A digital business case needs a causal chain

A business case should explain how the funded work changes an operating mechanism and how that change could produce a benefit. Licence acquisition, configuration completion and user registration are evidence of delivery or access. They are not the benefit itself. A programme may improve a service without producing a direct financial saving, or produce a saving only if another owner changes staffing, purchasing or operating practice. The case should state those conditions rather than move directly from product features to a claimed return.

In a fictional integrated-planning service, a shared availability view helps planners identify conflicts earlier. The potential benefit is not simply that data is visible. It may be less reconstructive coordination, fewer avoidable plan changes or a better basis for a promise. Each proposition requires a changed action: planners must use the view, operating teams must maintain its meaning and the decision owner must respond to the conflict. If the old decision process remains unchanged, the new interface can be used frequently without changing the outcome.

The case should distinguish benefit categories. Reduced manual work, improved decision quality, service continuity and strategic access can all matter, but they should not be converted into money casually or counted twice. A continuity replacement may avoid an exposure that is difficult to quantify credibly. The sponsor can still assess it through an explicit obligation and options comparison. False financial precision is not necessary to make a serious investment decision.

A direct operating benefit can depend on organisational action outside the technology team's control. If better scheduling is expected to reduce purchased overtime, an operating manager must change how labour is arranged. If a service reduces duplicated reconciliation, finance must determine whether the saved effort is released, redeployed or absorbed by new work. The programme should record the accountable action and its evidence. It should not call nominal task time a cash saving without a credible path to a financial effect.

The case also needs a reference practice. The enterprise should describe what it currently does and what would happen without the proposed investment. That reference may change because an existing service becomes unsupported or because the business itself changes. A stable fictional baseline should not be treated as an actual counterfactual. In real use, the sponsor must examine the relevant continuity option and operating alternatives before assigning benefits to the chosen platform.

An assumption is a declared proposition used for reasoning. Evidence is information that supports or challenges a proposition within a scope. A commitment is an authorised action the organisation agrees to perform. These categories should remain visible in the business case. A readiness judgment is an assumption; a tested workflow is evidence; an operating manager's funded policy change is a commitment. Treating all three as equally established can make a proposal appear more mature than it is.

The Transformation map uses a declared annual benefit multiplied by confidence and readiness judgments. These fields permit transparent sensitivity analysis. They do not establish a probability distribution, observed adoption rate or measured benefit. The declared values should be labelled as assumptions and linked to the operating mechanism they represent. A model can calculate the consequences of the values exactly while the evidence supporting them remains incomplete.

The business case should record why a judgment was chosen and what would change it. A confidence judgment might reflect uncertainty in the claimed causal mechanism; readiness might reflect an assumed ability to implement and use the capability. Neither label has a calibrated universal meaning in the app. Authors should define their interpretation for the case and avoid presenting a change in a multiplier as proof that an operating prerequisite has been completed. Where a mandatory condition exists, it should remain an explicit gate outside the multiplier.

Commitments should identify the role and resource needed for benefit realisation. The fictional planning service needs a data-maintenance obligation, a conflict-resolution decision and a change in planning practice. If these commitments are absent, a high benefit assumption remains conditional. The sponsor should either secure them, narrow the claim or defer the capability. A technology supplier's delivery obligation cannot automatically substitute for the enterprise's own operating commitments.

The case should preserve uncertainty honestly through approval. Early approval can fund discovery, evidence collection or a bounded pilot rather than the full rollout. This allows a programme to proceed while recognising that its benefit proposition is not yet demonstrated. The approval record should state which questions remain unresolved and which later decision depends on their answers. Calling a pilot approved is different from calling the enterprise benefit proven.

A collection of individually attractive cases can produce an overstated portfolio if they share the same outcome. The planning service and exception service may both reduce the work of reconstructing a changed instruction. If each claims the full effect, adding their benefits duplicates the proposition. The business case should identify shared mechanisms, allocate their effect deliberately and document where benefits are contingent on a combination. The companion app does not infer overlap or synergy from capability names or dependencies.

Enabling capabilities need particular care. A common data service can have a direct operating benefit and support downstream benefits. Those categories should be distinguished so the enabler is not credited with the entire downstream effect while the consuming capabilities are also credited. A dependency arrow means the prerequisite is included in the bundle. It does not establish a new benefit accounting rule. The author remains responsible for a coherent additive slate.

The cost boundary also matters. The app's cost field is expressed in USD thousands for a declared capability investment. It has no separate fields for continuing operation, recurring licences, change, security, retirement or migration overlap. An author can include a chosen cost scope in that field, but should state it consistently. A ranking that compares one capability's implementation-only cost with another's broad lifecycle cost would be difficult to interpret even if the arithmetic is correct.

Annual adjusted benefit divided by implementation cost is a ranking ratio in the app. It is not a discounted valuation, internal rate of return or payback period. Its units reflect annual benefit relative to a cost amount. The ranking can be useful within a coherent slate, but it does not establish financing suitability or enterprise value. The business case should use a more appropriate method where time profile, recurring cost, risk or a required financial criterion is material.

Resource constraints should remain distinct. The app considers both cost and effort, with effort in person-months. It does not infer which skills are available, when the effort can be supplied or which capabilities compete for a specialised team. A selected set can satisfy the declared aggregate effort envelope while being impossible to deliver under the actual calendar and skill constraints. The operating case should retain those constraints and direct a separate delivery assessment where needed.

The fictional sponsor approves a planning pilot whose first purpose is to test whether a shared availability view changes conflict resolution. The approved operating commitment is that planners use the view for the included case type, data owners maintain the defined availability states and an operating manager resolves identified conflicts through an authorised route. The pilot does not yet claim a reduction in overtime or service failure. It collects evidence about the mechanism on which those broader effects might depend.

The fictional planning pilot completes a mechanism comparison record. In the original route, a planner reconstructs availability through calls; in the proposed route, the shared view supplies the current state but an operating manager still resolves competing demands. A routine recognised conflict can test whether reconstructive work disappears. A disputed availability state tests whether that work merely moves to the data owner. A conflict resolved by a temporary senior intervention tests whether the intended authority route is actually usable. These contrasts keep a clearer screen from being mistaken for a changed service. They define observations to obtain rather than asserting a pilot frequency or time saving.

The sponsor's fictional decision matrix records different consequences for the same attractive presentation. A usable availability view with an unresolved decision route supports continued case-design work. A functioning decision route with a burdensome shared-state upkeep directs a resource and benefit revision. A clear route and sustainable upkeep can support evaluation of wider scope, while a result obtained only through exceptional senior intervention does not establish ordinary coverage. A concurrent staffing change is entered as a competing explanation for any later operating improvement. The point of these branches is to distinguish mechanism failure from insufficient evidence and from an additional operating cost, rather than classify every disappointing result as failed technology.

The resulting fictional business-case entry is complete without a return figure. Intended contribution: a planning case whose meaning can be reconstructed and whose conflict receives a legitimate decision. Included observation: routine and disputed availability cases. Displaced work: maintaining the shared state and resolving its exceptions. Competing explanation: changed staffing or demand. Current disposition: assess the mechanism before attributing a financial effect. A later expenditure reduction would require its own verified bridge from changed work to actual resources purchased. This instrument adds a causal comparison and a cost-location test; it retains the difference between a declared benefit in the app and a measured consequence in an operating service.