Goldman Sachs · Behavioral Stories
How would you handle conflict and data pressure?
TrueInterview
October 7, 2026 · 4 min read
Behavioral questions (respond to both)
1) Leading when a collaborator disagrees
Your manager has just assigned you as the team lead. A collaborator openly objects to that choice and challenges your leadership.
- What is your next step?
- How do you communicate with the collaborator, the team, and your manager?
2) Manager asks for sensitive project data
You are the only engineer on a project that includes sensitive data (such as customer PII, security credentials, regulated data, or confidential partner data). Your manager requests that you share or grant access to that sensitive data.
- How do you answer your manager?
- What actions do you take to handle the request properly while preserving trust and compliance? Overview: This question measures conflict resolution, leadership communication, ethical judgment, and secure-data governance for a software engineer handling collaborator disagreement and requests that involve sensitive information. Solution
Suggested approach (address both questions)
A. Use a structured response (STAR / CLEAR)
For both situations, interviewers look for:
- Calm conflict navigation (seek understanding, align on goals)
- Ownership with humility (lead without ego)
- Principled judgment (privacy, compliance, least privilege)
- Communication and escalation (document, involve correct stakeholders) A good structure:
- Clarify facts and intent
- Align on shared goals
- Explain constraints/decision criteria
- Resolve with next steps and owners
- Loop back with outcomes
1) Leading when a collaborator disagrees
1. Diagnose before defending
- Set up a 1:1 without delay.
- Use open-ended questions: “What concerns you about me leading?” “What risks do you see?”
- Figure out whether the disagreement concerns:
- Process (the way decisions get made)
- Competence (confidence in your ability)
- Ownership (they had wanted the position)
- Technical direction (actual risk to the project)
2. Reframe around shared outcomes
- Recognize their feelings without giving up your authority: “I hear you. Let’s focus on what will make the project succeed.”
- State plainly that the aim is working together well, not “winning.”
3. Define operating model (prevents recurring conflict)
Lay out expectations clearly:
- Decision-making: for instance, “disagree and commit” once discussion ends.
- Roles: who is responsible for design, implementation, and reviews.
- Communication cadence: a weekly sync, design review, and issue tracking.
- Conflict path: if it remains unresolved, timebox it and escalate with options.
4. Build credibility with actions
- Ask them to take charge of a sub-area (this grants autonomy).
- Give public credit for their contributions.
- Be open about trade-offs.
5. Escalate only when needed
Escalate when:
- They repeatedly undermine execution
- Their behavior is disrespectful or toxic
- Deadlines are in danger When you escalate:
- Provide concrete examples, the impact, and what you have already attempted.
- Suggest solutions: clarifying roles, mediation, or an updated RACI.
What a strong final answer sounds like
“I would meet 1:1 to understand their concern, align on project goals, and agree on how decisions will be made. I would give them clear ownership in an area and make collaboration expectations explicit. If undermining continues after coaching and clear expectations, I would involve my manager with concrete examples and a proposed resolution.”
2) Manager requests sensitive data
1. Find out what they need and why
Questions to ask:
- “What decision do you need this data to support?”
- “Do you need the raw data, or would aggregated or anonymized outputs be enough?”
- “Is there already an approved access route?”
2. Follow security principles (these are non-negotiable)
Principles to cite:
- Least privilege / need-to-know
- Data minimization (provide only the least amount needed)
- Approved access controls (audit logs, role-based access)
- Compliance requirements (PII/PCI/HIPAA/GDPR or internal policies)
3. Offer compliant alternatives (demonstrate that you are helping)
Rather than giving raw sensitive data:
- Give summary metrics, dashboards, or redacted samples
- Use masked or anonymized datasets
- Grant access through an approved system (a data warehouse with RBAC)
- Set up time-bound access with approvals and logging
4. If the request is not appropriate, hold your ground professionally
Ways to phrase it:
- “I want to help, but I cannot share raw sensitive data outside the approved access controls. Let’s use the correct process, or I can provide anonymized outputs that meet your need.”
5. Escalate through the proper channels
Escalate when:
- You are pressured to bypass policy
- The manager insists on direct access without approval Escalation route:
- The security or privacy team, data governance, or your skip-level (depending on company policy)
- Keep it factual and documented (what was requested and the safe alternatives you proposed)
6. Document decisions
- Record what data was requested, what was provided, which approvals were obtained, and where it is stored.
What a strong final answer sounds like
“I would clarify the business need and propose a safe alternative like aggregated or anonymized outputs. If access is truly necessary, I would route it through approved RBAC/audited mechanisms and required approvals. If I am asked to bypass policy, I would refuse politely and escalate to security/data governance, documenting the request and the compliant options offered.”