What Is a 360 Assessment? Format, Questions and Scoring (2026)

Somebody has asked you to run a 360 assessment. Maybe your manager, maybe a client. Either way you now need to know what one is, and what you're supposed to hand back at the end.

A 360 assessment asks the people around a colleague what they think of the way that person works. Their manager. Their peers. The people who report to them. And the person themselves.

You collect all of those answers into one report. Then you look at where they disagree.

What a 360 Assessment Is

You will see the same thing called 360 degree feedback, a multi-rater assessment, or a 360 review. It all means one thing: several people answer the same questions about one person.

Four groups usually answer.

• The person being rated, scoring themselves.

• Their manager, whose answers you keep separate.

• Their peers, usually four to eight of them, pooled into one score.

• Their direct reports, if they manage anyone, also pooled.

Some cycles add clients or partners. That matters for a consultant or an account lead, because most of their work happens outside their own team.

You ask about behaviour, not results. "Explains the reasoning behind a decision" is a good 360 question. "Hit 94% of quota" isn't. A colleague can't watch somebody hit a number, and your CRM already knows the answer.

The finding is the gap between how somebody sees themselves and how everyone else sees them.

Say a manager gives herself 4.6 on listening. Her team gives her 2.9. She has just learned something no performance review would have told her, and she learned it from a number instead of an argument.

How a 360 Assessment Differs From a Performance Review

Three things get mixed up here, and the difference decides what you build it in.

An engagement survey asks everybody the same questions about the company. Wide and shallow. An online survey builder handles it comfortably.

A 360 is the other shape. Several people answer about one person, and afterwards you have to bring those answers back together. That's the part that catches people out.

One rule matters more than the others: keep it away from pay. The moment a rating decides a bonus, people rate to protect each other. Every score slides to the middle and you learn nothing.

Who Runs 360 Assessments

Mostly HR teams. You run a cycle for your managers once or twice a year, and the scores feed a development plan.

The questions are the easy part. The work is everything around them: picking raters, chasing the ones who haven't answered, building a report for each person, then getting that report in front of them. That's what involve.me for HR services is for.

Learning and development teams run them around a leadership programme. One at the start, one a year later. The second round is the only way to see whether anything changed.

Consultants run them for clients, which is a different job. That one is further down.

What Goes Into a 360 Assessment Question Set

People give up on long forms, so keep it tight. Five parts:

1. Four to six competencies. Communication, decision making, developing others, dependability. Use the words your company already uses.

2. Three to five statements per competency. Each one should describe something a colleague could have watched happen. "Explains the reasoning behind a decision" beats "is an effective communicator". The second one asks the rater to make the judgement you wanted to measure.

3. A rating scale with labels. Five points, and write out what each one means. Otherwise a 3 from a manager and a 3 from a direct report are two different things.

4. A "not observed" option on every question. Leave it out and a peer who has never seen the behaviour picks the middle of the scale. Now your average is wrong and you can't tell.

5. Two open questions. What should this person keep doing, and what should they do differently. People read the comments first, so ask for them while your raters are still paying attention.

Six competencies is my ceiling. Past that, people stop reading the questions and start filling in the grid.

Ask for the rater's group first, because it changes what you should ask them next. A direct report can't answer a question about how their boss handles her own boss. Ask anyway and you get a wall of "not observed". Send each group to its own set of questions with conditional logic and everyone gets a shorter form they can actually answer.

Stuck on wording? Our list of employee feedback form questions is a faster start than a blank page, and most of it works in a 360 with small edits.

How a 360 Assessment Is Scored

The maths is simple. One rule: never blend the groups into a single number.

1. Give every point on the scale a value, usually 1 to 5. Drop the "not observed" answers instead of counting them as zero.

2. Average each competency inside each group. Now you have a manager score, a peer score and a direct report score, side by side.

3. Put the person's own score next to those. You're comparing it, not adding it in.

4. Sort the competencies by what the peers and direct reports said. The bottom two or three are the ones worth working on.

5. Keep each comment with the competency it was written about, so the number and the reason arrive together.

In involve.me the number comes from the score element, and it has two settings. "Correctly Answered Questions" counts right answers, one point each, which is what a knowledge test needs. For a 360 you want "Individual Score". It adds up whatever the rater picked, and no answer is right or wrong. Switch on "Individual Score & Calculation" on each question, give every point on your scale a number, and build the competency averages in the assessment builder.

Our own support desk hears about this a lot. Setting up calculations and score-based outcomes came up at least 20 times across every support conversation this summer. The hard part is never the maths. It's deciding what a 3 means before anybody starts rating.

What the 360 Assessment Report Should Show

This is where most cycles are won or lost. You can hand somebody a table of group averages. They will look at it once and do nothing.

Four things belong on the page they see:

• Their own score next to each group's score, competency by competency, so the gaps are obvious.

• A picture of each score, a gauge or a bar, so strong and weak stand apart.

• The written comments, grouped by competency, exactly as people wrote them.

• Two or three things to actually do, tied to the lowest scores.

In involve.me, you build that with Score-based Outcomes, which send different score bands to different outcome pages, plus the "Graph" and "Scorecard" elements. Those turn a number into a gauge, a ring or a bar chart, and you can set bands so a score reads as Low, Medium or High in your own words. It's the difference between sending somebody a development plan and sending them a spreadsheet.

Once again, our support team gets asked for exactly this. How do I run a questionnaire and give people a results page that changes with their answers came up 25 times across every support conversation this summer. In a 360, that results page is the job.

One thing to plan for. Each rater's answers arrive as their own submission, so you pull the per-person picture together from the Responses tab or an export. And the conversation after that, where a manager and their report work out what to do about a 2.9 on developing others, happens between two people and no tool. Book those conversations before you send anything out.

Keeping a 360 Assessment Anonymous

You told your raters that nobody would see who said what. Here's the number that keeps your promise: three.

Below three, the maths gives people away. Two peers, one score of 2, and everybody in the room can work out who wrote it.

1. Pool the peers, pool the direct reports, and report the manager on their own. The person already knows who their manager is.

2. If a group comes back with fewer than three answers, hold it back instead of showing a thin average.

3. Tell everyone the rule before you open the cycle. People answer more honestly when they know where the line is.

4. Collect nothing that identifies the rater. Your rater funnel doesn't need a "Contact Form" element at all, and leaving it out is worth more than any promise in the invitation email.

Name the person being rated, not the person rating. A hidden field in the URL carries who the assessment is about, so one funnel covers your whole cycle. Send each rater a link ending in something like ?subject=jane-doe&rater_group=peer. The value is saved with the submission, and you can drop it straight into the questions so raters read "Rate Jane on...". involve.me can also keep personal data separate from the answers in your funnel settings, which is worth switching on if you collect any contact details at all.

Our guide on how to collect feedback anonymously covers the same ground for a normal survey, including the settings that quietly break the promise.

How to Run a 360 Assessment Cycle

Building it takes a day. The cycle is where your time goes. Most of that you can automate once.

1. Pick who is being rated, and set a close date. One group at a time. A cycle with no deadline collects half its answers and stops.

2. Choose the raters. The person suggests them, their manager approves the list. Let people pick alone and you get a friendly panel and a flattering report.

3. Build the assessment once. One funnel, one set of questions, the rater group as question one, the person being rated in a hidden field.

4. Send everyone their own link. The name in the URL is what lets one funnel serve the whole group.

5. Chase the people who haven't answered. Two reminders, spaced out, is what works.

6. Hold back the thin groups and build the reports. Anything under three answers stays out.

7. Book the debriefs before you send a single report. A report that arrives with no conversation attached is where this all quietly dies.

Steps four and five are what automated email sequences are for. Build one workflow on the "Completed Submission" trigger, put a "Send Email" node on it to thank the rater, then a "Wait" node and a "Send Internal Email" node so you hear when somebody's answers are all in. No refreshing a dashboard for two weeks.

A template saves you the two weeks most teams spend arguing about competency wording. It's much easier to argue with questions once somebody has written them down.

For the wording itself, our free questionnaire templates are a useful second pass. The subject is different, the phrasing habits carry over.

Running 360 Assessments for Client Organisations

For a consultant this is billable work, not an internal process, and three things change.

You don't pick the panel. You're working with somebody else's org chart, so write the nomination step down and agree it. A client who picks their own raters picks kindly, then argues with the report.

The report is what they're buying. Inside an HR team a plain summary survives, because the debrief carries it. A client is paying for the document, so the outcome page and the export both have to look worth the invoice.

The follow-up is the offer. A cycle that ends with a report is a project. A cycle that ends with the next one booked for twelve months out, and development work in between, is a retainer. Build the second assessment at the same time as the first, on the same questions, and that conversation gets easy.

Six Mistakes That Sink a 360 Assessment

• Too many questions. Forty questions, six raters, a dozen people being rated. That's how a cycle stalls at 40% answered.

• A scale nobody defined. If you never said what a 3 means, your averages are noise with decimal places.

• The manager pooled in with the peers. One score inside an average of five bends it, and the manager's view is the one your reader most needs to see on its own.

• Letting it touch pay. Ratings flatten out the moment money depends on them.

• Promising anonymity you can't keep. Two answers in a group isn't anonymous, whatever your invitation said.

• No debrief. The report is the start of a conversation. Without the conversation it's a PDF in a folder.

Everything above is the part you can build. Whether the cycle was worth running gets decided in a room, by somebody who has just read a 2.9 on developing others and wants to know what to do on Monday.