360-Degree Feedback
- Introduced by
- Manuel London & James Smither; Richard Lepsinger & Anntoinette Lucia
- Year
- 1995
- Type
- Multisource performance-feedback measurement and development process
- Family
- Organizational behavior / performance feedback and development
What does it do?
360-degree feedback, also called multisource feedback, gathers ratings of a focal person's work behavior from the full circle of people around them, self, supervisor, peers, and direct reports, and sometimes customers, rather than from a single boss.
The aim is to give a more complete, less biased picture of performance and, especially, to prompt self-awareness and development by revealing how different observers see the same person. Manuel London and James Smither's 1995 article gave the practice a theoretical foundation, explaining when and why multisource feedback might change self-evaluations, goals, and behavior. Richard Lepsinger and Anntoinette Lucia's 1997 practitioner book laid out how to design and implement sound 360 systems. Conway and Huffcutt's 1997 meta-analysis documented the psychometric reality that different sources agree only modestly, which is precisely what makes multiple perspectives informative. 360-degree feedback became one of the most widely adopted leadership-development tools in organizations.
The intuition behind the method
A single supervisor sees only one slice of how a person works. They watch how an employee manages up, but rarely how that same person treats the people who report to them, collaborates with peers, or behaves when the boss is not in the room. Relying on one viewpoint is like judging a sculpture from a single angle. 360-degree feedback walks around the sculpture, collecting impressions from every direction, above, beside, and below, plus the person's own self-view. The value comes not from averaging these into one number but from comparing them. When a manager rates their own listening skills highly but their direct reports rate them low, that gap is the insight: it points to a blind spot the person could not have seen alone. The fact that sources disagree is therefore a feature, not a bug, because each vantage point captures genuinely different behavior. Delivered well, this layered, comparative picture creates the self-awareness that is the first step toward changing how someone actually behaves at work.
History
Multisource appraisal ideas appeared as early as mid-twentieth-century work on peer and subordinate ratings, but 360-degree feedback became a distinct, widely adopted practice in the late 1980s and 1990s as organizations sought richer, development-oriented alternatives to top-down appraisal. Manuel London and James Smither's 1995 Personnel Psychology article supplied a theoretical framework, drawing on control theory and goal-setting to specify when multisource feedback would change self-evaluations and behavior, and shifting the conversation from mere measurement to the psychology of feedback use. Richard Lepsinger and Anntoinette Lucia's 1997 book translated the approach into practical design and implementation guidance for practitioners. Conway and Huffcutt's 1997 meta-analysis grounded the field empirically by documenting the interrater reliabilities within and correlations between sources, clarifying that modest cross-source agreement is inherent to and informative for the method. The practice has since become a fixture of leadership development worldwide.
When to use it
Use 360-degree feedback primarily for development, to build self-awareness and guide the growth of managers and leaders, when multiple people genuinely observe the focal person's behavior and the organization can support the process with facilitation and follow-up. It is well suited to leadership development, coaching, and culture-building around feedback, and to roles where peers and subordinates see behavior the supervisor cannot. It is far more controversial and risky when used for high-stakes administrative decisions such as pay or promotion, because the anonymity, motivation, and rating behavior of sources change when consequences attach. It is poorly suited where rater pools are too small to preserve anonymity, where the culture cannot tolerate candid upward feedback, or where no follow-up support is provided.
Limitations and considerations
- Different sources agree only modestly, so a single combined score can obscure the divergent perspectives that give the method value.
- Ratings can be distorted by leniency, friendship, politics, and reciprocity, especially when used for administrative decisions.
- Effectiveness depends heavily on delivery, coaching, and follow-up that organizations often fail to provide.
- Evidence that 360 feedback reliably improves performance is mixed, with effects modest and contingent on many conditions.
Source conditions do not replace independent scientific review or validation by an analysis engine.
Strengths
- Captures behavior from multiple vantage points, giving a fuller and less single-source-biased picture than supervisor ratings alone.
- Surfaces self-other gaps and blind spots that are powerful catalysts for self-awareness and development.
- Incorporates upward and peer perspectives that the supervisor cannot observe, broadening the behaviors assessed.
- When paired with goals, coaching, and follow-up, has theory and evidence supporting genuine behavior change.
Applications
- Developing managers and leaders by revealing how different stakeholder groups perceive their behavior.
- Building self-awareness through structured comparison of self-ratings with others' ratings.
- Supporting executive coaching with concrete, multi-perspective behavioral data.
- Diagnosing team and organizational competency gaps by aggregating across many focal individuals.
- Fostering a feedback culture and normalizing upward and peer input on behavior.
How it works
How it works
A 360 process begins by defining the competencies, the behavioral dimensions of effective performance, on which the focal person will be rated, ideally derived from job or competency analysis so they reflect what the role and organization actually require. Lepsinger and Lucia stress that these competencies must be behaviorally specific and observable by the rater groups who will use them, since a peer and a subordinate can only rate what they can see. Getting the competency model right is foundational: it determines what the instrument measures and whether the feedback will be relevant, credible, and actionable for the person receiving it.
Formula
C = \{c_1,\dots,c_p\} \text{ behavioral competencies from job analysis}Ratings are then collected from multiple sources occupying different relationships to the focal person, typically self, supervisor, peers, and direct reports, each rating the same competencies from their own vantage point. The deliberate inclusion of several perspectives is the defining feature of the method, because each source observes the focal person in different situations and roles. Lepsinger and Lucia describe selecting appropriate raters, ensuring each has sufficient opportunity to observe, and designing the instrument so the questions make sense to every source, all of which shape the quality and usefulness of the data gathered.
Formula
r_{sc} = \text{rating of competency } c \text{ by source } s,\ s\in\{\text{self,sup,peer,sub}\}The individual ratings are aggregated within each source category, with peer and subordinate ratings averaged across raters and, crucially, kept anonymous, usually requiring a minimum number of raters such as three per group so that no individual can be identified. Conway and Huffcutt's meta-analysis informs this step by quantifying the interrater reliability within sources, which differs across source types and is often modest, meaning several raters are needed per source for a stable average. Aggregating by source rather than pooling everyone preserves the distinct perspective of each group, which is what the later comparison depends on.
Formula
\bar{r}_{sc} = \frac{1}{n_s}\sum_{k} r_{skc},\quad \text{anonymity for } n_s \ge 3 \text{ non-self raters}The aggregated source scores are then compared, with particular attention to discrepancies between self-ratings and others' ratings and among the different other-source ratings. London and Smither's theory highlights these gaps as the engine of the process: a positive self-other gap signals overestimation and a potential blind spot, while agreement across sources lends confidence to the picture. Conway and Huffcutt's finding that sources correlate only modestly underlines why these comparisons are meaningful, since the perspectives are genuinely different rather than redundant. Analyzing the gaps turns a set of ratings into developmental insight.
Formula
\Delta_c = \bar{r}_{\text{self},c} - \bar{r}_{\text{other},c},\quad \Delta_c>0 \Rightarrow \text{overestimation}Finally, the results are delivered to the focal person, ideally with a coach or facilitator, and translated into development goals and action, since feedback alone rarely changes behavior. London and Smither's theory-based account specifies the conditions under which multisource feedback affects self-evaluations, goal setting, and subsequent performance, emphasizing that the person must accept the feedback, set specific goals, and receive support and follow-up. This delivery-and-action step is where the method either succeeds or fails: well-facilitated feedback linked to goals and follow-up can change behavior, while a report handed over with no support typically does not.
Formula
\text{feedback} \to \text{self-awareness} \to \text{goals} \to \Delta \text{behavior (theory-based)}Common pitfalls
- Using 360 feedback for pay or promotion, which corrupts rater motivation and undermines the candor the method needs.
- Pooling all sources into one average and discarding the between-source differences that carry the developmental signal.
- Allowing rater groups too small to protect anonymity, which suppresses honest upward and peer feedback.
- Handing over the report with no facilitation, goal-setting, or follow-up and expecting behavior to change on its own.
Frequently asked questions
Why do different rating sources disagree, and is that a problem?
Conway and Huffcutt's meta-analysis showed that ratings from supervisors, peers, subordinates, and the self correlate only modestly with one another. This is largely because each source observes the focal person in different situations and roles, so they are genuinely seeing different behavior rather than measuring the same thing with error. For developmental purposes this disagreement is the point: the gaps between perspectives, especially between self and others, reveal blind spots and are the most useful part of the feedback. Disagreement becomes a problem only if one tries to collapse the sources into a single 'true' score, which discards exactly the comparative information that makes 360 feedback valuable.
Should 360-degree feedback be used for pay and promotion decisions?
Most experts advise caution. 360 feedback was designed primarily for development, and its key conditions, candid raters protected by anonymity, willingness to give honest upward feedback, depend on the absence of high stakes. When ratings drive pay or promotion, raters become lenient or strategic, reciprocity and politics intrude, and the candor that makes the method useful collapses. London and Smither's work and subsequent research emphasize that the psychology of feedback use differs sharply between developmental and administrative contexts. Some organizations do use 360 input administratively, but doing so requires great care, and many keep it strictly developmental to preserve the integrity of the ratings.
Does 360-degree feedback actually improve performance?
Sometimes, but the effects are modest and conditional rather than automatic. London and Smither's theory specifies that change requires the recipient to accept the feedback, set specific goals, and receive ongoing support and follow-up; without these, simply delivering a report rarely changes behavior. Meta-analytic and longitudinal research finds small average improvements, larger for those who initially overrate themselves and who engage with the feedback, and dependent on coaching and a supportive climate. The practical lesson is that 360 feedback is not a stand-alone intervention but the front end of a development process, and its value is realized only when the action and follow-up steps are taken seriously.
Related methods
Sources
Lepsinger, R., & Lucia, A. D. (1997). The Art and Science of 360 Degree Feedback. Pfeiffer.
ISBN: 9780787909581Copied.London, M., & Smither, J. W. (1995). Can multi-source feedback change perceptions of goal accomplishment, self-evaluations, and performance-related outcomes? Theory-based applications and directions for research. Personnel Psychology, 48(4), 803-839.
Open DOI referenceCopied.Conway, J. M., & Huffcutt, A. I. (1997). Psychometric properties of multisource performance ratings: A meta-analysis of subordinate, supervisor, peer, and self-ratings. Human Performance, 10(4), 331-360.
Open DOI referenceCopied.