Agile Coach & Scrum Master Interview Q&A¶
Last reviewed: November 2026
Questions for Scrum Masters and Agile Coaches working with engineering teams, including AI and data teams: Scrum and Kanban basics, facilitation, coaching, metrics, scaling and handling resistance. Behavioural answers follow STAR (Situation, Task, Action, Result), then what you learned. Use your own real examples.
Terms follow the 2020 Scrum Guide. See the Agile Coach & Scrum Master role guide for the interview loop, rubric and a 14-day plan.
Scrum foundations¶
What is the role of a Scrum Master?
Short answer: The Scrum Master is accountable for the Scrum Team's effectiveness: helping everyone understand Scrum theory and practice, and serving the Developers, the Product Owner and the wider organisation.
In depth:
- For the team: coach self-management, help remove impediments, make sure events happen and are useful.
- For the Product Owner: help with Product Goal definition and backlog management techniques.
- For the organisation: help it adopt Scrum and remove barriers between teams and stakeholders.
Follow-up they'll ask: How is that different from a project manager?
Explain the Scrum events and their purpose.
Short answer: The Sprint contains Sprint Planning, the Daily Scrum, the Sprint Review and the Sprint Retrospective; each is a chance to inspect and adapt.
In depth:
- Sprint: a fixed length of one month or less in which a usable Increment is created.
- Sprint Planning: why the Sprint is valuable (Sprint Goal), what can be done, and how.
- Daily Scrum: Developers inspect progress toward the Sprint Goal and adapt the plan.
- Sprint Review: inspect the outcome with stakeholders and adapt what comes next.
- Sprint Retrospective: inspect how the team worked and plan improvements.
Follow-up they'll ask: Which event do teams most often get wrong, and why?
What are the Scrum artifacts and their commitments?
Short answer: The Product Backlog with its Product Goal, the Sprint Backlog with its Sprint Goal, and the Increment with its Definition of Done.
In depth: The commitments give each artifact a focus: the Product Goal is the long-term objective, the Sprint Goal is the single objective for the Sprint, and the Definition of Done is the quality bar an Increment must meet to count.
Follow-up they'll ask: What happens to work that does not meet the Definition of Done?
What makes a good Sprint Goal?
Short answer: One clear, outcome-focused objective that gives the team a reason to work together and flexibility in exactly which items deliver it.
In depth: "Customers can reset their password without contacting support" is better than "finish tickets 12 to 19". A good goal helps the team make trade-offs during the Sprint and gives stakeholders something meaningful to review.
Follow-up they'll ask: The team keeps missing its Sprint Goal. What do you do?
How do you run a retrospective that leads to real change?
Short answer: Make it safe to be honest, focus on one or two improvements the team owns, make them concrete actions, and check them at the next retrospective.
In depth: Vary the format to avoid fatigue. Use data from the Sprint, such as flow metrics or incidents, alongside feelings. Limit actions to what the team will actually do, and put them in the Sprint Backlog if needed.
Follow-up they'll ask: Retros have become complaint sessions. What do you change?
Kanban and flow¶
When would you recommend Kanban over Scrum?
Short answer: When work arrives continuously and unpredictably, such as support, operations or platform requests, or when fixed Sprints add overhead without benefit.
In depth: Kanban focuses on visualising work, limiting work in progress and improving flow. Scrum gives a cadence for planning and review. Many teams combine them. Choose based on the work and the team's needs, not dogma.
Follow-up they'll ask: How would you introduce WIP limits to a team that resists them?
What flow metrics do you use and why?
Short answer: Work in progress, cycle time, throughput and work item age, used by the team to see and improve flow, not to judge individuals.
In depth: High WIP usually lengthens cycle time. Work item age shows items getting stuck before they finish. Throughput helps forecasting. Read them as trends and discuss them in retrospectives.
Follow-up they'll ask: How would you forecast a delivery date using these?
How do you feel about story points and velocity?
Short answer: They can help a team plan its own capacity, but they are easily misused; never compare velocity between teams or use it as a performance target.
In depth: Many teams forecast just as well by counting items and using throughput. If you use points, keep them a team tool. Watch for point inflation when velocity becomes a target.
Follow-up they'll ask: Leadership wants to compare teams' velocity. How do you respond?
Coaching and facilitation¶
What is the difference between coaching, mentoring, teaching and facilitating?
Short answer: Teaching shares knowledge, mentoring shares your experience, coaching helps people find their own answers through questions, and facilitating helps a group reach its goal in a conversation.
In depth: A good coach moves between these stances deliberately, depending on what the person or team needs. For example, teach the basics of a Retrospective to a new team, then coach them to run their own.
Follow-up they'll ask: Give an example where you switched stance mid-conversation.
How do you help a team become self-managing?
Short answer: Gradually hand over decisions with clear boundaries, step back from running events, and coach the team to solve its own impediments.
In depth: Make explicit which decisions the team owns. Let them facilitate events in rotation. When they bring a problem, ask what they could try before offering an answer. Celebrate them solving things without you.
Follow-up they'll ask: How do you know when to step in?
A team says the Daily Scrum is a waste of time. What do you do?
Short answer: Find out why, then help them make it useful: short, focused on the Sprint Goal, run by the Developers, and ending with any re-planning needed.
In depth: Common causes are status reporting to a manager, too many people, or no shared goal. Try changes such as walking the board right to left, focusing on blocked items, or timeboxing strictly. Let the team decide the format within the purpose.
Follow-up they'll ask: What if the manager insists on attending?
How do you handle a dominant person in team meetings?
Short answer: Use facilitation techniques that give everyone a voice, and if needed, talk privately with the person about the impact.
In depth: Techniques include silent writing before discussion, round-robins, and small-group work. Privately, ask them to help draw others in; often they do not notice their impact.
Follow-up they'll ask: What if that person is the most senior engineer?
Organisation and scaling¶
How do you handle resistance to agile ways of working?
Short answer: Understand the concerns first, start with small changes that solve problems people already feel, and show results rather than arguing about frameworks.
In depth: Resistance often comes from past bad experiences, fear of losing control, or extra overhead. Involve sceptics in designing the change. Measure the outcomes they care about.
Follow-up they'll ask: Tell me about a sceptic you won over.
How would you coordinate several teams working on one product?
Short answer: Start simple: one Product Backlog and Product Goal, shared Sprint cadence where useful, regular cross-team refinement and a joint review; add a scaling framework only if simpler coordination fails.
In depth: Frameworks such as Nexus, LeSS or SAFe offer structures for this, with different trade-offs. Reduce dependencies by organising teams around features or value streams rather than components where possible.
Follow-up they'll ask: What are the risks of adopting a heavy scaling framework?
How do you measure the success of an agile transformation?
Short answer: By outcomes: faster delivery of valuable work, better quality, happier customers and teams, not by the number of teams "doing Scrum".
In depth: Combine delivery measures, such as cycle time and DORA metrics, with customer outcomes and team health surveys. Set a baseline at the start and look at trends.
Follow-up they'll ask: What is a vanity metric in this context?
Working with AI and data teams¶
How do you apply Scrum to an AI or ML team where results are uncertain?
Short answer: Frame Sprint Goals as learning outcomes when needed, time-box experiments, and include evaluation in the Definition of Done.
In depth: A Sprint Goal like "know whether the new retrieval approach beats the current one on our evaluation set" can work, but Scrum still expects a usable Increment every Sprint, so pair learning goals with something usable, or consider Kanban for mostly exploratory work. Make experiment results visible in the Sprint Review. Keep a balance between exploration and delivering usable Increments.
Follow-up they'll ask: What does Done mean for a model change?
What would you add to a Definition of Done for a data or AI team?
Short answer: Checks that make the work trustworthy in production, such as data tests, evaluation results against an agreed bar, monitoring, documentation and access reviewed.
In depth: For example: pipeline has data quality tests; model or prompt change passes the evaluation set; dashboards or alerts exist; privacy and access reviewed for new data. Let the team own and evolve the list.
Follow-up they'll ask: The team says the Definition of Done slows them down. What do you say?
How do you help a data team that is overwhelmed with ad hoc requests?
Short answer: Make all incoming work visible, agree an intake and prioritisation process with stakeholders, and protect focus with WIP limits or a rotating support role.
In depth: A Kanban board for requests plus a clear owner for prioritisation helps. Track how much time goes to ad hoc work and discuss it with leadership. Look for repeated requests that self-service could replace.
Follow-up they'll ask: How do you say no to a powerful stakeholder?
About you¶
Tell me about your biggest success as a Scrum Master or coach.
Short answer: Pick a story with a clear before and after, your specific actions, and evidence of the result.
In depth: Keep the focus on what the team achieved and how you helped, with concrete signals such as faster delivery, fewer incidents or better team survey results. Be honest about what was hard.
Follow-up they'll ask: What would you do differently?
How do you keep growing as a coach?
Short answer: Show deliberate practice: feedback from teams, peer coaching, reading or training, and reflecting on what did not work.
In depth: Mention specific habits, such as asking teams for feedback after each retrospective or keeping a learning journal. Certifications are fine to mention, but experience and reflection matter more.
Follow-up they'll ask: What is one belief about agile you have changed your mind on?
What would you do in your first month with a new team?
Short answer: Observe and listen first, build trust with one-to-ones, understand the team's goals and pain points, then agree a few improvements together.
In depth: Attend events without changing them at first. Map stakeholders and how work flows. Share what you see and let the team choose where to start.
Follow-up they'll ask: What if the team does not want a Scrum Master?