Assumption-Based Planning

Assumption-Based Planning, or ABP, is an approach to strategic planning originally developed by RAND for the U.S. Army. It is a fascinating methodology that can help bake agility into planning to reduce overall risk. The concept was introduced by Dewar, Builder, Hix, and Levin in their 1993 book, Assumption-Based Planning. It was then further developed by James Dewar in his 2002 book of the same name. Here we will give an overview of the approach and discuss each of its five steps.


ABP is not a substitute to planning. It is more a way of analyzing an existing plan in order to improve it. Any plan ABP is applied to should be further enough developed that it is already making tough concrete choices, but not so far along that it cannot be revisited and improved by analysis. For ABP to be effective, it needs both leadership buy-in and participation. If you don’t have this, then it is not likely that plan will be revisited based on the output of the exercise, which renders such effort waste.   

Every plan is based on a set of underlying assumptions. The farther in the future you plan, the riskier these assumptions become. Plans that are evenly distributed across time incorrectly assume that risk is evenly distributed across time. ABP therefore stresses a focus on near-term actions to reduce risk and keep options alive while deferring other decisions until uncertainty is reduced. This insight, Dewar et al. argue, should lead one to distinguish between “means-uncertain” and “ends-uncertain planning”.

In the near term, desired ends should be known. The focus is on experimenting to discover how to achieve such target outcomes or ends. This is the domain of means-uncertain planning. The farther out you go, however, the less likely you will know what target outcomes should be. This is the domain of ends-uncertain planning. This helps you become more strategic since, as Dewar et al. put it, it places the focus squarely on the contemplation of alternative ends, thereby helping you both shape the world and hedge your bets.

Step 1: Identify Important Assumptions

As described by Dewar et al., strategic planning is about doing what can be done now to prepare for an unknowable future. The assumptions you are making frame your approach. They form a conceptual ecosystem in which only certain plans take root. You don’t want to put all your eggs in one basket, and you don’t want to put all your decisions in the same narrative. Step 1 in ABP is to identify the important assumptions at play within the planning time horizon, the timeframe relevant to the plan being analyzed.

In Dewar et al., an important assumption is any whose negation would lead to a significant change in the plan. Dewar’s later work instead calls these “load-bearing assumptions”. Explicit assumptions, the assumptions that are easy to talk about, are usually not the most important, the most likely to be disrupted. What you’re really after are the important implicit assumptions. Dewar et al. point out these are often baked into forgone conclusions, into what is simply glossed over as “fact”. Whether implicit or not, however, assumptions are also constraints. The most valuable assumptions to identify cause surprise when surfaced. They constrain your thinking in ways you didn’t realize.

All plans assume a particular world—or scenario—to which they are well suited. This is sometimes called the “ghost scenario” as it tends to go unseen. Such a presumed future goes hand in glove with a lack of counterfactual thinking. This is omission neglect, ignoring what should instead be done if the world turns out to be otherwise. Consider, for instance, your vision statement. This should set out where your organization would like to be at least several years out. This vision needs a strategy to get you there. Whatever it is, it is assuming a given future in which it works as a strategy. This is the ghost scenario.

To apply ABP, take your existing vision, strategy, and roadmap and start thinking of them as a set of decisions. That’s all a plan is, after all, a series of decisions. In general, the better-conceived the plan, the clearer the assumptions made about the future. Often, however, plans are focused on actions to be taken and the assumptions made are implicit, which necessitates discovery.

The trick then is to take statements revealing the choices made and unpack what they imply about the world. Especially helpful are statements that communicate concrete choices. (“We will do A and not B.”) In what kind of world would these be smart decisions? What would need to be true about the world for this plan to be a good one? Elicitation can be done using an interview protocol. Dewar suggests beginning by asking for general impressions regarding the plan, the idea being that if people need to vent this will get it out of the way up front.  

Elicitation can also be done as a facilitated activity. One option could be to have a diverse group of subject matter experts brainwrite (freelist) responses. You can then cluster the inputs and label resulting categories. These can be sent back to participants, asking them to add any new assumptions that come to mind after seeing the categories that emerged. Now, ideally with leadership, rank the assumptions in terms of importance. For now, take two or three of the most important assumptions, and no more, and move into Step 2, vulnerabilities.

Step 2: Identify Vulnerabilities

As Dewar et al. define it, an assumption is vulnerable if something could plausibly cause it to fail within the timeframe of the current plan. This timeframe is important as, given enough time, all assumptions are ultimately vulnerable. Thus, to identify vulnerabilities the focus is placed on how the world may change within the planning time horizon in ways that would impact the plan. These alternative futures are what the U.S. military sometimes calls “worlds” and corporate planners call “scenarios”.

When identifying possible vulnerabilities, it’s not about your best guess of what is likely to happen. It’s about what could happen. The more diverse the inputs the better the overall results. A helpful technique here is the Delphi Process, also developed by RAND. (In his later work Dewar refers to this as the “Rip Van Winkle technique”.) This technique helps groups generate lists of possible vulnerabilities without the contamination of dominant group members.

The framing of the exercise helps the group ideate possible changes to the environment within the next several decades that could impact current the plan. As Dewar et al. note, the restriction to yes-or-no questions keeps participants focused on concrete aspects of the future. Banning questions contingent on prior answers keeps people focused on long-range issues. To apply this, take each assumption carried forward from Step 1 and have a diverse group of experts ideate ways in which they might be violated.

Keep in mind that experts from outside your organization will often identify vulnerabilities insiders wouldn’t think of. Feel free to use the Delphi Process or any other facilitated activity. (This can also be done virtually or even via email.) This step will leave you with a list of assumption-vulnerability pairs. As with Step 1, the group, and again ideally with leadership, should then rank order the list by risk. Risk here can be thought of as importance and vulnerability.

This should not, however, suggest that assessing risk is a simple matter of creating a 2×2 matrix. Such an approach would gloss over that seemingly very different tradeoffs might still be nearly equivalent in terms of actual expected value or loss. Thus, as Dewar notes, the actual ranking that results will often come down more to an organization’s particular philosophy and values. See, for instance, the example below.  

Step 3: Define Signposts

In general, you don’t want to just wait around for assumptions to fail. That’s where signposts come in. A signpost is an event monitored as a leading indicator of an assumption’s violation. Goals and objectives are a good place to look for signposts. For example, if achieving Objective A is necessary to achieve Objective B, then the achieving of A is both an important assumption as a well as a signpost for B. Signposts, then, can be thought of as dependencies.

Dewar recommends starting with an assumption’s vulnerability and imagining the path along which that assumption could fail. A signpost then would be any indication that you are already on this path. Quantitative signposts are more likely to be tied to quantitative assumptions. Qualitative signposts are fine. As an assumption’s vulnerability changes, plans should also change. The actions taken should change and adapt. That is the function of signposts.

Step 4: Define Shaping Actions

ABP emphasizes action that can be taken now to reduce risk while deferring other decisions until later. Such near-term actions are divided into two types: shaping and hedging actions. A shaping action is any action taken to control the vulnerability of an assumption. Dewar discusses two types of shaping actions, those meant to reduce an assumption’s vulnerability and those meant to change the nature of vulnerability. An example of the former might be branding and advertising. An example of the latter might be R&D or business acquisitions.

Step 5: Define Hedging Actions

Hedging actions are taken to prepare the organization for the failure of an important assumption. Notice the difference in focus: Shaping actions nudge the world to make it friendlier to assumptions. Hedging actions nudge the organization to “hedge its bets” in case an assumption fails. Let’s say you hope it doesn’t rain on a walk. If you could alter the weather, that would be a shaping action. You can’t, however, so you resort to hedging actions, such as scheduling the walk based on rain forecasts and/or bringing an umbrella.

Recall that signposts indicate the changing vulnerability of an assumption, which means the actions taken should adapt. Plans will already have shaping and hedging actions baked into them, which means the changing vulnerability of an assumption doesn’t so much mean you are introducing such actions as much as you are changing them. Low risk or inexpensive hedging actions should often just be taken immediately. The more costly or disruptive a hedging action is, the more it will make sense to wait for particular signposts.

ABP and Scenarios

Though ABP is technically separate from scenario planning, it nevertheless greatly benefits from it. The prioritized assumption-vulnerability pairs that come out of Step 2 can easily be fleshed out as scenarios. If doing so, you don’t want more than several. As Dewar et al. note, six scenarios will become unwieldly, whereas fewer than three suggests a lack of creativity. Keep in mind these add up quick. If you only prioritize two assumptions from Step 1 and then choose two vulnerabilities for each, that is already four scenarios.

Scenarios help readers feel an important assumption failing. They flesh out details and help suggest near-term actions and signposts deemed compelling evidence that this future is approaching. To achieve this, as Dewar notes a scenario must be both relevant to the current plan and credible, meaning the people using them for ABP must find them believable.

This is an iterative process, with priorities regularly reconsidered given new discoveries. For instance, considering what could go wrong (vulnerabilities) is necessary to generate scenarios, and the process of writing scenarios in turn helps uncover additional ways things could go wrong. This cycle can also surface additional implicit assumptions, which may even be deemed more important than the ones already prioritized.

It is important here to distinguish between the role of planners and leaders. As Dewar et al. put it, planners, with leadership’s help, identify threats and actions to counter (pivots). Leaders, however, decide which planner-identified actions to take. One of the most important roles of planners is to produce a small and robust set of alternative actions to be considered in the near term. In identifying responses (shaping and hedging actions), only consider one scenario at a time. Look for actions with the lowest cost-to-risk ratio. In the first planning cycle consider only the most urgent of issues. Other issues can be considered in subsequent cycles.


OK, to close (for now!), Dewar et al. note that while these steps do not necessarily have to be done in this order there is nevertheless a sort of logical flow to the process. First, to apply ABP, there must be an existing plan, which, whether acknowledged or not, is based on an underlying set of assumptions. Plausible events or changes can occur within the planning time horizon that will violate important assumptions, which means they are vulnerable.

Identifying vulnerabilities helps indicate signposts and shaping actions. Signposts of vulnerable assumptions, as well as assumptions that have failed, indicate changes in hedging actions.

The magic of ABP lies in its approach to agility. By confining ends-uncertain planning to near-term actions and signposts it eliminates much of the downside of the traditional planning process. Traditional planning demands details out the end of the planning time horizon, which, sans a real crystal ball, is nothing but waste.

There will be more to come on scenario planning in future articles.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: