Empathy Assessment

The Empathy Assessment looks at four dimensions of team health: collaboration, psychological safety, pain point visibility, and team cohesion. Together they tell you whether people share information freely, feel safe telling the truth, raise problems early, and hold together under pressure. The first two questions are about the assessment conversation itself, because the quality of what you learn depends on how that conversation is run. Below is a breakdown of each scored question: what it measures, why it matters, and concrete steps to improve if your team scored low.

Score Interpretation

How to read your total maturity score (0 to 100).

1 to 20: Struggling. People are protecting themselves rather than the work. Mistakes are held against individuals, help is not asked for, and problems surface only when they become unavoidable. Start with how leaders respond to bad news, because nothing else moves until that changes.

21 to 40: Developing. The team functions, but candor is selective. People share within their own role, raise the safe issues, and stay quiet about the ones that matter. Focus on giving pain points a consistent venue and on proving that raising something leads to action.

41 to 60: Norming. The basics hold. Information moves, most people will ask for help, and issues get raised in retrospectives. The next gains come from disagreement upward, clear ownership across roles, and closing the loop on improvement actions so the team sees results.

61 to 80: Performing. This is a healthy team. Trust is real, blockers get swarmed, and leadership knows what is actually hard. Tune the harder edges: comfort with unconventional ideas, resilience after conflict, and connection that goes beyond task coordination.

81 to 100: Thriving. Candor is the default and the team recovers as a unit. People raise problems early because they have seen it work, and they disagree without it costing them anything. Protect this deliberately through team changes and growth, because this state erodes quietly.

Cameras On

Were cameras on?

Why it matters: This assessment rests on what an interviewer can observe, and most of what matters in a team health conversation is not in the words. Cameras are how you catch the pause before someone answers a question about safety, or the glance that says the real answer differs from the spoken one. On an audio only call, hesitation reads as thinking and silence reads as agreement, and both readings are often wrong.

To improve your score: Set the expectation in the invite and give the reason rather than the rule, that this is a conversation where you want to see each other. Turn your own camera on first and keep it on. Keep the session short enough that being on camera is a reasonable ask. If people consistently decline, treat that as information about how the conversation is landing rather than as a compliance problem, and ask directly what would make it easier. In person conversations satisfy this automatically.

Session Length

How long did the meeting last?

Why it matters: Thirty to sixty minutes is the range where trust has time to build and the conversation still holds attention. Under thirty minutes you get the safe answers, because people rarely disclose a real concern in the first ten minutes of talking to someone. Past an hour, energy drops and the interview starts to feel like an interrogation, which changes what people are willing to say.

To improve your score: Book sixty minutes and plan for forty five minutes of content. Spend the first five minutes on context and confidentiality so people know what happens with what they say. Ask your hardest question in the middle rather than at the end, when rapport is highest and the clock is not yet visible. If the conversation needs more time, schedule a second session instead of extending, since a tired participant gives shorter answers.

Proactive Information Sharing

How often do team members proactively share information relevant to others’ work, without being asked?

Why it matters: Proactive sharing is the cheapest form of coordination a team has. When it is missing, information moves only when someone thinks to ask, so people find out about a changed requirement or a broken dependency after they have already built against the old assumption. Teams in that state look busy and produce rework, and the rework is rarely traced back to the communication gap that caused it.

To improve your score: Make sharing a low effort default rather than an act of initiative. Use one visible channel per team for decisions and changes so people do not have to guess who needs to know. Add a standing prompt in the daily standup asking what you learned that someone else should hear. When someone shares something that saves another person time, name it publicly, because recognition is what turns a behavior into a norm.

Response to Blockers

When someone is blocked, how quickly does the team typically mobilize to help?

Why it matters: How fast a team converges on a blocked teammate tells you whether people see the work as shared or as individually assigned. Slow mobilization usually means everyone is protecting their own commitments, and it produces work that sits in progress for days while the rest of the board keeps moving. It also teaches the blocked person that asking is not worth it, which pushes the next blocker underground.

To improve your score: Make blockers visible and expensive to ignore. Flag them on the board, review them first in the daily standup, and set an explicit norm that a blocker older than one day belongs to the team rather than to the individual. Try swarming on the oldest blocked item once a week and watch what it does to cycle time. Pay attention to how the most senior people respond, because their reaction sets the price of asking for everyone else.

Cross Role Coordination

How well does the team coordinate across roles, such as development, QA, and product, versus working in silos?

Why it matters: Silos between roles show up as work that passes a handoff and comes back. QA finds a requirement nobody agreed on, product discovers a technical constraint after committing to a date, and each function concludes the other is the problem. The cost is carried quietly as rework and late surprises rather than showing up as an obvious coordination failure.

To improve your score: Bring the roles together earlier instead of adding coordination meetings later. Use the three amigos approach, where a developer, a tester, and a product representative agree on acceptance criteria before work starts. Write a definition of done the whole team signs, including test and product acceptance. When a handoff repeatedly fails, walk the path together once and fix the specific step rather than escalating the relationship.

Ownership Clarity

How clear is ownership when a task spans more than one person or role?

Why it matters: Shared ownership without a named owner is the most reliable way for work to stall while everyone assumes someone else has it. Ambiguous ownership also makes following up socially awkward, so the item goes quiet instead of getting chased. The pattern is easy to spot in a retrospective, where several action items appear that nobody remembers agreeing to own.

To improve your score: Name one accountable owner for every item that crosses a boundary, even when several people do the work, and record that name where the work lives rather than in meeting notes. Ask the question out loud before any conversation ends: who owns this, and by when. Distinguish who is accountable from who contributes, so a single name never gets read as one person doing all of it alone.

Response to Mistakes

If someone makes a mistake on this team, is it held against them?

Why it matters: This is the closest thing to a single diagnostic for psychological safety, and it predicts almost everything else this assessment measures. When mistakes are held against people, the rational response is to hide them, and hidden problems get expensive. A team that punishes error does not become a team that makes fewer errors. It becomes a team that finds out about errors later.

To improve your score: Run blameless post incident reviews that ask what made this failure possible rather than who did it, and publish what you found. Have leaders describe their own mistakes and what they changed as a result, because the behavior has to be modeled before anyone will believe it. Watch what happens in the first ten seconds after bad news arrives, since that reaction, not the stated policy, is what the team is actually learning from.

Asking for Help

How comfortable are team members asking for help or admitting they don’t know something?

Why it matters: The cost of not asking is invisible and large. Someone spends two days on a problem a colleague could have resolved in twenty minutes, and nobody ever sees that trade. Discomfort here usually means the team has signaled, often unintentionally, that competence means knowing, so people protect their standing by struggling quietly. New joiners and the most senior people are frequently the most affected.

To improve your score: Set an explicit norm, such as anyone stuck for more than an hour brings it to the team, which converts asking from a judgment call into a rule. Give people a low stakes route, like a help channel or a designated buddy, rather than requiring them to interrupt the whole group. Have experienced people ask questions in the open, because one senior person saying they do not know something changes the room more than any policy will.

Raising Concerns and Dissent

How safe do team members feel raising a concern or disagreeing with a decision, even with someone more senior?

Why it matters: Teams that cannot disagree upward make expensive decisions confidently. The information needed to avoid a bad call usually already exists somewhere in the room, and the only question is whether it is safe to say it. Silence here is hard to detect from the outside, because a meeting where everyone agrees looks like alignment and can just as easily be the absence of candor.

To improve your score: Have the most senior person speak last so their view does not anchor the discussion. Ask for the counter case explicitly, for example by asking what would have to be true for this to be the wrong decision. Assign someone to argue the opposite position so dissent comes with a role attached rather than requiring courage. When someone does push back, thank them out loud and say what you changed, because visible impact is the only real proof that dissent is welcome.

Proposing New Ideas

How comfortable is the team proposing new or unconventional ideas?

Why it matters: Comfort with unconventional ideas is a step beyond comfort with dissent, and it is where improvement comes from. A team safe enough to disagree but not safe enough to suggest something odd will keep executing the current approach well and will never find a better one. The usual killers are not hostility. They are the reflexive we tried that before, and the idea that disappears into a backlog nobody revisits.

To improve your score: Separate generating ideas from judging them, so the first pass collects options without evaluation. Give ideas a real path, such as a timeboxed spike or a small experiment with a defined success measure, so proposing something does not require winning an argument first. Track what happened to the last several suggestions and share it, because people watch the outcomes far more closely than they hear the invitation.

Early Surfacing of Obstacles

When a team member hits a recurring obstacle, how likely are they to surface it before it becomes a bigger problem?

Why it matters: Recurring obstacles are the most fixable problems a team has and the least likely to get reported, because each individual occurrence feels too small to raise. A flaky test, a slow approval, a tool that fails twice a week, each one costs twenty minutes and nobody escalates twenty minutes. Across a quarter that is a large amount of capacity disappearing where nobody can see it.

To improve your score: Make small friction easy to record, with a simple visible list where anyone can add an irritant in under a minute. Review that list in the retrospective and commit to fixing one item per iteration so it demonstrably moves. Ask specifically about the second and third occurrence of a problem, since what makes something worth raising is repetition rather than severity.

A Consistent Venue for Pain Points

Does the team have a consistent venue, such as a retrospective or daily standup, where pain points actually get raised?

Why it matters: Without a reliable place to raise problems, surfacing an issue means choosing a moment and a person, and most people will simply not do it. A consistent venue removes that decision. It also creates a shared record, so a problem mentioned by three people over a month becomes visible as a pattern instead of three unrelated complaints.

To improve your score: Hold the retrospective on a fixed cadence and protect it the way you protect planning, because the first event a busy team drops is the one without an obvious deliverable. Vary the format so it does not become a script people can sleepwalk through. Give people a way to contribute in writing before the session, which changes who participates, since the people with the most useful observations are often not the fastest talkers.

From Raised to Resolved

Once a pain point is raised, how often does it lead to a visible action or change?

Why it matters: This question measures whether the team believes raising something is worth the effort, and it quietly determines the score on every other question in this section. Teams that raise issues into a void stop raising them within a couple of iterations. Nothing suppresses candor faster than a retrospective whose output is a list that reappears unchanged the next time.

To improve your score: Cap it at one or two improvement actions per iteration, each with a named owner and a due date, and put them on the board next to the delivery work so they compete for real capacity. Open the next retrospective by reporting on the previous actions, including the ones that did not happen and why. When something is genuinely outside the team’s control, say so plainly and escalate it visibly, because a clear no preserves more trust than silence does.

Leadership Awareness of Pain Points

How aware is leadership or the interviewer of the team’s current top pain points?

Why it matters: Leaders who are unaware of the team’s real obstacles allocate attention to the wrong things and then wonder why delivery is slow. The gap is rarely caused by leaders not caring. It is caused by information being filtered as it moves up, because each layer softens the message slightly, and what arrives at the top is a status that is technically true and practically useless.

To improve your score: Ask the team to name their top three obstacles in their own words and carry those words upward without translating them into safer language. Give leaders a direct and informal route to the team, such as sitting in on a retrospective as a listener rather than a participant. Close the loop by reporting back what leadership did with the last set of issues, because awareness only counts when the team can see evidence of it.

Trust Between Team Members

How would team members describe their sense of trust in one another?

Why it matters: Trust is what lets a team move without verification overhead. Where it is low you see the symptoms rather than the cause: duplicated review, defensive documentation, decisions escalated that the team could have made, and a general slowness no process change seems to fix. Low trust is expensive precisely because it never appears as a line item anywhere.

To improve your score: Build trust through reliability rather than through social events. Make small commitments visible and keep them, and when a commitment is going to slip, say so early instead of at the deadline. Give people work that requires depending on each other, then let them deliver it without a manager brokering the interaction. Where trust was damaged by a specific event, address that event directly, because generic team building does not repair a specific breach.

Goal Alignment and Shared Definition of Done

How aligned is the team on shared goals and what “done” looks like?

Why it matters: Misalignment on done is where a large share of rework starts. One person considers a story finished when the code merges, another expects tests and documentation, and a third assumed it would be deployed. Everyone acted reasonably against a different standard, and the disagreement surfaces at the worst possible moment, usually in front of stakeholders at the demo.

To improve your score: Write the definition of done as a short checklist the team agrees on and keep it visible during planning and review rather than filed in a wiki. Confirm the goal at the start of each iteration in one sentence anyone could repeat, then test it by asking two people separately what the team is trying to achieve. When acceptance criteria are written, have someone other than the author read them back, because ambiguity is invisible to the person who wrote it.

Resilience After Setbacks

How resilient is the team when facing setbacks or conflict? Do they recover as a unit or fragment?

Why it matters: Every team hits a failed release, a missed date, or a real disagreement. What separates teams is what happens in the week afterward. Teams that fragment start protecting their own scope, communication narrows to the minimum required, and the next setback lands on a weaker foundation. Teams that recover as a unit come out with a clearer understanding of the system and of each other.

To improve your score: Debrief setbacks deliberately and soon, focusing on the conditions that produced the outcome rather than on the people involved. Give the team a shared next step quickly, because one specific small win restores confidence faster than reassurance does. Address conflict while it is still about the work, since unresolved disagreement about a decision reliably turns into disagreement about a person.

Connection Beyond Tasks

How connected do team members feel to each other beyond just task coordination?

Why it matters: Connection is what makes the difficult conversations possible. People extend more benefit of the doubt, ask more direct questions, and offer help sooner when they know something about each other beyond a work item assignment. Distributed and frequently reshuffled teams lose this by default, because every interaction is scheduled and every scheduled interaction arrives with an agenda.

To improve your score: Create small amounts of unstructured time rather than large organized events, such as a few minutes at the start of a recurring meeting that is not about status. Pair people across roles on real work, which builds connection faster than any social activity. Notice who is consistently outside the informal conversation, often the newest member or someone in a different time zone, and give that person a deliberate route in.