360 feedback when you have no company
What the method gets right, the three places it breaks, and what to do if nobody is going to run one for you.
Somebody at your last job ran a 360 on you. You got a deck with radar charts and a column of scores against a competency model. There was a page of anonymous comments, two of which you could tell exactly who wrote.
You read it twice, felt slightly bruised, and did nothing differently. Then you left, and now there is no HR department to run another one, and you would quite like to know what people actually think.
What a 360 is
A structured way of collecting ratings about one person from the people around them. A manager, some peers, some direct reports, sometimes a customer. Everybody rates the same person against the same list of competencies, usually on a numeric scale, usually anonymously.
The whole method rests on one true idea. The people around somebody can see things about how they operate that the person cannot see about themselves, so ask several of them and compare.
That idea is correct and it is the same idea we are built on. What is worth arguing about is the machinery bolted to it.
What it does well
It gets several perspectives at once, from people in genuinely different positions relative to you. Somebody who reports to you sees things your manager never will, and a good 360 puts those side by side.
It is comparable. Because everyone answers the same items on the same scale, a score can be tracked over two years, and a group of managers can be looked at together. If you need to compare a hundred people, this is the method.
It also has authority. It arrives through a process, from an institution, which makes it harder to dismiss than a friend's opinion. That matters more than it should and it does matter.
Where it breaks
The scale eats the observation
Asked to rate somebody on communicates effectively from one to five, a colleague has to convert everything they know into one digit. What made the four rather than a three is exactly the part you needed, and it is gone before the data is collected. You end up with a number and a competency name, and no idea what anybody was picturing when they chose it.
Anonymity that is not anonymous
With four raters and one written comment box, most people can identify the author of most comments. Everybody involved knows this, so the comments get written with that in mind and the useful ones do not get written at all. The result is a document that is technically anonymous and functionally signed, which is the worst of both.
It is inside a power relationship
Answers about your boss are given by people whose pay that person influences. Answers about a peer are given during a promotion cycle. The material is being collected inside the same system that will use it, and everybody answering knows that.
The evidence on whether it changes anything is mixed
Decades of research on multi-source feedback has not produced a clean finding that receiving it reliably improves behaviour. Some people improve, some do not, and the conditions that separate them are still argued about. We are not going to attach a percentage to that, because we would be making it up.
What made the four rather than a three is exactly the part you needed, and it is gone before the data is collected.
How to read a 360 you have already been given
Most people read one of these looking for their score. That is the least informative thing in the document, and here is what to do instead.
Look for the split, not the average
If four people rated you a two and one rated you a five, the average is a respectable three and it describes nobody's experience of you. The three is an artefact of arithmetic. It is not an observation anybody made. The disagreement is the finding: it usually means you are one person with one group and a different person with another.
Read the lowest score from the group with the least power over you
Your manager's rating is written by somebody who has to keep managing you, and who may have to defend the number in a calibration meeting. The ratings from people who report to you had the most reason to be careful, so anything critical that survived got through despite that pressure. A low score from that direction is the one to take seriously.
Expect leniency and a crowded middle
Raters are lenient, meaning scores drift high, because a low number feels like an accusation and takes effort to justify. Raters also cluster in the middle, avoiding both ends. Between the two effects, the apparent difference between a 3.6 and a 3.9 is mostly noise. So do not rank your competencies by score and work on the lowest: that ranking is largely an artefact.
If a finding is too abstract to act on, go and get the episode
Needs to develop executive presence is not something you can do anything about on Monday. The scale converted somebody's actual memory into a phrase, and the memory is the part you needed. Go back and get it from one person you trust, asked about one occasion, in private, weeks after the deck landed.
SayThe 360 said something about executive presence and I cannot act on that. Can you think of one meeting where you saw it? Just what I did.
Ask about the occasion, not about the score. Nobody wants to defend a number they gave anonymously, and asking them to will end the conversation. Asking what happened is answerable.
What you can do without a company
You have one advantage the corporate version does not: nobody's pay is involved. People who are not being evaluated alongside you can afford to be straight with you, if you make it safe.
Pick people who see you in different settings rather than the four people from one team. Ask about specific occasions rather than competencies. Ask more than one person the same question, because the value is in the overlap. The guide on asking has the exact wording that works.
How what we do differs
We ask five people you choose, and we do it in a conversation rather than on a scale. Nobody rates you on anything. Each person is asked what happened, and then asked about the specifics of what they just described.
The anonymity is real rather than nominal. You never see a sentence any of them wrote, so nobody is writing with an eye on being identified. The single exception is a common word supplied independently by three or more of them, which cannot be traced to any one person.
The trade is honest and worth naming. We cannot compare you to a hundred managers, we produce no score to track across two years, and we are not a performance process. If you need those things, you need a 360 and this is not it.
What we produce instead is the comparison between what you said about yourself and what five people said separately. How that works is set out in full, including what gets refused.
The two methods side by side
| A 360 review | What we do | |
|---|---|---|
| What each person is asked to produce | A rating on a competency, usually one to five | An account of what happened, then questions about the specifics |
| What survives collection | A digit and a competency name | The occasion, in the words of somebody who was there |
| How anonymity behaves | Technically anonymous, functionally signed with four raters and a comment box | You never see a sentence any of them wrote, so nobody writes with an eye on being identified |
| Who is in the room | People whose pay and promotion the subject influences, inside the system that will use it | Five people you choose, with nobody's pay involved |
| What you can compare against | A hundred managers on the same items, and yourself two years ago | Nothing. There is no score and no benchmark |
| Where the authority comes from | A process and an institution, which makes it harder to dismiss | Only the material itself, which is the harder position to hold |
| What it is for | A performance process, tracked over years | One comparison: what you said about yourself against what five people said separately |
This page compares two methods of collecting information about a person. It does not assess or name any specific commercial instrument.
The part this page cannot do
Choosing a method is not the hard part of this. Reading something true about yourself that you did not already suspect is the hard part, and no page can do it for you.
The thing both methods are reaching for is the gap between what you meant and what people actually experienced. That gap only becomes visible when several people describe it without knowing what the others said.
Start the interviewIt takes about fifteen minutes and it is free. The $29 Portrait is only charged after five people finish and the evidence clears.