possible futures, plural, and then thinking through what each of them might mean.

Scenario - a fictional potential future that acts as a "what-if." Many forecasting techniques revolve around generating a range of possible scenarios for the future and using those scenarios to ask questions about each of the possible futures they represent. You've likely already done scenario-based planning, in the form of wondering what the future would look like if some event occurred - imagining your life if you won the lottery is scenario planning!

Scope - the scope of a forecasting workshop or session is the size of the question and the variety encompassed in it. A large scope might ask "what does the country look like in five years," while a smaller scope might ask "what is our town's government going to look like in one to three years?" Scope encompasses both the size of the question and the distance into the future it's being looked at. Changing the scope of a forecasting session significantly changes the feel of it, as well as the type and specificity of results and the time required to achieve them. Getting the scope right based on the resources available to you and what you hope to learn is a perpetual challenge in forecasting.

Indicator - an indicator is a data point you can track over time that "indicates" a particular future might be becoming more or less likely. It essentially answers the question "if this future was going to happen, what measurable things would occur beforehand?" More detail about what makes a good indicator is provided later on in this article.


Most importantly, before doing any forecasting work, identify what you hope to achieve. Forecasting is well-suited for strategic planning, but not meant for concrete answers or numerical results. It's a way of thinking about the future and answering "what-if" questions. Starting out with a clear, shared set of questions that you want answered is essential for useful forecasting work, even if the questions themselves aren't very specific! Keep in mind as well that the scope of your topic will restrict the types of questions you'll be able to answer - asking huge questions about global trends far in the future is unlikely to be able to provide you specific ideas about what might happen locally to you in the next few months.

At its core, forecasting is a way to create a set of shared stories that you can think about collaboratively as a group. Those stories can then be applied to answer questions you have about the futures those stories represent.

Simple Scenarios Technique

The Simple Scenarios Technique, as explained by this short guide starting on page 8, is a straightforward (but not necessarily simple) process. It's best suited for short to medium range forecasting - within the next few years - of complex scenarios that are influenced by multiple major factors. This technique is most useful with a dedicated facilitator and a small group of participants.

The Simple Scenarios Technique consists of three major steps:

  1. Identify key drivers
  2. Generate mutually-exclusive scenarios
  3. Identify indicators for each scenario

Let's dive into each step.

Before we get started - I've also made a handout for attendees that can help keep track of the process and record insights. It's meant to be printed on US letter sized paper, double sided, in black and white.

Identify Key Drivers

The first task, and one of the largest, is to generate a list of "key drivers" for the portion of the future you're aiming to forecast. There are a variety of ways to create this list, and the quality of the entire process depends mostly on the quality of the key drivers. One of the most common is to conduct a brainstorming session:

  • The facilitator explains the purpose of the session, and the topic being brainstormed. Make sure each participant understands what a "key driver" is, but don't be too specific - getting unusual or unexpected ideas is part of the point of this process.
  • Set a timer for 5-10 minutes.
  • Each participant writes out key drivers on individual post-it notes (or similar) during this time, without sharing them yet.
  • Once time is up, work together as a group to clump together similar post-it notes and refine your list of key drivers. Participants can keep adding drivers during this time if the collaboration sparks new ideas. Post-it notes are helpful because the group can use a wall to physically group similar ideas.

Ideally, you'll want to emerge from this process with 3-10 key drivers, depending on the scope of your forecasting session. This is a "divergent-convergent" process, where the initial brainstorm is meant to generate a wide range of ideas without the participants influencing each other or dismissing any ideas, then the group portion helps everyone converge back on a list of key drivers that captures as much variability as possible.

Generate Scenarios

With the list of key drivers in hand, the facilitator should lead the group in generating a list of scenarios. Generally, you want to select 3-5 scenarios that represent a reasonable worst-case scenario, a mainline or widely-accepted analysis, a new opportunity, and up to two others that seem interesting or useful to consider. If the group is somewhat skeptical of the forecasting process, consider including a recognized or extremely credible scenario so the participants aren't faced with only scenarios they consider strange or unlikely.

Generating these scenarios is the most creatively demanding part of the process, and for most groups it relies heavily on the facilitator. What you're aiming to do is consider what the future might look like if the key drivers you identified progress in different ways. The table below, replicated from the short guide introduced earlier, demonstrates a variety of key drivers (on the left) being used to generate scenarios (across the top) for a fictional country facing a variety of problems.

Imperfect Peace Descent Into Order Pockets of Civility Fragmentation
Government Capacity + - - -
Insurgency - + +
Economy + -
Paramilitaries - + +
Civil Society +
Drug Trade +
External Actors

You can read the first column of this table as asking "what if government capacity strengthened, the insurgency weakened, the economy strengthened, and paramilitaries weakened, while the other key drivers remained neutral?" Using that set of factors, the "imperfect peace" scenario can be imagined and sketched out. What would the imperfect peace look like? What impacts would it have? How can we be ready for it? What can we do to make it occur (or not occur)?

Identifying the list of scenarios you want to consider is extremely difficult and open-ended. The number of scenarios to create depends largely on how much time and energy you have to analyze them, as well as your ability to organize them all and keep them clear to participants (this is where something like access to a whiteboard in the space you're using can be important). Each scenario must also be sketched out in enough detail that everyone can grasp them and understand what the scenario "looks like," while also remaining mutually exclusive - each scenario needs to stand on its own, and any of the scenarios occurring should mean the others can't happen. Generally, 3-5 scenarios are about the limit of what a group can handle without either long breaks or a multi-day process.

Each scenario selected for use needs to balance realism (the "alien invaders" scenario might be interesting, but is probably not very useful for thinking about the future) and the value of diverse futures for analysis (if every scenario is essentially what everyone expected to happen anyway, there's not a lot of value in thinking through them all - they're all just the same future with small tweaks). As demonstrated in the table above, some key drivers may end up not being used - consider what key drivers might not matter as much for the analysis you're trying to do.

Generating these scenarios is difficult, but it's where the real work of forecasting lives. It relies on the group's ability to think about the future and what it might look like, and then consider the implications of that future. It's difficult, and it can be frustrating, but it's a skill that improves with practice, and the results of spending an hour or two thinking over a set of future scenarios can be incredibly useful.

Identify Indicators

Identifying and thinking through the scenarios is valuable in itself, as it gives the group a great way to anticipate and prepare for a variety of possible futures. However, going a step further can make the process even more valuable by giving you some early warning of which scenarios might be more or less likely to occur over time. This final step is identifying indicators for the scenarios you identified. Indicators are a consistent way to track your scenarios out in the real world, and each should be:

  • observable/collectible
  • valid (actually tied to the scenario's likelihood)
  • reliable (objective, consistently reported)
  • stable (consistent over time)
  • unique (measures only one thing)

The full set of criteria are aspirational; the first two are essential, the third and fourth are very important but can be difficult, and the final characteristic is extremely difficult to guarantee.

To generate indicators, work through each scenario and ask "what would have to happen for this scenario to become reality?" From that, identify roughly three to ten indicators that meet the criteria above. These indicators provide you a set of fixed data points that you can track over time and can provide early warning of which scenarios might be more or less likely to occur. By consistently reviewing these indicators (ideally on a fixed schedule - every month or every other week or similar), you might see the indicators for one specific scenario begin to all point toward that scenario coming to pass, giving you advance warning to start preparing or engage in more in-depth planning.


Using the Simple Scenarios Technique of identifying key drivers, generating a set of scenarios, then analyzing and collecting indicators for those scenarios, you can prepare for a variety of potential futures and use indicators as a tool to gain early warning before they become reality.


Grabo, Cynthia M. Anticipating Surprise: Analysis for Strategic Warning. Center for Strategic Intelligence Research, Joint Military Intelligence College, 2002.

Pherson, R. H. "Strategic Foresight - Nine Techniques for Business and Intelligence Analysis." (2016).

Pherson, Randolph H., and Richards J. Heuer Jr. Structured Analytic Techniques for Intelligence Analysis. CQ Press, 2020.

Smith, Scott, and Madeline Ashby. How to Future: Leading and Sense-making in an Age of Hyperchange. Kogan Page Publishers, 2020.