Analyze interacting change impacts

1. Describe the outcome and contract boundaries

In a fictional funds service, the last eligible request arrives at 17:00. The worker starts final processing only after cutoff and takes twenty elapsed minutes. A CFT file departs at 17:30, only after all processing is complete.

Transfer takes ten minutes and the recipient needs five to validate the file. The business condition requires validation by 18:00 inclusive. All times use the same clock and day; durations do not overlap.

The reference situation finishes at 17:45. This contract supports change analysis because it identifies events, sequence, population and boundary. “File sent” and “file validated by the recipient” are different outcomes.

Renaming a dashboard state does not change the commitment. The BA should confirm with owners which times can be controlled, which data enter the final batch and which evidence establishes each event. Figures are invented for learning; they are not any bank’s timings, SLA or procedures.

Project person-days, considered in another section, are also different from these operating minutes.

2. Analyze the combination as well as individual proposals

Proposal A extends cutoff to 17:15 and moves departure to 17:40. With twenty-minute worker processing, the file is ready at 17:35 and validation finishes at 17:55. Proposal B changes the transfer route, increasing its duration from ten to twenty-five minutes.

Applied alone to the reference, it retains departure at 17:30 and finishes at 18:00, still within the inclusive limit. Combining A and B, however, gives departure at 17:40, twenty-five-minute transfer and five-minute validation: completion at 18:10.

Each proposal’s isolated success does not prove joint compatibility. The BA should show which contract terms each team changes and how the changes interact even when requests have different sponsors or identifiers. The calculated miss follows the supplied assumptions; no actual failure was observed.

Analysis also does not automatically decide which proposal to reject. First expose incompatibility and identify options preserving priority needs, with costs, risks and evidence still required.

3. Calculate the departure window satisfying both sides

A file cannot depart before completion and must depart early enough to satisfy the recipient commitment. In the combined case, earliest departure is 17:15 plus twenty minutes, or 17:35. Latest departure is 18:00 minus twenty-five transfer minutes and five validation minutes, or 17:30.

The permitted interval would require departure both at or after 17:35 and at or before 17:30: it is empty. Choosing an intermediate time, averaging boundaries or transferring an incomplete file does not solve it. To retain a 17:15 cutoff and 18:00 deadline, the worker would need to finish within fifteen minutes, permitting 17:30 departure.

This is a derived condition for option assessment, not an implemented improvement. If worker processing could take ten minutes, departure could fall between 17:25 and 17:30. That window width is scheduling flexibility under assumptions, not a guarantee against every failure.

Changed sequence or partial processing would need a new business contract and its own analysis.

4. Sum work units and identify new work

Effort analysis has a different boundary. Proposal A has four person-days of its own work and includes shared review T of three, totaling seven. B has five of its own and includes the same three-day T, totaling eight.

This exercise’s agreement permits one T execution for combined scope, but joint integration additionally requires J at two person-days. The total is 4 + 5 + 3 + 2 = 14, neither fifteen from adding requests nor twelve from merely removing duplication.

T’s identity must include the context making reuse valid. If the two T references represent environments or versions requiring distinct executions, the agreement no longer justifies deduplication. J also cannot be omitted because it did not exist in isolated requests.

The BA should help connect new tasks to identified effects and confirm estimates with those performing the work. This sum measures forecast effort; capacity, skills, dependencies and waiting remain necessary to estimate the project calendar.

5. Handle uncertainty without changing the decision criterion

For J, now suppose owners justify only a range from one to four person-days. Remaining combined effort is still twelve. Forecast total lies between thirteen and sixteen.

A fifteen-person-day limit is neither guaranteed nor inevitably exceeded. Demonstrating that limit would require justifying J at no more than three, or deciding a scope or capacity change through the applicable route. Operationally, a measured twelve-minute worker mean also does not prove the at-most-fifteen-minute condition for the relevant batch.

Do not confuse mean, observed maximum, assumed bound and future guarantee. If the exercise explicitly supplies twenty as the admissible upper bound, the combined window remains infeasible at that extreme. If new evidence justifies at most fifteen within the required context and population, analysis may demonstrate temporal feasibility under that bound.

Retain the assumption version and supporting evidence, separating them from decision status. Supplied deterministic ranges are neither statistical confidence intervals nor probabilities of compliance.

6. Match the decision to combined scope

In this case, fictional governance binds each approval to the assessed package and baseline. A and B were analyzed separately against B1. One decision on A alone and another on B alone do not themselves represent a decision on the incompatible combination identified by analysis.

The BA prepares differences, consequences, new work and conditions for applicable authority while retaining earlier decisions. Renaming the baseline or combining approval records should not simulate joint assessment that never occurred. If the decision maker authorizes building a faster-worker option, authorization still confirms neither fifteen-minute performance nor production acceptance.

Earlier tests can contribute where they remain applicable to requirement, version, environment and context; they do not automatically verify the new composition. Analysis should identify what can be reused, what needs adaptation and which gaps prevent conclusion.

The matching policy is supplied by the scenario rather than presented as a universal PMI or financial-institution workflow.

7. Compare options without erasing needs

Several responses to incompatibility are possible: change processing, revise departure within a feasible window, negotiate another recipient commitment or reformulate the cutoff need. Each changes assumptions and may affect value, risk and work.

If 17:15 and 18:00 are current mandatory conditions, an option working only by moving cutoff to 17:10 or deadline to 18:10 does not yet satisfy them. It may be assessed as a change but needs the appropriate decision. Transferring data before batch completion may also change completeness, identity and recovery; it should not be treated as simple scheduling optimization without business validation.

For example, files with 110 and 60 operations sharing twenty identifiers contain 150 distinct operations if each list has unique identities and cross-list deduplication is confirmed. If every identifier belongs to the eligible population of 150, that coverage is established.

That does not prove timing: if arrivals are 17:50 and 18:00 and three-minute merging starts only after both, followed by five-minute validation, completion is 18:08. Population is correct, but the timing condition fails. For the committee, begin with the interaction: “Each change fits the reference separately; together they miss the recipient deadline by ten minutes.”

Then present the option, scope, effort, what has been demonstrated and who needs to decide. APS, transfer owners, development, QA and Business should receive relevant consequences including procedures, monitoring and support for future operation.

8. Exercise: recommend a coordinated change

Before reading the solution, write a package with three outcomes: timing comparison, effort and decision state. Use A and B from earlier sections, with mandatory 17:15 cutoff, mandatory validation by 18:00, twenty-five-minute transfer and five-minute validation.

The team proposes reducing worker processing to fifteen minutes and moving departure from 17:40 to 17:30; this option has not been tested. J has a confirmed two-person-day estimate and T is the agreed single review. Solution: the current combination finishes at 18:10.

The modified option would have file readiness and departure at 17:30, finishing at 18:00 under the at-most-fifteen-minute worker assumption. Effort for the supplied work scope is fourteen person-days, but any extra work needed to achieve optimization must be identified and added; this figure is not a complete budget for that unknown option.

The recommendation should expose the technical assumption, estimate and evidence gaps and route the joint decision. Do not declare acceptance merely because timing arithmetic is feasible. In the workshop, change one condition at a time, retain the reference and explain when a new combination makes earlier decisions or results insufficient.

Common pitfalls

Combine isolated approvals; advance an incomplete file; confuse mean with bound; remove duplication but forget integration work; present partial effort as a complete budget.