360 feedback frequency: how often to actually run one
Monthly is too noisy, annual is too slow to course-correct. How often to run a 360 for actual development — and how to match the cadence to your goal.
On this page
Frequency is the most under-thought variable in 360 feedback. Everyone argues about the questions and the anonymity and the rating scale, and then defaults, without much thought, to once a year — because that's when the company runs its review, and the 360 got bolted onto it. But annual is a cadence chosen for the fiscal calendar, not for how people actually change, and it quietly guarantees the 360 won't work as a development tool. By the time the results land, the moments they describe are months cold, and you get to sit with the feedback for a full year before anyone checks whether you did anything about it. Get the frequency wrong and the best questions in the world won't save the round. Get it right and a mediocre questionnaire still moves you.
Why "annual" is the default, and why it's wrong for development
Annual reviews exist to solve big-company administrative problems: syncing everyone to a comp cycle, producing a record, calibrating across managers. A 360 attached to that machinery inherits its once-a-year rhythm by default, not by design. That's fine for the administrative job. It's close to useless for the developmental one, for two structural reasons.
First, one data point isn't a trajectory. The entire value of a 360 for your own growth is the change between rounds — is the thing you're working on actually shifting, or does it just feel like it? A single annual snapshot can't tell you that. You need at least two points, close enough together to connect. Annual gives you your first comparison a full year in, which is far too late to course-correct.
Second, there's no feedback loop. Development runs on a loop: see how you land, change one thing, measure again. At annual cadence that loop takes a year to close once — so you effectively get one attempt per year at improving anything. That's the same reason continuous feedback beats the annual review: the loop, not the ceremony, is where growth happens.
The two ways to get frequency wrong
Run it too often
Best for: survey fatigue and noise
Monthly rounds exhaust the people answering — quality craters as they start clicking through — and they measure change that hasn't happened yet. Behavior doesn't shift month to month, so you're reading noise as signal and burning goodwill you'll need later.
Run it too rarely
Best for: a snapshot you can't act on
Annual (or "when we remember") means the moments are cold by the time you read them, you get no second data point to see a trend, and the improvement loop closes once a year. It documents; it doesn't develop.
The right cadence lives between these, and it's set by a real constraint: behavior changes slowly. Work from the NeuroLeadership Institute on how behavior change actually happens is consistent on this — new habits take sustained, repeated practice to stick, on the order of months, not weeks. Measure faster than the thing changes and you're just adding noise and fatigue. So your interval should be long enough that a real change could plausibly have occurred, and short enough that you can still course-correct. For most individual leaders, that lands on quarterly.
What each cadence is actually good for
Frequency isn't one-size-fits-all — it depends on what you're using the round for.
- Monthly — almost never right for a full 360. Behavior doesn't change this fast, and the fatigue destroys answer quality. The exception is a single, tiny pulse question during an intense, deliberate change push ("this month, did I interrupt less in meetings?") — one question, not a round.
- Quarterly — the default for active development. Long enough for a behavior to plausibly shift, short enough to adjust course three more times this year, and it maps naturally onto how work is already organized. If you don't have a reason to pick something else, pick this. It's why the quarterly structure works so well as a rhythm.
- Twice a year — a reasonable step down once a skill is stable and you're in maintenance rather than active change. Enough to catch drift, light enough to sustain indefinitely.
- Annual — fine as a broad backstop (a wide "how am I doing overall" once a year), but never as your only cadence if you're actually trying to change something. On its own, it's a snapshot, not a development tool.
Notice that the useful cadences are all tied to the pace of change and the pace of work — not to December.
How to choose your cadence
Start from what you're trying to change
If you're actively working on a specific behavior, run quarterly so you can see whether it's moving and adjust. If you're just monitoring a skill that's already solid, twice a year is plenty.Respect respondent fatigue
Every round spends goodwill from the people answering. Ask too often and they start phoning it in, which corrupts the data. Quarterly is roughly the ceiling before quality drops for most small teams — go slower if your people are stretched.Match the interval to how fast the behavior moves
A communication habit might show change in a quarter; a deeper pattern like how you handle conflict may need two. Space your rounds so a real delta has time to appear — measuring faster than change happens just reads noise.Keep the questions stable across rounds
Whatever cadence you pick, ask the same core questions each time. The comparison only works if you're measuring the same thing — a changing questionnaire gives you fresh snapshots but no trajectory.
Consistency beats frequency
If you take one thing from this: a decent round run reliably every quarter beats an ambitious round run once and abandoned. The compounding value of a 360 comes entirely from the sequence — round two compared to round one, round three to round two — and a sequence requires you to actually keep going. Most people over-design the first round and never run a second, which is the one outcome guaranteed to produce zero development, because with one data point there's nothing to compare. Put concretely: a leader who runs a plain five-question round every quarter for a year ends up knowing far more about how they've actually changed than one who commissions an elaborate thirty-question 360, reads it once, and files it — the first has three comparisons and a visible trend, the second has a single snapshot and a good intention. Pick a cadence you can genuinely sustain, keep the questions steady, and let the deltas accumulate. That's also the backbone of a real feedback-driven development plan — the plan is only as good as the re-measurement behind it. Korn Ferry's research on 360 design lands in the same place: what separates 360s that change behavior from ones that don't is far more about the repeated loop than the instrument.
Where Mirorly fits
Mirorly is built around the sequence, not the one-off. You answer a set of behavioral questions about yourself, send the same ones to the people you work with, and read your view next to theirs — then run the identical round next quarter and watch the gaps move (or not). You can set a round to recur and get a reminder when it's time, so the cadence you chose actually holds instead of quietly lapsing after round one. The quarterly check-in template is built for exactly this rhythm. If you're still deciding what to ask, start with questions that surface useful answers and keep them stable from there.
Common questions
The one-line summary
Frequency is the variable that quietly decides whether a 360 develops you or just documents you: annual is a calendar choice that gives you one uncomparable snapshot and a year-long loop, monthly measures change that hasn't happened and exhausts your respondents, and quarterly — the same questions, run on a cadence you can actually sustain — is where the deltas accumulate and real change becomes visible.