Prompt and context
The question asks for a real experience in which you had no reporting line or approval power but still changed a practice, priority, or decision. The point is not how many people you convinced. It is how you understood their constraints, created a shared outcome, reduced the cost of trying, and proved the change with evidence.
What the interviewer is testing
The interviewer is testing whether you can own an outcome without relying on authority. A strong answer names the stakeholders, the factual conflict, your specific actions, the choices the other person retained, and measurable results. Amazon’s official interview guidance recommends STAR stories that explain personal actions, data-backed outcomes, and lessons learned; Google re:Work identifies decision processes, role clarity, and psychological safety as parts of team effectiveness.
Clarifying questions to ask first
Clarify whether the situation involved cross-functional work, a technical proposal, or a process change; which authority you lacked and who owned the final decision; why people might resist; what success meant and over what time window; and which results can be shared in anonymized form. “Everyone liked my idea” is not a result.
A 30-second answer framework
Say: “I first aligned on the shared outcome and each group’s constraints, then used evidence to locate the disagreement and invited key skeptics to design a low-cost, reversible test. Before the test we agreed on metrics, an owner, and a review date; afterward the evidence determined whether to expand, change, or stop. I improved decision quality while leaving final authority with the named owner.”
Step-by-step deep analysis
1. Define the shared outcome, not your solution
Translate the dispute into an outcome everyone can recognize, such as less rework, customer protection, on-time delivery, or lower on-call load. Ask engineering, operations, sales, and support for their non-negotiables before presenting your preferred approach.
2. Separate facts, assumptions, and constraints
Collect current data, user feedback, dependencies, and the time window. Label the evidence strength for each conclusion and state what remains unknown. People can challenge an assumption without first accepting your position.
3. Involve the skeptics in the design
Learn privately or in a small group what the objection would cost, then invite the person to modify the plan with you. Make ownership, support needs, and failure criteria explicit. Participation improves information quality and preserves the other person’s professional judgment.
4. Lower risk with the smallest reversible test
Choose a representative scope, short window, and named owner to test the key assumption. Define a stop condition, rollback action, and review date. Without formal authority you can coordinate the test, but you cannot bypass approval or promise someone else’s resources.
5. Record the decision and dissent
Write the goal, options, evidence, unresolved risks, decision owner, and next check in a short record. The record keeps execution consistent and gives new evidence a clear route back into the decision; it is not a scorecard for who was right.
6. Close the loop with results and learning
Cover adoption, user or business impact, delivery quality, and collaboration cost. Explain who changed which behavior, how the metrics moved, which assumption failed, and what you would do earlier or design differently next time.
High-quality sample answer
“I owned a cross-team settlement change but had no management authority over payments or support. The proposed flow moved manual checks until after shipment; payments feared compliance risk and support feared more contacts. My task was to protect the date and refund accuracy, not to make everyone accept my design. I grouped six weeks of refund anomalies and support tickets and found that 70% came from two edge-order types. I invited both owners to design a three-day, 10% traffic test: edge orders stayed manual while other orders used the new flow. The payments owner had stop authority and support logged contacts. We preregistered a refund-accuracy floor of 99.5%, a support-contact increase limit of 10%, and daily reviews. Accuracy stayed at 99.8% and contacts rose 4%, so the team expanded to 50%. I recorded the untested international-order risk and scheduled the next check. The influence came from defining risk together, giving skeptics stop authority, and using results to move the decision; the payments owner still held final authority.”
Common mistakes and improvements
- Making influence about charisma: Describe the other person’s constraint, your evidence, and observable actions.
- Saying only “everyone agreed”: Name the behavior that changed, the owner, and the metric that verified it.
- Bypassing the formal owner: State approval boundaries and stop authority; coordination is not permission to overrule.
- Running a test without an exit: Set thresholds, rollback, and a review time before starting.
Follow-up questions and responses
What if the final owner still rejects your proposal?
Confirm the decision, the retained risk, and your execution responsibility, then stop repeating the pitch. If the risk crosses an agreed threshold, use the established escalation path with new evidence; otherwise execute the decision and review it at the agreed time.
What if a skeptic will not share data?
Explain which decision the data supports, the minimum fields needed, and how sensitive information will be protected. If sharing remains impossible, narrow the test or choose a validation that does not depend on that data, and record the information gap as a risk.
How do you prove the change was more than short-term compliance?
Check multiple review points and real delivery outcomes. Look for the practice in a process, runbook, or monitor rather than a single meeting promise. If the metric falls back, return to the assumptions and constraints.
How is this different from persuading people?
Persuasion emphasizes accepting your view. Influence means helping a team make a better decision through a shared outcome, clear responsibility, and verifiable evidence. The conclusion may be someone else’s proposal; you should still show how your work reduced risk.