Mini 360 Feedback Question Builder

A full 360 asks a wide circle of colleagues about many skills. The Center for Creative Leadership (CCL), which sells 360 assessments, describes the usual cycle as every 12 to 24 months or so, covering a broad range of behaviors². It adds that 360 data can also be gathered more often, on fewer behaviors.

A mini 360 is that narrower pass. Goldsmith, whose coaching network uses this process, recommends picking only one to three areas for behavioral change¹. Within six months, his process calls for a two-to-six-item mini-survey with the leader's key stakeholders.

Time it so colleagues have had a chance to notice a change, with time left to act on what they say. If the goals aren't written down yet, settle them first in a leadership development plan.

Most development goals are too abstract to rate. Nobody can score "be more strategic" or "stop being a bottleneck" from the outside. For each goal, ask what a colleague would notice if it were being met. Then name that act and a time window, such as the last 90 days.

Ask about acts, not traits. "Asks for others' views before giving their own" is something a colleague can watch happen. "Is a good listener" asks for a verdict on someone's character.

Keep the set short: one or two questions per goal, plus one open question. Ask about one thing at a time. Pew Research Center, which runs national opinion surveys, warns that questions about more than one concept are difficult to answer and often lead to responses that are difficult to interpret³. "Does she listen and decide quickly?" is two questions.

Next, decide how to show change. One way is to re-ask an item from the earlier 360, word for word. Pew's advice is that when measuring change over time, it is important to use the same question wording³. Reword it, and a shift could come from the wording, not the leader.

The other way is to ask raters directly whether the leader has changed. In a 2004 study, Goldsmith and Howard Morgan analyzed more than 86,000 mini-survey responses for 11,480 managers⁴. At seven of the eight companies, co-workers rated each leader's change in effectiveness on a scale from -3 to +3.

A new question with no earlier version, such as how often something happened, shows only where things stand now. Report it as a first reading.

End with one open question that looks ahead. For follow-up, Goldsmith suggests questions like "Based upon my behavior last month, what ideas do you have for me next month?" It draws suggestions the leader can use, not a debate about the past.

Pick raters who see the leader at work often. CCL advises inviting colleagues who are in the best possible position to observe your behavior on a frequent basis⁵. Usually that means the manager, a few peers and a few direct reports.

Then route each question. A direct report can't see how the leader handles their own manager, and a manager may never sit in on team meetings. Send each question only to people who could have seen the behavior, and note everyone else as skipped, with the reason.

Keep the first round's raters wherever you can, since only they can compare before and after. Add newcomers who see the work, but note when they started: their answers are first impressions.

Have the leader ask each rater personally, explaining the goals and why that person's view matters. CCL notes that this personal outreach will increase your response rate⁵. Then send the questions with the time window, who sees the answers and the deadline.

Count each question's response rate against the people you asked. The National Center for Education Statistics, a US federal statistics agency, divides an item's answers by the number of respondents who are asked to answer that item⁶. If a question went to four of six raters and all four answered, its rate is 100%, not the 67% you'd get by dividing by six.

Measure change on the same people. If the panel changed, a whole-panel average compares two different groups. Work out the change for raters who answered both times, show the whole-panel figure beside it, and put the number of raters next to every score.

With one or two answers to a question, skip the average: read them as individual accounts and note who didn't reply. Across the report, CCL's advice is to look for overarching themes and patterns instead of latching onto any single data point⁵. If there are many written comments, code them into themes. If you interviewed stakeholders instead, pull the themes from your interview notes.

Close the loop with the raters. In Goldsmith and Morgan's study, leaders who discussed their priorities with co-workers and followed up regularly showed striking improvement, with or without a coach. Those without that ongoing dialogue improved barely more than random chance⁴.

The tool at the top of this page does the setup: paste the goals and raters, and River drafts the questions, who answers each and the request.