Setup Considerations: 360 cycle design

What this achieves

A 360 cycle is expensive in other people’s time, and most of what determines whether it produces anything useful is settled before it launches. This page covers those decisions, and which of them you cannot revisit afterwards.

Decision 1: what the cycle is for

The cycle type is not a formality — it decides who is asked and what the answers can be used for.

Option	Description
360 Review	Multi-rater feedback from peers, managers, and direct reports. The widest picture and the most time spent.
Peer Feedback	Feedback from selected peers only.
Upward Feedback	Direct reports give feedback on their manager.
Manager Review	Managers evaluate their direct reports.
Self Assessment	Employees assess themselves.

Run development-focused feedback and rating-focused review as different cycles, or accept that reviewers write for the more consequential of the two. A peer told their words may affect somebody’s pay writes differently from one told this is for development, and only one of those is what you asked for.

Example: HC Corp running an upward-feedback cycle on its engineering managers is asking direct reports to comment on the person who writes their review. That works only where anonymity is genuine and the reviewer pool is big enough to make it so.

Decision 2: the peer minimum is the anonymity control

Min Peer Reviewers and Max Peer Reviewers look like sample-size settings. On an anonymous cycle, the minimum is the setting that decides whether anonymity is real.

Peers per subject	What anonymity means
Two	A comment is attributable by anyone who can count.
Four or five	A genuine pool, and a workable sample.
Eight or more	Strong anonymity, and a lot of other people’s time.

Example: an anonymous HC Corp cycle with a minimum of two peer reviewers displays the word anonymous and delivers none of it. With a minimum of four, both the screen and the reality say the same thing.

The maximum matters for a different reason. Every reviewer assignment is an hour of somebody’s week, and a subject with twelve reviewers costs twelve people that hour.

Decision 3: nominated or assigned peers

Allow Peer Nomination lets the subject pick their own peers within the range.

Option	Description
On	Reviewers who genuinely worked with the subject. Also a friendlier sample, chosen by the person being reviewed.
Off	Reviewers assigned by the cycle. Less flattering, and sometimes less informed.

Neither is wrong. Nomination suits a development cycle where the point is useful input. Assignment suits a cycle whose results feed a rating, where the subject choosing their own judges is a problem.

Decision 4: cohort size and who can be identified

Scope decides how many people are reviewed, and how identifiable the reporting is afterwards.

Example: a cycle scoped to HC Corp’s Engineering department covers 34 people, and a departmental figure says something about a group. Scoped to the People department it covers 6, and every breakdown of the results — by department, by relationship, by anything — is a breakdown of six identifiable colleagues.

Decide before launch what you will report at what size, and apply it evenly. The bias report and the analytics page both break results down by department, so this is not hypothetical.

Decision 5: one template or several

Templates are reusable, and the same library serves performance review cycles. That has two consequences worth planning around.

Editing a template affects whatever picks it up next. Duplicate rather than edit once a template is in live use.
Changing the question set between cycles means two cycles that cannot be compared. If you intend to track movement over time, the question set has to hold still.

Example: HC Corp asking a different set of competency questions each year gets a rating each year and a trend never. Holding the template steady for three cycles gets a trend, which is the thing worth having.

Decision 6: how much time reviewers actually have

A cycle’s window is the period reviewers have to write. Under-running it produces rushed answers; over-running it produces answers submitted on the last two days regardless.

Plan the chasing too. Reminders exist and work, and a cycle that reminds weekly for a month trains people to ignore reminders. Two well-timed nudges beat six.

What is expensive to change later

Decision	Cost of changing later
Anonymity	Cannot be changed once the cycle launches. Reviewers were told one thing on the form.
Target scope	Cannot be changed. The interface locks it at creation — cancel and recreate, losing everything submitted.
Cycle type	Cannot be changed. It determines who was assigned.
The peer minimum, before launch	Low. It is an editable field on a running cycle, but raising it after reviewers are assigned does not retrofit anonymity to responses already written.
The template’s question set, between cycles	High. Two cycles with different questions are two cycles, not a trend.
Dates	Low. Start and end dates stay editable on a running cycle.

What happens next

Build the template first, because it is the reusable part and the cycle is the disposable one. Then create the cycle, settle the switches above, and launch it — which is the point at which participants and reviewers are created and the scope stops being editable.