Relevant costs and comparison horizons

1. Build a common basis before adding values

A migration request can arrive with three apparently comparable proposals: retain middleware, upgrade its version or move the application to another service. Before calculating, state the outcome the business needs and the operating conditions.

Options must cover the same relevant scope, including support, recovery, interfaces and retirement of old components where applicable. An installation price excluding operation is not directly comparable with one already including three support years.

Record currency, horizon, starting point and inclusions. In this lesson, amounts are fictional, services are equivalent under the stated conditions, and calculations exclude discounting, taxes and inflation unless explicitly stated. These simplifications support studying the decision; they do not replace a real project’s financial conventions and requirements.

2. Separate history from costs the decision can still change

The initial study cost twelve thousand euros, already paid and unrecoverable whichever option is chosen now. That amount belongs in project history but does not differentiate future alternatives. Charging it only to the option abandoning the original proposal artificially favors continuation.

The useful question is which costs and outcomes the decision can still change. A future cancellation fee can matter; a refundable deposit can also have different consequences across options. Do not classify everything already paid as irrelevant without checking recoverability and alternative use.

Keep historical expenditure visible in project monitoring, separately from the incremental comparison. The sponsor can then understand both how much has been spent and which alternative better uses remaining resources. Classification erases neither accountability nor accounting records.

3. Identify alternative use of an existing resource

The application can use equipment the organization already owns. Its historical purchase cost was ten thousand euros and cannot be reversed. An authorized, feasible sale for four thousand is now available only if the application does not use the equipment.

Using it means foregoing that four-thousand alternative even without another acquisition payment. Distinguish opportunity cost from cash outlay: the former helps compare resource uses; the latter supports cash and budget planning. Do not automatically add the historical ten thousand and foregone four-thousand sale as two new payments.

Also establish that the alternative is achievable, including timing, authorization and necessary transaction costs. An imagined valuation without a buyer or feasible alternative use does not become recoverable value merely by appearing in a spreadsheet.

To compare A, which uses the equipment and requires eight thousand in future spending, with B, which requires fourteen thousand and permits the four-thousand net sale, choose a consistent basis. Future cash gives A=8 and B=14−4=10. An opportunity-cost basis charges A=8+4=12 while B=14.

A costs two less in both presentations. Adding four to A while also subtracting the same four from B counts the resource difference twice, artificially reversing the choice.

4. Include transition, maintenance and retirement

Over a common twelve-month horizon, retaining the service costs four thousand euros monthly and requires a six-thousand mandatory upgrade, totaling fifty-four thousand. Migration costs eighteen thousand for preparation, two thousand monthly for all twelve months, three thousand additional coexistence cost and six thousand for retirement at the end, totaling fifty-one thousand.

Additional amounts are not included in monthly fees, and this exercise has no other flows. The projected difference favors migration by three thousand without turning the difference into approval. Omitting retirement yields forty-five thousand and exaggerates the advantage; omitting the baseline upgrade yields forty-eight thousand and changes the conclusion.

Ask each owner to confirm estimate scope: FinOps may know consumption, APS support and another team decommissioning work. No isolated total guarantees inclusion of all these elements.

5. Find when the comparison changes

Another comparison uses thousands of euros and m complete service months. A costs 5 + 3m; B costs 17 + m. Equality occurs when 5 + 3m = 17 + m, giving m=6.

At three months, A costs fourteen and B twenty; at twelve, A costs forty-one and B twenty-nine. The cheaper option depends on the horizon even when both provide the same service. If allowable duration ranges from four to eight months, do not claim B is always cheaper.

Show both sides of the switching point and identify who can clarify likely duration or the decision policy under uncertainty. Do not invent a uniform probability distribution for the interval. Equal cost at six months also establishes no equality in risks or characteristics omitted from this model.

6. Separate total advantage, funding and benefit timing

An option can cost less overall while requiring more money before operation. If migration needs twenty-one thousand initially and the authorized limit is twenty thousand, its projected three-thousand advantage over the horizon does not remove the initial constraint.

Present the funding gap and alternatives for the decision owner to assess. Benefit timing also needs to be explicit. If actual expenditure reduction begins at the start of month four and amounts to two thousand monthly, months four through twelve comprise nine months and eighteen thousand, not twelve months and twenty-four thousand.

Check whether the saving is already reflected in a lower cost line to avoid counting it again as an additional benefit. A schedule change is not neutral merely because the monthly amount stays unchanged.

7. Explore uncertainty that can change the recommendation

In the twelve-month example, migration costs forty-five thousand before retirement. If retirement can cost between four and ten thousand, total cost ranges from forty-nine to fifty-five thousand. Retaining the service costs fifty-four thousand.

Migration is cheaper when retirement costs less than nine thousand, ties at nine thousand and costs more above it. This interval establishes no guaranteed advantage but identifies relevant information to collect. If an inspection establishes a maximum of eight thousand under applicable conditions, total cost is bounded at fifty-three thousand and the cost ranking becomes stable within that scope.

A cost margin proves neither risk acceptance nor satisfaction of other conditions. Nor does it justify unlimited investigation: information’s usefulness should be compared with the cost and time needed to obtain it.

8. Exercise: prepare a traceable recommendation

Prepare a summary before reading the solution. The baseline costs fifty-four thousand over the year. Migration costs forty-five thousand before retirement, which remains between four and ten thousand.

It requires twenty-one thousand immediately against an authorized limit of twenty thousand. The twelve-thousand study has already been paid and is unrecoverable under every option. Solution: the future comparison excludes the common study while retaining it in history.

Migration ranges from forty-nine to fifty-five thousand, so it can cost less or more than the baseline. The retirement switching value is nine thousand. A separate initial funding gap of one thousand exists.

The recommendation should expose the estimate needing greater precision and the required funding decision. Presenting the midpoint of fifty-two thousand as an observed cost or authorization to start is insufficient.

Common pitfalls

Different horizons; historical expenditure charged to only one option; owned equipment treated as having no alternative use; omitted retirement; double-counted savings; intervals treated as probabilities; total advantage confused with available funding.