My answer is that a 360 leadership assessment changes behaviour only when it measures behaviour other people can observe, protects the people who rate, and ends with a small visible commitment that someone follows up. Without those three things it produces a tidy report and no change. The tool matters far less than the design around it.
A 360 rarely fails because it was inaccurate. It fails because nobody can act on it by Monday. Colourful charts, averaged scores and a paragraph of themes tell a leader how they are perceived. They rarely say what to do differently in the next executive meeting. That gap between perception and action is what this article is about.
Why a 360 leadership assessment needs more than a report
A 360 is a multirater instrument. In the words of a 2024 paper from researchers at the Center for Creative Leadership and Wake Forest University, leaders rate themselves on competencies and are assessed by several raters such as peers, direct reports and bosses. The same paper, which analyses a sample of nearly 17,000 leaders and nearly 150,000 raters, describes scoring as one of the thorny challenges of the method. Even among specialists, in other words, turning many opinions into a single trustworthy number is difficult.
That should make a buyer cautious about treating the score as the product. The score is raw material. What the leader and the organisation do with it is the product, and the research on feedback is clear that this part can go wrong. Kluger and DeNisi's 1996 meta-analysis, which covered 607 effect sizes and 23,663 observations, found that feedback interventions improved performance on average (d = .41) but that over a third of them decreased performance. Their explanation is that feedback loses effectiveness as it shifts attention away from the task and towards the self.
For anyone running a 360, that finding has a practical consequence. A report that reads as a verdict on the person invites defence. A report that describes behaviour in specific situations invites experiment. The words used in the debrief decide which of the two you get. I coach leaders to move from the question "What does this say about me?" to "What does this show me about how I act in a particular meeting?"
Design the 360 leadership assessment around behaviours people can see
Start with the business problem, not the instrument. Are you fixing uneven leadership quality, preparing a succession slate, or shifting executive behaviour after a merger? Each of those needs a different list of behaviours. I push leaders to define five to seven observable behaviours for the role before anyone completes a survey. A COO might be assessed on decision clarity, cross-functional accountability and escalation discipline. A CHRO might be assessed on talent judgement, courage in conflict and follow-through. If the behaviour cannot be seen in a meeting, a decision, an escalation or a customer outcome, it does not belong in the assessment.
This is the practical end of measuring leadership capability. You are not measuring personality. You are measuring repeatable behaviour under pressure, and I set out how that connects to organisation-wide evidence in measuring leadership capability across an organisation.
The 360 behaviour-to-outcome architecture
- 1. Behaviour design: A short list of behaviours drawn from the real pressure points of the role, not a generic competency model copied from a vendor deck.
- 2. Rater quality: Raters who see the leader in different settings and have enough recent exposure to judge. A direct report in weekly meetings gives better data than a peer who only sees polished presentations.
- 3. Trust and bias control: Clear rules on anonymity, enough raters to protect identity, and a process that reduces social pressure. If the organisation has a history of political use of feedback, address it openly.
- 4. Debrief to insight: A skilled person turns scores into patterns, contradictions and examples. A report is not a debrief.
- 5. Visible follow-up: Two or three commitments tied to recurring work moments, reviewed on a schedule, with a named person who will challenge drift.
The five parts work together. Behaviour design gives the assessment focus, rater quality gives it credibility, trust control gives it honesty, the debrief gives it meaning and follow-up gives it force. Organisations tend to spend heavily on the tool and then starve the debrief of senior time, which is backwards. For the wider system view, see what a leadership operating system does.
How to choose raters, reduce bias and protect confidentiality
A 360 fails quickly when raters are chosen casually or when people believe the results will be used as a political weapon. Raters need regular contact, recent exposure and direct experience of the behaviour in question. Twelve raters with four who have enough exposure is noise. Separate direct reports, peers, managers and internal partners so you can see patterns by relationship. A leader who looks decisive upward and evasive sideways is telling you something about where the pressure sits, and a single averaged score hides it.
- Set rater criteria before names are suggested: Agree the rules first: regular contact, recent exposure and direct experience. Otherwise executives nominate allies and the feedback comes back flattering and empty.
- Use role-based rater groups: Keep groups separate in the report so patterns by relationship type stay visible.
- Watch for response distortion: Inflated scores, identical rating patterns and sudden outliers often signal fear, loyalty or confusion. Context matters: people rate cautiously when they do not know who will remain after a restructure.
- Explain confidentiality exactly: Say how anonymity works, who sees raw data and how comments will be used. If you are vague, people assume the worst and the quality of feedback falls with trust.
- Keep the rater set practical: Too many raters create false precision. Too few create weak signal. The right number depends on role scope and exposure.
Run the debrief so the gap between self-view and other-view becomes usable
The sharpest signal in any 360 leadership assessment is the gap between how a leader sees themselves and how others experience them. I pay close attention to that gap because it usually marks a habit the leader has normalised. Treat it as data, not as a character flaw. Some leaders are unaware of their tone. Others misjudge the effect of their speed. Some believe they are being decisive when colleagues experience being cut off.
A good debrief converts the gap into one experiment. Choose one recurring meeting, one behaviour to test and one person who will report honestly afterwards. The aim is to get from reaction to a visible commitment before the session ends. If the leader is angry, that is information, and a skilled facilitator lets it be spoken while keeping the conversation attached to the work. Hard feedback without follow-up produces resentment. Hard feedback with a clear plan produces movement.
- Translate each theme into one observable behaviour.
- Tie the behaviour to a recurring business moment, such as the weekly executive meeting or a quarterly review.
- Choose one metric, one meeting or one relationship to track.
- Name the person who will challenge drift.
- Review progress every 30 days, not every quarter.
- Stop plans that do not change what others can see.
Keep the number of commitments small. A senior leader leaving the debrief with two or three concrete actions has a chance of doing them. An action like "be more inclusive" will not survive the first busy week. An action like "invite two quieter voices to speak first in the Monday leadership meeting" can be observed by the people affected. I also ask executives to share their commitments in plain language with the three people most affected, because leadership is public and accountability works best when it is visible.
Purpose is the first design decision and the one most often skipped. A 360 used to develop a leader and a 360 used to inform a promotion are different instruments in practice, even when the questionnaire is identical. Raters who believe their comments will cost a colleague a role will soften them, and leaders who believe the same will lobby their raters. Decide which job the assessment is doing, tell everyone involved, and keep the two uses apart. If you want both, run two processes with two communications.
Some leaders handle hard feedback well and many do not. The ones who struggle tend to do one of three things. They explain it away, they challenge the raters, or they promise change and then disappear into their diary. A good debrief anticipates all three. It slows the conversation, separates the emotion from the pattern, and asks where the pattern shows up and who will be able to observe a change next week. The leader does not have to agree with every comment. They do have to leave with one commitment that colleagues can see.
What boards and CHROs should settle before launching a 360
Before any launch, four questions need plain answers. What business problem are we solving? Which behaviours matter in these roles? Who owns the follow-up after the report lands? What will we do if the leader ignores the feedback? If those answers are missing, you are buying activity rather than change, and the process will drift into an annual compliance exercise that the serious people ignore.
| Decision | What to settle | Who owns it |
|---|---|---|
| Purpose | Development, succession input or both, stated in writing before raters are invited | CHRO with the CEO |
| Behaviours | Five to seven observable behaviours per role | Role sponsor and HR |
| Confidentiality | Who sees raw data, who sees themes, how comments are edited | CHRO |
| Follow-up | Debrief owner, 30-day and 90-day reviews, consequence for inaction | Sponsor |
| Link to systems | How results feed succession, performance or team effectiveness | CHRO and executive team |
Governance matters because an assessment that sits outside succession, performance or team effectiveness drifts. I suggest each 360 has a clear owner, a clean timeline and a visible link to one of those systems. If you want help connecting individual evidence to the wider picture, the 360 leadership assessment and the Leadership Capability Diagnostic are the entry points I use, and Capability AI supports the practice between check-ins. For team-level patterns, read building executive teams that perform and how to fix uneven leadership quality.
If you are a CEO, CHRO or board member, stop asking whether your 360 process is comprehensive. Ask whether it changes what people do in the next meeting, the next review and the next decision. If it does, you have a capability system. If it does not, you have an expensive file. The way I work with leadership teams starts from that question.