Value analysis
Build a transparent value hypothesis for a proposed change. The aim is to make the decision logic inspectable, not to manufacture an ROI claim or dictate a price.
When the decision concerns a customer proposal, engagement, retainer, or offer, use the proposal-pricing mode below and read `references/proposal-value-anchors.md`. Read `references/spreadsheet-model.md` only when the user explicitly requests a spreadsheet. Read `references/host-operations.md` when current public evidence must be collected or for an authorized durable artifact action.
Workflow
1. Define the decision, baseline, affected population or process, time period, scope, and evidence source. 2. Test each value path separately: direct KPI value, proxy value, replacement cost, avoidable cost, and intangible benefit. 3. For each path, state the formula, assumptions, confidence, dependencies, and evidence. Mark a path unavailable rather than replacing missing data with a confident estimate. 4. Check overlap before totaling. Time saved and an eliminated vendor invoice may describe the same benefit; count it once. 5. Separate capacity value from cash savings. Released staff time is not a payroll reduction unless the labor cost actually changes. 6. Compare alternatives on matched scope, time period, adoption, and implementation cost. Keep intangible benefits qualitative unless an evidence-supported economic link exists. 7. State a decision range or next measurement step. Economic value informs a price discussion but does not establish an acceptable fee by itself.
Proposal-pricing mode
For a proposal or commercial offer:
1. Confirm the customer or buyer, use case, proposal type, current range if supplied, and whether the user wants analysis only, a saved memo, or an optional spreadsheet. 2. Classify evidence maturity: early or under-instrumented, developing with partial metrics, or metric-mature. 3. Evaluate every value path and label it **primary**, **supporting**, **future-measurement-only**, or **rejected for now**. Weak baseline data downgrades a KPI path; it does not justify invented data. 4. Choose one primary anchor and one secondary check. Early cases normally use replacement cost or cost avoidance; developing cases use a proxy plus a replacement-cost check; metric-mature cases may use direct KPI value. 5. Match the pricing frame to the anchor: ROI, better than hiring, better than an outsourced stack, cost avoidance, or foundation plus retainer. Do not use an ROI frame when the buyer cannot measure or credibly accept it. 6. Compare the proposed range with the cost to hire, outsource, coordinate vendors and tools, or continue manual work. Use conservative public wage or vendor evidence where material, and round the recommendation to a clear commercial range rather than false precision. 7. Check the proposed range against supplied seller-side economics: cost to serve, delivery capacity, contribution or margin requirements, and any approved minimum economics. If those inputs are unavailable, return a buyer-value range and state that commercial viability is unverified; do not label it a viable price recommendation. 8. Before creating or updating a durable memo or spreadsheet, present the anchor, range, frame, evidence gaps, and risks, then obtain explicit authorization unless the user already authorized that artifact.
Value paths and reconciliation
Assess direct KPI improvement, proxy metric, replacement cost, avoidable cost, and intangible benefit separately. For each path record baseline, formula, units, period, adoption, evidence date, confidence, one-time or recurring status, maturity, and dependencies. A proxy needs a stated link to an economic outcome; an intangible benefit stays qualitative unless that link is evidenced.
Build an overlap table before totaling. A released hour, avoided hire, and cancelled outsourced invoice can describe one economic effect; include only independently incremental value. Use ranges when assumptions differ. Separate cash savings, redeployable capacity, and revenue or margin. Name objections, alternative explanations, and measurement needed to raise confidence.
Reuse supplied, source-backed benchmarks before external research. Refresh a benchmark when it is stale or material to the recommendation. State whether a benchmark supports the first decision, only a future measurement plan, or no usable conclusion. If no good benchmark exists, say so and name the baseline data to collect.
Capture source URLs during research, cite each non-trivial external claim next to the claim, reproduce all sources at the end, and verify every URL-to-claim pairing before finalizing.
Output
For a general value decision, return a baseline, value-path table, formulas and assumptions, overlap check, range, evidence gaps, and recommended next decision or measurement.
For proposal-pricing mode, also return sources read, the primary anchor and secondary check, recommended pricing range and rounded anchor, pricing frame, other path statuses, target and proxy metrics, conversation framing, risks and unknowns, and three to five discovery questions. Put the range immediately after the anchor. For recurring prices show annual value first and the monthly equivalent in parentheses. Distinguish sourced component assumptions from internal planning assumptions.
Boundaries
Do not invent benchmarks, adoption, ROI, savings, a price ceiling, or a mandatory percentage-of-value fee. This method does not approve pricing, execute purchases, update financial records, or create or update a durable artifact without explicit authorization.
Deliver a value hypothesis and evidence gaps to pricing, proposal, or investment decision work; do not represent it as a signed business case.