Frequency Cap¶

Delivery-frequency rule — instantiates Saturation Avoidance

Limits how often the same input reaches the same recipient within a window, spacing exposures so the channel keeps responding instead of going numb.

Some channels saturate not from total volume but from repetition — the fifth reminder, the tenth ping, the same message hitting the same person again and again until they stop responding to any of it. Frequency Cap is the rule that bounds repeat exposure per recipient per window: at most N deliveries to one target in a given period, with enforced spacing between them. Its defining move — the thing that makes it this mechanism and not a general admission limit — is that it counts per unique recipient over time, treating each person or subsystem as a small channel with its own recovery budget, rather than governing the aggregate queue. The gap it inserts between exposures is not incidental; the recovery interval is the point, because a saturated receptor recovers with rest, not with more input.^[1]

Example¶

A product team's lifecycle-messaging system can send onboarding tips, feature nudges, billing notices, and re-engagement pushes — all from different internal owners who each think their message is the important one. Left alone, a new user gets nine notifications in their first week and mutes the app. A Frequency Cap sets a per-user budget: no more than three non-critical pushes in any seven-day window, and no two within twenty-four hours. When onboarding, growth, and billing all queue a message on the same morning, the cap admits one, spaces the next, and drops the third into the following window.

The measured effect is counter-intuitive to the senders: total sends fall by roughly a third, but tap-through per send rises and unsubscribes drop — because the recipient's attention channel was allowed to recover between exposures. The cap didn't just reduce noise; it protected the very responsiveness each sender was competing for.

How it works¶

Count per recipient, not per campaign. The cap is enforced against a unique target's rolling exposure history, so many senders share one budget instead of each spending it independently.
Enforce a minimum gap. Beyond a ceiling-per-window, require spacing between consecutive exposures — the recovery interval that keeps the receptor from staying numb.
Reserve the budget for what matters. Critical or transactional messages are exempted or carve out their own allowance, so the cap suppresses the low-value repeats first.
Roll the window. Track exposures on a sliding window rather than a calendar reset, so a burst can't cluster at a boundary.

Tuning parameters¶

Cap height and window — exposures allowed per period. Lower protects responsiveness and starves legitimate reach; higher reaches more and risks numbing the recipient.
Minimum inter-exposure gap — the enforced recovery interval. Longer gaps preserve response but slow any campaign that needs cadence.
Counting unit — per person, per device, per household, per account. The wrong unit (per channel, per campaign) lets the recipient saturate while each sender's books look clean.
Exemption set — which message classes bypass the cap. A generous exemption list quietly defeats the cap; a strict one may block something genuinely urgent.

When it helps, and when it misleads¶

Its strength is protecting a shared receptor from many independent senders who would each, rationally, over-send. By pricing repeat exposure against a common budget with built-in recovery, it keeps the channel responsive and usually lifts per-exposure effect even as raw volume falls.

Its classic failure is capping the wrong unit: a per-campaign or per-channel cap looks disciplined while the actual human still gets buried, because their saturation is cross-sender and cross-channel. It also can't tell a valuable exposure from a wasted one — it only rations frequency, so a well-tuned cap paired with junk content still delivers spaced junk. And exemptions are where discipline leaks: enough "critical" carve-outs and the cap is decorative. The guard is to cap at the unit that actually saturates (the recipient, across senders) and to audit the exemption list as ruthlessly as the cap itself.

How it implements the components¶

Frequency Cap fills the archetype's per-recipient input-rationing components:

input_reduction_rule — it is a rule for slowing input: at most N exposures per window, excess deferred or dropped.
recovery_buffer — the enforced minimum gap between exposures gives the receptor recovery time so it does not live permanently saturated.
saturation_budget — the per-recipient exposure allowance is a standing budget many senders draw against.

It does not decide which competing message wins a scarce slot by value or urgency — that ranking is Queue Admission Limit's priority_admission_rule — and it neither detects saturation (Saturation Dashboard) nor reroutes overflow (Alternate Pathway Routing).

Instantiates: Saturation Avoidance — rations repeat input to keep a per-recipient channel responsive.
Sibling mechanisms: Ad Frequency Cap · Attention Cap Management · Queue Admission Limit · Saturation Dashboard · Channel Capacity Management · Alternate Pathway Routing · Worker Caseload Limit · Overflow Queue · Capacity Expansion Trigger · Graceful Degradation Mode

Notes¶

Frequency Cap is the general, cross-domain rule; Ad Frequency Cap is its advertising-specific instance, tuned to impressions and creative wear-out. Where both could apply, use the general cap for the shared-receptor budget and reserve the ad variant for the campaign-lift specifics — they are not two competing tools so much as one tool at two altitudes.

References¶

[1] Habituation — the progressive decline in response to a repeated, unchanging stimulus — is the mechanism a frequency cap fights. Spacing exposures (and varying them) is the standard countermeasure, which is why the enforced recovery gap matters as much as the ceiling. ↩