Five Case Model
Five perspectives on one proposal: strategic, economic, commercial, financial, and management. Develop them together with detail suited to the decision.
-
-
Sunk Cost
A past license payment does not decide whether to use the product. Compare future costs and benefits, including what owned resources could do elsewhere.
-
Build or Buy
A purchased product can include a custom adapter. Compare complete approaches, verify mandatory requirements, and agree who can maintain and move the result.
-
True-up
A current user count may miss a contractual payment obligation. Check what the agreement counts, when additions are reported, and when reductions take effect.
-
Technical Lock-in
Pay-as-you-go means you can stop at any time in theory. In practice the architecture, the missing equivalents, and the team’s skills decide whether you can.
-
Unit Economics
A lower unit cost can coexist with a higher bill. Define the cost, output, period, and service conditions before interpreting the change.
-
Switching Value
Financial break-even and a change in the preferred option are different boundaries. A variable may have several boundaries, or none in its feasible range.
-
Optimism Bias
Correct systematic forecast error using comparable evidence. Show how mitigation, its cost, and remaining uncertainty change the estimate and reserve.
-
Planning Interval
Joint planning aligns known dependencies and exposes open questions. Integration, feedback, and necessary decisions continue within the interval.
-
Three Horizons
Manage the established business, emerging opportunities, and future options concurrently. These growth horizons are not fixed technology budget categories.
-
Line of Sight
A documented link to a goal makes an investment claim inspectable. Its contribution still needs evidence, assumptions, and a suitable test.
-
Digital Take-up
Digital channel completion and end-to-end use of a new internal system are different measures. Define their boundaries before interpreting adoption.
-
Benefits Realization
Faster work can release capacity without reducing spending. Benefit reviews distinguish forecasts, observed outcomes, attributed effects, and cash savings.
-
Parallel Run
In a fictional telecom, a parallel run compared month-end figures from both systems and exposed a rounding rule the legacy report had applied for nine years.
-
Migration Fallback
In a fictional telecom, a standalone backup lacked six days of changes. Returning required a tested procedure to recover state, transfer write authority, and reconnect clients.
-
Migration Wave
In a fictional hospital, the pharmacy system had few dependencies in the list. The nightly batch that read it was in neither the list nor the wave. The morning shift found out first.
-
First Mover
Early migrations yield working systems and lessons within tested conditions. Select candidates by learning value, failure impact, and what can transfer.
-
Lift and Shift
Rehosting preserves application design, but infrastructure, responsibilities, and costs can change. Verify the move and reassess application problems.
-
Continuous Improvement Planning
Regular assessment and funded remedies help limit legacy accumulation. Support dates, uncertainty, dependencies, and residual risk all belong in the plan.
-
Technology Roadmap
External deadlines, internal commitments, targets, and forecasts mean different things. A roadmap makes those meanings and the choices behind them visible.
-
Strategy Review
A useful review can change a delivery plan or challenge the diagnosis. Its scope follows the evidence, with clear ownership and authority to act.
-
Operating Model
Shared processes and data guide organizational choices, but do not automatically determine authority, funding, or a single implementation.
-
Foundation for Execution
Shared capability can be funded centrally, jointly, or by its first consumer. Reuse needs clear responsibilities and resources for continued operation.
-
Riskiest Assumption
Choose which assumptions to test by comparing uncertainty, consequences, dependencies, and test cost. Model performance can be one of them.
-
Pilot Exit Criteria
A pilot needs a defined next decision. Missing evidence does not automatically justify an extension, and a successful test does not authorize every rollout.
-
Strategy Kernel
A strategy kernel connects a diagnosis to a chosen approach and coordinated actions. Its quality depends on evidence and feasible execution.
-
Business Capability Map
Several systems can support one capability without duplicating it. Map their responsibilities and compare gaps, overlap, and improvement options.
-
Now-Next-Later Roadmap
Now, Next and Later express different levels of confidence. Discovery continues in each column, while meaningful dates remain explicit.
-
Cost of Delay
Waiting time is evidence to investigate, not necessarily lost labor. Cost of delay makes the consequences of postponement explicit, with units and assumptions.
-
User Journey Map
A journey map follows one perspective toward a goal. Distinguish observed experience, assumptions, and future designs when choosing what to improve.
-
Service Blueprint
Connect the user journey to people, systems, and processes. Use operational evidence to test which dependencies and defects contribute to a delay.
-
Product Discovery
Discovery investigates user needs, candidate solutions, and remaining risks so a team can justify its next investment and learn during delivery.
-
Exit Cost
Estimate planned and urgent exits separately: data transfer, logic changes, rebuilt operations, contracts, and transition work all affect the cost.
-
Chargeback
Chargeback moves technology costs onto the budgets of the teams that use them. It sharpens accountability — and punishes any flaw in how costs are attributed.
-
Showback
Showback shows each team what its share of technology costs is, without moving money between budgets — with workload control and value review needed to decide what to change.
-
Cost Attribution
Cost attribution uses recorded workload identity to trace consumption reliably. Shared costs can also be allocated by a declared policy, which is distinct from evidence of actual use.
-
Full-Time Equivalent (FTE)
FTE standardizes staffing hours. Available capacity, headcount, shift coverage, and the chosen full-time denominator remain distinct quantities.
-
OpEx
OpEx concerns expense recognition. Payment timing, asset depreciation, and minimum capacity commitments must be assessed separately.
-
System Retirement
Retire a system completely: confirm no dependencies, keep retained records accessible under their obligations, then remove credentials, jobs, and costs.
-
Cutover
Plan a cutover with a rehearsal, a timed runbook naming owners and evidence, go/no-go criteria agreed in advance, a stop rule, and a decision owner.
-
Coexistence
Run old and new systems together with an authority rule, a shared comparison point, no duplicated side effects, and a plan for changes after the copy.
-
System Dependency
Find system dependencies beyond application code: jobs, files, procedures, reports, service accounts, manual steps, and rare month-end work.
-
Legacy System
Understand a legacy system through support, changeability, dependencies, and operating risk rather than age, using the GOV.UK criteria.
-
Modernization and Migration
Distinguish modernization, a change made to meet needs better, from migration, the act of moving data, workloads, or users, and why one is not the other.
-
Double Counting
Recognize double counting when one benefit or cost appears under several criteria, check what each criterion includes, and keep weights honest.
-
Sensitivity Analysis
Use sensitivity analysis to test whether plausible changes in weights or inputs change a ranking, locate the reversal point, and name deciding evidence.
-
Weighted Scoring
Learn how weighted scoring adds anchored scores, why a weight is the value of a swing rather than abstract importance, and what the model assumes.
-
Scoring Anchors
Define scoring anchors that map raw measurements to 0 and 100, handle values outside the range, and see how fixed anchors differ from relative scaling.
-
Mandatory Gate and Preference
Separate mandatory gates with pass, fail, and unverified states from preferences that are scored, and see why a high total cannot offset a failed gate.
-
Evaluation Scorecard
Understand an evaluation scorecard as gates, anchored scores, weights, evidence, and a conditional recommendation, and what its total cannot claim.