Two senior stakeholders deadlock: the VP of Sales insists the new CRM must 'auto-assign every inbound lead to the nearest rep instantly,' while the VP of Operations insists 'no lead moves without a manager approving the routing.' Both cite their targets as non-negotiable. What is the most effective first move to break the conflict?
- A. Escalate to the mutual executive sponsor and ask them to pick one requirement so the sprint is not blocked
- B. Separate the stated solutions from the underlying needs (speed-to-contact vs. routing accountability) and facilitate a session around those interests ✓
- C. Implement both as a configurable toggle so each VP gets their preferred behavior in their own region
- D. Build the auto-assignment first since it is quantitatively tied to a revenue target, and defer approval to a later phase
Correct answer: B. Conflicting stated solutions are usually reconcilable once you surface the underlying needs, because both interests (fast contact and accountability) can often be satisfied together rather than by choosing a winner.
A developer rejects your acceptance criterion 'The search results should load quickly and show relevant matches,' calling it untestable. Which rewrite best resolves the objection?
- A. 'Search must return relevant, high-quality results within an acceptable time under normal load'
- B. 'Given a catalog of 100k SKUs, when a user searches a valid keyword, then the first result page returns within 2 seconds at the 95th percentile and the top 10 include all exact-title matches' ✓
- C. 'Search performance and relevance must meet or exceed the current legacy system as judged by the product owner'
- D. 'The system shall provide fast and relevant search as defined in the non-functional requirements document'
Correct answer: B. A verifiable acceptance criterion binds a concrete precondition, action, and a measurable, observable threshold, whereas the other options retain subjective words like 'quickly,' 'acceptable,' or defer to a person's judgment.
You are building a requirements traceability matrix and the sponsor asks what 'bidirectional' traceability actually buys the project. Which statement most accurately captures its distinct value?
- A. It lets you trace forward from a business need to design and tests, and backward from any test or code artifact to the originating need, so orphaned or gold-plated work is exposed ✓
- B. It ensures every requirement is written in Given-When-Then format so both business and QA can read it
- C. It guarantees that changes to a requirement automatically update the linked test cases in the test management tool
- D. It provides a two-column mapping of functional to non-functional requirements to prevent scope gaps
Correct answer: A. Bidirectional traceability's unique benefit is detecting both unimplemented needs (forward gaps) and unjustified artifacts with no originating need (backward gaps, i.e., gold-plating).
A client keeps introducing new 'urgent, must-have-now' changes mid-sprint, each threatening the committed scope. Which response best protects delivery while preserving the relationship?
- A. Absorb the changes into the current sprint since the client is the paying customer and urgency is their call
- B. Accept each change but silently extend the sprint end date to accommodate the added work
- C. Run impact analysis on each request, and if truly urgent, swap it in against an equivalently-sized committed item the client agrees to drop, keeping sprint capacity fixed ✓
- D. Freeze all changes until the sprint ends and route every request to a change control board regardless of urgency
Correct answer: C. Protecting a fixed sprint capacity by trading in a new item only in exchange for de-scoping an equal item forces the client to own the prioritization trade-off without derailing the timeline.
An executive says 'implement an AI chatbot to cut support costs.' Before writing requirements, which framing is the most defensible starting point for a BA?
- A. Define the target outcome and metric (e.g., deflect X% of tier-1 tickets at maintained CSAT), then scope intents, escalation paths, and NFRs around that ✓
- B. Draft functional requirements for natural-language understanding, multi-turn dialog, and 20 supported languages to cover future needs
- C. Benchmark three vendor chatbots and recommend the one with the highest published accuracy score
- D. Specify a precision and recall target of 95% for intent classification as the primary success criterion
Correct answer: A. A vague executive ask must first be anchored to a measurable business outcome and scope boundary; model-level metrics like precision/recall are downstream and meaningless until the outcome and intents are defined.
A feature shipped exactly to spec, but the target metric (checkout conversion) did not move. Which investigative approach is most sound before concluding the feature failed?
- A. Roll the feature back immediately since a metric that did not move proves the hypothesis was wrong
- B. Confirm the feature was actually reaching and being used by the intended segment, check instrumentation validity, then decompose the funnel to locate where the drop-off persists ✓
- C. Ship a second, larger feature targeting the same metric to overpower the lack of movement
- D. Compare the aggregate conversion rate before and after launch and attribute any difference to the feature
Correct answer: B. A flat metric can stem from exposure, instrumentation, or funnel-stage issues rather than the feature itself, so you must verify reach and measurement and localize the drop-off before judging the feature; a naive before/after ignores confounders.
You inherit a stalled project: scope has ballooned, the original baseline is stale, and stakeholders distrust the team. What is the correct sequencing of your first recovery actions?
- A. Re-baseline scope against current reality, re-establish a single prioritized backlog with stakeholder agreement, then commit to a small near-term deliverable to rebuild trust ✓
- B. Immediately commit to the original deadline to signal confidence, then work backward to fit scope into it
- C. Add contractors to increase velocity so the accumulated backlog can be cleared before renegotiating scope
- D. Escalate to leadership that the project is unrecoverable and recommend a full restart with fresh requirements
Correct answer: A. Recovery requires establishing a realistic re-baselined scope and a shared prioritized backlog first, then demonstrating credibility with an early tangible win—adding people to a bloated backlog (Brooks's Law) or clinging to a dead deadline deepens the failure.
During elicitation, a critical SME is deliberately non-cooperative and withholds process knowledge, and you have no authority over them. Which approach is most likely to unlock the information?
- A. Document the SME as a project risk and proceed to write requirements from your own assumptions
- B. Report the SME's non-cooperation to their manager to compel participation
- C. Understand what is driving the resistance (fear of automation, workload, past burns), reframe their involvement around their interests, and use lower-friction techniques like observation or document analysis to reduce the ask ✓
- D. Bypass the SME entirely by interviewing their junior team members who are more willing to talk
Correct answer: C. Influence without authority works by diagnosing and addressing the source of resistance and lowering the participation cost, whereas escalation or bypassing typically hardens resistance or yields lower-fidelity requirements.
You must choose between a UML activity diagram and a BPMN process model to hand off to both business stakeholders and developers. In which situation is BPMN the clearly better choice?
- A. You need to depict the internal control flow and decision logic of a single system operation for developers
- B. You are modeling a cross-organizational process spanning multiple participants with message exchanges, lanes, and business-level events for stakeholder validation ✓
- C. You want to show the sequence of steps inside one use case's main flow at a fine-grained level
- D. You need a lightweight sketch of algorithmic branching within a component
Correct answer: B. BPMN's pools, lanes, and message flows are purpose-built for cross-participant, collaboration-heavy business processes, while activity diagrams are better suited to modeling control flow within a system or use case.
A product team wants to A/B test a redesigned onboarding flow. The PM proposes ending the test 'as soon as the new variant is winning.' Why is this the wrong call?
- A. Because the sample must always run for a full 30 days regardless of the observed effect
- B. Because peeking and stopping the moment significance appears (optional stopping) inflates the false-positive rate; the sample size and duration should be fixed in advance from the minimum detectable effect ✓
- C. Because A/B tests require an equal number of users in each arm before any result is valid
- D. Because you should instead pick the variant with the higher raw conversion count irrespective of statistical testing
Correct answer: B. Stopping a test the instant it looks significant is optional stopping, which dramatically inflates Type I error; the sample size and horizon should be pre-computed from the minimum detectable effect and power.