STR.4 - Construct Scenarios for Strategic Choices

Type: Method pattern Status: —

STR.4:1 - Problem frame

Use this pattern when a strategic choice depends on future conditions that one forecast cannot responsibly settle. Begin with a condition that could change which direction is useful or feasible, then explain a plausible path by which that condition could obtain.

The working result is a small, justified set of future situations for the same strategic question. Each scenario connects assumptions, a causal account and consequences for choice. It helps the practitioner see where a direction works, where it fails and what a materially different future would require.

A scenario describes possible conditions. A forecast makes a predictive claim; a target expresses what someone intends to achieve. Distinguish them before using any of them to support a commitment.

Do not undertake a scenario exercise when a sufficient current comparison or a justified forecast already answers the question. A known constraint can rule out a proposal without any future story. For consequential reliance, obtain the domain evidence and professional judgement needed for the mechanisms and protected limits; an internally coherent story is not evidence that its events will occur.

STR.4:2 - Problem

Several beautifully named futures can imply the same action. Their narrative variety hides the absence of decision-useful diversity. Another set may be different but combine incompatible assumptions, such as unrestricted customer-data reuse and a restriction that makes that reuse unavailable.

A particularly tempting error turns a preferred future into the baseline and treats the others as deviations. The preferred direction then appears resilient because the scenario set was built around its success. Counting scenarios or assigning equal probabilities does not repair the bias.

STR.4:3 - Forces

Scenarios need enough structure to explain consequences without pretending to predict a detailed future. A small set is usable but can miss a shared failure; a larger set can consume the attention needed to make a decision.

Familiar developments deserve consideration even when they lack novelty. Unfamiliar conditions can expose hidden dependencies but require a clear basis for treating them as relevant. The practitioner is balancing consequential coverage, explanatory coherence and effort, not maximizing dramatic difference.

STR.4:4 - Solution

Construct future conditions around the choice they can change. Explain their mechanisms, test their differences and limits, and return the set with the decision questions it exposes.

STR.4:4.1 - State the choice and horizon the scenarios must serve

Recover the strategic subject, contribution sought, continuing obligations and decision window. Use STR.3’s formulation and unresolved questions if they already exist. Identify what the recipient might choose differently under another condition: a direction, option, scale, timing or protected limit.

Keep the time of commitment distinct from the horizon of consequences. A four-day preparation choice can concern a twelve-month direction; a future event matters to that preparation only if it changes the attainable contribution of those four days.

Reuse an existing scenario set when its assumptions, horizon and consequences fit this use. An attractive industry report about a different population or time window supplies ideas, not automatic coverage.

STR.4:4.2 - Choose the conditions with leverage over the decision

Start from uncertain developments and dependencies that can change the comparison. Examples include customer willingness and ability to buy, a necessary permission, rival responses, supply access or a shared service bottleneck. Use source evidence to distinguish an observation, an extrapolation and a deliberately posed stress condition.

Select the smallest productive set. A two-axis matrix can help when two meaningful uncertainties and their combinations account for the differences needed. Use another construction when those axes hide a third condition, create impossible combinations or force irrelevant quadrants. There is no required number or geometry.

Do not confine generation to a fixed list when a new beneficiary, problem or feasible means changes what should be compared. Return that discovery to STR.3 or STR.5. A different direction space is more than a new score on the old scenarios.

STR.4:4.3 - Explain how each situation could arise

Write a short causal account: the starting conditions, material changes, their interaction and the resulting situation at the selected horizon. State the assumptions that make the account work. Include only detail that changes interpretation, feasibility, consequences or recognition of the condition.

Then challenge the account. Could the proposed sequence occur within the time available? Do its assumptions fit together? Does it count the same resource twice, rely on an unavailable right or combine incompatible customer behaviours? Retain an uncertain mechanism as uncertain rather than making the prose more confident.

If one uncertain mechanism carries the choice, request its qualified domain answer or retain conditional branches. The scenario’s explanatory role does not substitute for that answer.

When another participant’s choice can change the comparison, connect each serious own move to that participant’s possible responses. State who chooses, what they can observe before choosing, the actions then available, and the supported or hypothetical preferences and constraints. Include the timing and means of any further response of your own. Reuse an adequate current response account; no separate interaction model is needed when such responses cannot change the present choice.

Obtain the qualified game-theoretic or other behavioral result needed for this mechanism. Use a sufficient direct result without building another model account; use PSD.5 when complementary or conflicting model contributions need to be chosen and connected for this question. In a finite sequential model with complete and perfect information, backward induction can resolve later choices before earlier ones. Its relevant common-knowledge and sequential-rationality assumptions must apply: each mover compares outcomes by their own preferences at each possible decision, including contingencies outside the proposed outcome path. Retain materially different tied continuations unless a justified tie-breaking rule resolves them. Return the resulting conditional action profile with its premises, not just a preferred final score.

A declaration alone does not establish the response. If preferences, information, available actions or behavior remain unknown, keep the consequential response branches or request the missing determination. A modelled best response is not an observation or assurance that the participant will behave that way. Simultaneous moves or incomplete information need their appropriate Method, not this sequential calculation by analogy.

STR.4:4.4 - Test diversity by its strategic consequences

Compare the conditions against the serious directions or option families. Ask which one becomes preferable, inadmissible, too late or dependent on a different capability. Keep a scenario when that difference matters. Combine redundant stories unless their different mechanisms require different observation or response.

A qualitative account is normally enough: “A full-horizon purchasing freeze removes the opportunity for a paid service trial; a data-use restriction instead changes which service arrangement is permitted.” FPF C.17 supplies the comparison basis and supported-difference discipline if a novelty or diversity claim is needed. It does not generate the scenarios or turn their novelty into likelihood.

Look for a consequential failure common to the options. If every expanding direction depends on the same scarce field-service team, variety in customer futures does not diversify that capacity risk. Examine the shared condition far enough to determine whether it changes scope, generation or the eventual choice.

STR.4:4.5 - Return the usable set and its limits

For each scenario, retain the conditions, mechanism, consequential assumptions and implication for the strategic question. Also state what the set leaves outside. A short comparison with explanatory paragraphs can carry this result.

Name the observations that would make a condition more or less relevant, and whether they could arrive in time to change the decision. STR.12 can later connect actual observation to reconsideration and authorized action. A scenario signal is not yet a standing monitoring assignment.

Stop when further credible conditions would not change the present decision enough to justify their examination. This is a bounded judgement, not a claim to have listed every future. Return to STR.5 for responsive directions or STR.9 for comparison when those are the next missing contributions. A probability-based decision remains available when its probability claims have a suitable basis; the scenario count supplies none.

STR.4:5 - Archetypal Grounding

STR.4:5.1 - SensorCo: three conditions that change the same choice

All facts here are constructed inputs. SensorCo faces declining generic AI-inspection prices, but its existing contracts can sustain current device and service obligations for the next twelve months. The Board seeks a less price-exposed repeat paid contribution while protecting that service and its reserve.

Two interested customers may discuss a bounded paid trial next quarter at a price ceiling, subject to agreement and permitted use. Four funded internal days before the next Board consideration can complete costing and specify that trial. They could instead improve device diagnostics. The twelve-month service direction has not been selected.

The team builds three conditions around what could change that preparation choice:

These are conditional accounts, not three equiprobable outcomes. The third condition does not automatically eliminate every service; the second does not imply that existing contracts vanish. Those distinctions keep the device-only rival viable and the service proposal properly conditional.

The team also checks a common failure: service expansion, device improvements and platform work may compete for the same people. The three market conditions do not establish operational capacity. STR.8 and STR.14 address the missing contributions and their joint demands.

A procurement notice that only delays one customer by a week does not establish the full-horizon freeze. A complete prohibition on every relevant service-data use can instead close that dependent service route. The team returns these discriminating observations with the set, not an automatic instruction to pivot.

STR.4:5.2 - When two stories should be one

An adviser presents “rapid digitalization” and “AI everywhere” as different futures for a professional’s next-year practice. Both assume the same clients, available time, tools and willingness to pay; neither changes a direction or inquiry. Their names add no useful diversity.

If one instead introduces a client requirement for human accountability that changes the delivery arrangement, the distinction becomes material. The practitioner retains that condition and its basis, rather than collecting more titles.

STR.4:5.3 - A provider’s promise can leave its later choice unchanged

An independent workshop is considering integration with a provider’s interface. In this constructed case, it can continue its current arrangement or integrate; after observing integration, the provider can maintain or restrict that interface. Assume these are the complete alternatives for the question, actions and preferences are known to both participants, prior moves are observed, and the sequential-rationality and common-knowledge assumptions of the model apply.

The indices below only order each participant’s own outcomes. They are not money, probabilities or comparable value between participants.

Begin with the provider’s later decision. Under these premises, restriction is preferable to maintenance for the provider: 4 exceeds 2. The workshop therefore compares integration’s induced −2 with continuation’s 0, not with the favourable 6. A prior announcement that the provider will maintain the interface leaves its available actions and the stipulated preferences unchanged. Taking that announcement as a fixed future would reverse the workshop’s answer without establishing the premise.

Now consider a different arrangement. Qualified technical and rights determinations establish that an independently usable interface can be transferred before integration, removing this provider restriction action throughout the required window. A separately supplied full-consequence account, including the transfer’s burden and lost useful freedom, orders the resulting workshop/provider outcomes at 5 and 2. That conditional arrangement can support integration instead of continuation under the model. It remains a candidate until the provider agrees, the transfer is effective, the necessary permissions obtain and the other dependent conditions are met.

An announced future transfer or an installed client does not establish those conditions. If restriction remains possible, examine the original response again. If the provider’s preferences or information are unknown, restriction remains a relevant conditional case rather than a prediction of its behavior. Neither a favorable calculation nor the workshop’s authority establishes the provider’s agreement or a usable interface.

STR.9 compares the options with their induced responses. STR.10 compares the changed arrangement with its actual burden and authority conditions. Reuse a sufficient current response account instead of rebuilding this example for an already answered decision.

STR.4:6 - Bias-Annotation

Scenario authors often elaborate the future they want and leave the unwelcome one implausibly crude. Give a serious rival enough mechanism and detail to change the choice. Ask whose loss or access constraint disappears from an optimistic account.

Corporate competition is only one source of strategic uncertainty. Public-service and personal uses may be driven by obligations, care, access or available time. Translate the causal account and decision criterion; do not import a growth objective through the scenario format.

STR.4:7 - Conformance Checklist

• The set serves one identified strategic choice and horizon.
• Each retained condition has an intelligible mechanism and explicit consequential assumptions.
• Observations, forecasts, scenarios and targets remain distinguishable.
• Differences change strategic action, feasibility, consequences or useful observation.
• Shared failures and incompatible assumptions have been examined where they can change the choice.
• Exclusions and the basis for stopping are stated without claiming exhaustive foresight.
• Any probability, causal or permission claim has its own applicable support.

STR.4:8 - Common Anti-Patterns and How to Avoid Them

Four quadrants by habit. Two axes produce tidy but irrelevant stories. Choose conditions from the strategic question and keep a matrix only when its combinations are useful.

Three labels, one consequence. Every future recommends the same proposal for the same reasons. Determine whether that is a supported robustness result or merely redundant construction; introduce a serious discriminating condition if one is missing.

The desired future presented as expected. Customer acceptance follows because the strategy requires it. State the adoption premise and compare the condition in which it fails.

Late signals. An annual indicator cannot inform a commitment that becomes irreversible next month. Return the timing mismatch before relying on adaptation.

STR.4:9 - Consequences

The recipient can compare directions under conditions that expose different dependencies and failure regions. Assumptions become contestable, and the set can reveal an option family the original forecast hid.

Scenario construction still costs time and can make a speculative narrative memorable. Keep the qualified causal account and the decision consequence visible, so vividness does not replace evidence. Reuse or retire a scenario when the receiving question no longer needs it.

STR.4:10 - Rationale

The strategic contribution of scenarios is conditional comparison. Their value comes from making a consequential difference legible, not from forecasting success or maximizing the number of imagined worlds.

Causal explanation and decision discrimination do different work. The former makes the proposed situation intelligible; the latter justifies spending strategic attention on it. Both are needed when the scenario is retained for this use.

STR.4:11 - SoTA-Echoing

Which scenario approach is worth using for this decision? Cordova-Pozo and Rouwette’s 2023 review of reviews distinguishes scenario approaches and their construction and effectiveness questions. Adapt its method-selection lesson in sections 4.2–4.4: choose coherent, decision-discriminating conditions instead of requiring one two-axis technique. The serious lower-effort alternative is a justified single forecast or direct conditional comparison. Use it when additional scenarios would not change the decision. The review supports a reasoned method choice, not a guarantee that this scenario exercise will improve performance.

For choice when model or future agreement is unavailable, adapt the robust-decision line discussed by Lempert and colleagues in 2024: preserve consequential conditional information and examine failures across futures. This changes section 4.4’s shared-failure question and section 4.5’s qualified return. Its climate-assessment setting supplies methodological support with a transfer limit; it does not establish probabilities for SensorCo.

Could another participant’s response defeat our move? Use backward induction only where its finite, complete/perfect-information model and rationality assumptions fit the question. Muhamet Yildiz’s 2012 notes, reused in MIT’s Fall 2025 course, explain the later-to-earlier comparison and commitment contrasts in chapter 9, sections 9.1–9.5, with subgame and information-structure limits in chapter 11, section 11.1. This established Method informs section 4.3 and the provider case; current teaching reuse is not empirical evidence that real participants obey the model. A sufficient existing response account is the lower-effort alternative. Different information or behavioral premises require an appropriate qualified result.

Reopen this choice of method when a new mechanism defeats coherence, a new condition changes action, a sufficient forecast becomes available or maintaining the set costs more than its attainable decision contribution.

STR.4:12 - Relations

STR.3 supplies decision-relevant formulations when framing is unresolved. STR.4’s conditions can change STR.5’s directions, STR.6’s option families and STR.9’s robustness comparison. None requires completing these patterns in number order.

STR.8 and STR.14 address capability and joint-work conditions exposed by a scenario. STR.12 uses timely observations to support later reconsideration. FPF C.17 supports a claimed difference relative to a stated basis; PSD.5 connects complementary or conflicting model contributions when that account is needed, and PSD.10 supports an uncertainty representation when the scenario’s qualification needs it. A direct game-theoretic or behavioral determination remains with its appropriate professional Method.

STR.4:End