In this article
Before you start
This question belongs to the engineering manager interview questions guide, where the people-management round is mapped.
Why interviewers ask this
The question separates managers who label people from managers who diagnose problems. Interviewers in the people-management round, as described in what the people management round is, want to see whether you can tell two different problems apart before you act on either.
A weak answer says "I would talk to them and make them work it out," which treats a missing skill and a personal clash as the same thing. A stronger answer names how you would find out which one you face, what you would do for each, and how you would know it worked. A Staff-level answer adds the cost of waiting: personal friction is not harmless, and the evidence below says so. It also states when you stop coaching and move to a formal process, because a plan with no end point is only a hope.
Skills gap versus personality conflict from first principles
- Skills gap
- A skills gap is a difference between what a role needs a person to do and what that person can do today, which training, practice and stretch work can close without any change in attitude.
- Relationship conflict
- Relationship conflict is a kind of team conflict that is about the people rather than the work, shown as irritation, mistrust or dislike in how they treat each other, and which a manager can address by describing specific behaviors and agreeing rules.
In plain words
Picture a kitchen where a dish keeps coming out wrong. Either the cook has not learned the technique, or the cook knows it. Then the waiter keeps changing the order without telling anyone. The first needs a lesson. The second needs a talk between two people. The analogy breaks in one place. In a team both can be true at once. A weaker reviewer and a faster author can annoy each other until a gap feeds a conflict.
What managers call a personality conflict, research calls relationship conflict; this page uses "relationship conflict" for it.
In one precise sentence: close a skills gap with teaching, practice and a checkpoint, and settle a relationship conflict by turning it into specific behaviors and agreements that the people involved own.
Step 1: Separate the skill, the will and the relationship
Diagnosis starts from behavior you or others observed, not from a verdict about the person. "She is difficult" is a verdict. "She rejected three reviews in a row without comment" is a behavior. For any problem, ask three questions: can the person do the task to the standard, do they want to, and does something between them and a colleague block the work?
Input: the observed behaviors and their effect on the work. Output: one of three working labels: skill, will or relationship. Ask the person in a one-on-one before you decide, and ask a neutral third person what they saw. Several causes can overlap, so write down the main one and what would change your mind. You now have a label to act on, and the next step differs for each.
Step 2: Match the action to the cause
A skills gap gets teaching, pairing and stretch work. Name the specific skill and the standard, give the person a smaller task that exercises it, pair them with someone strong, and agree how you will look at the result. The growing engineers questions cover the coaching side in more depth.
A relationship conflict gets a different sequence: talk to each person separately, describe the behavior and its impact on the team, then bring them together around the shared goal and write working agreements. For the sharper version of that scene, see two engineers who can barely stand each other.
Input: the label from step 1. Output: a plan with an owner and a date. If the label is will, the conversation is about expectations and consequences, which is the territory of handling a low performer.
| Question | Skills gap | Relationship conflict |
|---|---|---|
| Evidence | Work misses the bar | Friction between two people |
| First move | Name the skill and the bar | Talk to each separately |
| Main tool | Teaching, pairing, stretch work | Mediation, working agreements |
| Check | Work quality over weeks | Handoffs run smoothly |
Step 3: Check the result and escalate if it does not move
A plan is only a plan until the outcome changes, so decide the check before you start. For a skills gap, look at the quality of the work a few weeks later, not at how hard the person tried. For a relationship conflict, look at ordinary handoffs: do reviews get requested and answered, and does the rest of the team stop routing around the pair?
Input: the plan and the agreed signal. Output: continue, adjust or escalate. Escalation has three rungs: coach or mediate first, adjust the plan once, then use a formal process for repeated missed expectations or a change of reporting lines for a conflict that agreements do not settle. Do not wait for the problem to age: De Dreu and Weingart (2003) report "strong and negative correlations between relationship conflict, team performance, and team member satisfaction." Correlation is not cause, but it gives no support for letting personal friction run.
Step-by-step description
- Observed behavior points to a cause.
- Teach, reset expectations or mediate.
- Look at the result, then adjust or escalate.
- A skills gap gets teaching and stretch work, a relationship conflict gets mediation and agreed rules, will gets expectations, and all end in a check of the result.
- Diagnosis starts from observed behavior and points to a cause.
Text description of the diagram
- Escalation has three rungs: coach or mediate first, adjust the plan once, then use a formal process or a change of reporting lines.
- Escalation has three rungs, and the formal process comes after you adjust the plan once.
Is some conflict useful?
Task conflict is disagreement about the work itself, such as which design to build. Reported findings differ. De Dreu and Weingart (2003) found task conflict negatively correlated with performance, with correlations they describe as "strong and negative (instead of the predicted positive)". De Wit, Greer and Jehn (2012) found the link to group performance near zero overall and sometimes positive, for example in top-management teams and when task and relationship conflict are only weakly correlated. The safe reading is that debate about the work is neither automatically harmful nor automatically helpful. Give it an owner and a deadline, and watch whether it turns personal.
Knowing it is half the job. Saying it is the other half.
Interview Clozo is an AI interview copilot for real interview rounds: it hears the question and puts a strong answer on your screen, and you say it in your own words. See the features or how it works.
Start Free Trial →The Staff-level answer
How would you address a skills gap or a personality conflict on your team?
I start by asking which problem I actually have, because the fix differs. If the work misses the bar, I name the specific skill and the standard, agree a plan with practice, pairing and a stretch task, and check the work in a few weeks. If the skill is there but the effort is not, I reset expectations and consequences.
If two people clash, I talk to each separately, describe the behavior and its impact on the team, bring them together around the shared goal, and write down working agreements. The trade-off is time: coaching and mediation cost me hours now, but personal friction is strongly linked to worse team performance and satisfaction, so leaving it alone costs more. If the plan does not move the outcome, I move to a formal process.
What makes it Staff-level: it diagnoses before acting, gives each cause its own action, states the cost of waiting with evidence, and names the point where coaching stops. Say it in the order diagnose, act, check, escalate.
The rubric
| Dimension | Weak answer | Strong answer | Staff answer |
|---|---|---|---|
| Diagnosis | "I would talk to them." | Asks what the cause is. | Separates skill, will and relationship from observed behavior. |
| Action | One generic fix for everything. | Coaching or mediation, chosen per case. | Plan with owner, date and working agreements. |
| Trade-offs | Not mentioned. | Mentions the time cost. | Cost of waiting, with evidence, set against the cost of acting. |
| Follow-up readiness | Stuck when asked "what if it fails?" | Mentions escalation. | States the check, the signal and the formal next step. |
Three variants of this question
A report keeps missing the bar and says the code review feedback is unfair
A missed bar plus a complaint about review tone mixes a possible skills gap with a relationship signal. Check the reviews against the written standard first; if they hold up, treat it as a skills gap, and if the reviewer's tone is the problem, address that behavior with the reviewer separately.
Two senior engineers argue in every design meeting
Ask whether the argument is about the work or about each other before you step in. If it is about the work, give the decision a clear owner and a deadline; if it is personal, use separate conversations and working agreements.
What if the skills gap is yours as the manager?
Name it, find someone who has the skill, and ask for feedback on that area. A manager who hides a gap loses credibility; one who names it and closes it models the behavior asked of the team.
Common mistakes
- "She is just difficult." (wrong) That is a verdict about a person. (right) "In the last three reviews she rejected the change without comment, and the author waited days."
- "I would let them fight it out." (wrong) The evidence links personal friction to worse team performance and satisfaction, so doing nothing is a choice with a cost. (right) "I would step in early with separate conversations and agreed rules."
- "I would send them on a training course." (wrong) A course is an input, not an outcome. (right) "I would agree the skill, a stretch task and a check on the work itself."
- "Disagreement is always healthy." (wrong) Reported findings differ: De Dreu and Weingart (2003) found task conflict negatively correlated with performance, and de Wit, Greer and Jehn (2012) found the link near zero and sometimes positive. (right) "Debate about the work needs an owner and a deadline."
- "I would fix it myself and not tell anyone." (wrong) The team still sees the pattern. (right) "I would tell the people involved what I saw and what I expect."
Follow-up questions
How do you tell a skills gap from a motivation problem?
Ask whether the person can do the task at the standard when the stakes are clear and support is given. If the work improves with teaching and a stretch task, it was a skill; if it does not improve although the skill is there, the conversation is about expectations and what is blocking them.
When do you step in on a conflict and when do you let it play out?
Step in when the friction shows in behavior toward people, or when the work slows. Let a disagreement about the work play out only if it has an owner, a deadline and respectful tone. The team health questions cover the signals that tell you which case you are in.
What if the person with the gap is senior?
The approach is the same, but the framing changes: ask for their view of the gap first, agree the standard together and make the support specific. A senior person often closes a gap faster when it is treated as a shared problem, not a ranking.
How do you know the plan worked?
You agree the signal before you begin: the quality of the work for a skills gap, and smooth handoffs between the two people for a conflict. If the signal does not move after the checkpoint, adjust the plan once and then escalate.
- Diagnose from observed behavior whether the cause is skill, will or relationship before you act.
- A skills gap gets teaching, pairing and stretch work; a conflict gets mediation and working agreements.
- Decide the check and the escalation point before you start the plan.
- Evidence links personal friction to worse team performance and satisfaction, so waiting has a cost.
- Describe behaviors and effects, never verdicts about the person.
Sources
- Task versus relationship conflict, team performance, and team member satisfaction: a meta-analysis (De Dreu and Weingart, 2003) — Journal of Applied Psychology 88(4), 741-749, 1 Aug 2003
- The paradox of intragroup conflict: a meta-analysis (de Wit, Greer and Jehn, 2012) — Journal of Applied Psychology 97(2), 360-390, 1 Mar 2012
- Task versus relationship conflict: a meta-analysis (De Dreu and Weingart, 2002 conference version) — Academy of Management Proceedings, 1 Aug 2002
