How this works

A person answers questions about themselves. Five people who know them answer questions about them, separately, without seeing each other. Then we compare the two and write about the difference.

That comparison is the whole product. Every other test in this category has one source, which is you, and it can therefore tell you what you already think. We have six sources and five of them are not you.

One person can be wrong about you in a way that says more about them than about you. Fifty people would mostly not know you well enough to say anything specific, and the specific part is the only part worth reading.

Five is the number where a claim can be checked. Say four of them describe something independently and the fifth describes the opposite. That is a finding with a shape: the pattern, and the person who does not see it. We report both. An average would delete the second half, and the second half is usually the interesting one.

Their answers never reach you word for word. Not as a quote, not as an anonymous quote, not as a paraphrase close enough to identify who said it.

That rule is not a courtesy. It is the reason a person is willing to be honest about someone they care about, and the honesty is the raw material. A product that leaked one sentence would get truthful answers exactly once.

A part of the picture that nobody said anything about stays empty. We do not fill it in from the pattern, and we do not fill it in from your own answers wearing a different grammar. An empty section is an honest section and this product would rather ship one than invent the alternative.

If the whole set does not clear the bar, there is no charge. That is a rule in the code rather than a policy in a document. One check decides whether a Portrait may be released and whether it may be charged for, so the two cannot disagree. The evidence gate is the page about what that bar actually is.

It is not clinical. It produces no diagnosis, no condition, no score against a disorder, and nothing about health. It is not an employment assessment and must not be used as one. It does not predict how a relationship will go or how somebody will perform at work. We have no outcome data, and inventing a forecast from a personality read is the exact thing this category gets wrong.

It also does not tell you your percentile. Our comparison base is small, so a percentile would be a number with a decimal point and no meaning behind it. The day that base is large enough, we will say so and publish the number that made it true.