Skip to main content
Back to blog

Tailoring WhatsApp Send Cadence Under Meta's Marketing-Messages-per-User Cap

Meta's marketing-messages-per-user cap throttles how many marketing template messages one recipient will receive from your business, and Meta does not publish the threshold. This operator playbook turns cadence from a fixed calendar into a per-recipient budget — cohorts, stage windows, tenant-owned frequency caps, quiet-hours pacing, and uncapped re-engagement lanes.

Orbit Editorial Team

Quick answer: Meta's marketing-messages-per-user cap makes one thing operationally true: every opted-in recipient now has a finite, unpublished marketing budget with your business, and a recipient who hits their ceiling stops receiving your marketing sends until their window resets or they re-engage. The cadence that survives that is not a fixed calendar — weekly blasts to the whole list treat a full-fan and a dormant contact as equally worthy of a slot. The cadence that works allocates sends per recipient: recent engagers get the campaign, lapsed contacts get a tighter cadence and fewer lifecycle stages, dormant names graduate to the uncapped lanes instead of more templates. On Devotel Orbit the allocation is enforced where it should be — tenant-owned frequency caps per contact, quiet-hours release windows that spread a campaign across recipient-local days, segment exclusion so lapsed cohorts never enter the campaign at full volume, and the utility and inbound lanes that never touch the cap at all.

The cap explainer covers what Meta counts, what stays uncapped, and why the unpublished threshold cannot be reverse-engineered into a quota. This post is the other half: the operating cadence. It assumes the mechanics and goes straight to the levers — six of them, in the order an operator pulls them.

The mental model: a budget, not a calendar

A marketing calendar allocates sends per campaign. The cap allocates tolerance per recipient. The conflict between the two shows up in a specific place: a campaign that fires on your schedule regardless of what the recipient has already absorbed from you this week. Under the cap, the recipient's window is the scarce resource, and every send spends it — the promotion that converts and the lifecycle nudge nobody opens spend it identically from Meta's side, because Meta counts delivery attempts to a recipient, not your offers' conversion rates.

So plan the way the recipient experiences you. Rank your sends by expected value — the offers and moments a recipient would miss — and spend the budget there. The recipients who engage reset their standing through reads and replies; the recipients who never open get a tightening ceiling and give you nothing back for the spend. Daily-to-everyone is the calendar a cap punishes first, because it burns the most budget on the cohort least able to extend it.

One guardrail before the levers: Meta does not publish the threshold, so no cadence here claims to sit "under" it. What each lever does is move cap risk onto your side of the line — as your own throttled or suppressed sends — where you can see it, tune it, and change it before Meta starts failing deliveries on their side.

Lever 1 — Segment by engagement, then step the cadence down

Full-fan campaigns protect themselves; lapsed and dormant recipients do not. A three-band cohort structure is the working minimum:

  • Full-fan — engaged recently (opened, replied, or messaged you inside your engagement window). This cohort earns the campaign cadence: your normal frequency, your best templates, first allocation of any new send.
  • Lapsed — no engagement in the recent window but not dormant. This cohort gets a stepped-down cadence — fewer lifecycle stages, no daily promotions, one win-back attempt before demotion.
  • Dormant — nothing for a long window. This cohort graduates out of template cadence entirely and re-enters through the uncapped lanes, covered in lever 6.

On Orbit the cohorts live under Audience → Segments, built on last-engagement traits, and each band maps to its own cadence rules. Outbound → Audiences — the outbound hub's lists-and-segments view — then includes the full-fan band at full volume and excludes lapsed and dormant names from the high-frequency campaigns — the campaign never sees the recipient whose budget you intend to conserve. Exclusion at the audience layer beats throttling after the fact: a send that was never scheduled spent nothing, cost nothing, and needs no recovery.

Cohort membership has to move. A lapsed contact who opens or replies promotes back to full-fan; a full-fan contact silent through your window demotes. Static cohorts rot into exactly the shape the cap punishes — high volume to a band that quietly went dormant inside it.

Lever 2 — Interrupt lifecycle series on engagement

A lifecycle series — welcome, post-purchase, win-back — is a pre-committed cadence. Left alone it fires its full stage list at every enrolled contact regardless of what the recipient does between stages, which means it spends budget at the contacts the series was designed for and the contacts the series has already lost, indiscriminately.

The fix is interrupt-driven stages: what happens to the recipient decides the next stage, not the clock. A welcome series keeps its second and third sends for contacts who opened the first, and steps unopened contacts down to a single reminder before exiting. A win-back series caps at one or two attempts against a recipient who has not opened either — the loop that re-fires an unopened template every 48 hours spends the most budget on the cohort Meta tightens fastest, and the cap explainer's failure-mode walkthrough shows what that loop looks like from the recipient's side.

The same interrupt applies to transaction-adjacent sends: a recipient whose reply opened a service window does not need the promotional nudge your calendar had queued for them two days later. They are already in a free, uncapped conversation — the template send is redundant spend, and on a cap-tight recipient it is the send that fails. Design the series to release on engagement, not to persist through silence.

Lever 3 — Set tenant-owned caps, per contact

Segmentation is a coarse control — it steps the cadence at the cohort boundary. The fine control is a per-contact ceiling you own. Under the frequency caps settings you configure how many marketing sends one contact receives per day and per week, per channel; Orbit evaluates the contact's recent send count at send time and suppresses what would exceed the limit.

Set the per-contact ceiling where your engaged recipients tolerate volume — observed from your own delivery and read data, not from a number Meta has never published. Two consequences follow. First, the cap risk surfaces as your own suppression events in the send log, per contact, before it can surface as Meta-side delivery failures — a suppressed send is a free lesson; a per-recipient failure on Meta's side costs a spend and teaches you the same thing later. Second, the contact-level view turns "approaching the cap" from an invisible Meta property into something an operator reads on the contact profile, which is what makes cadence tuning an evidence loop instead of a guess.

Tenant caps and cohort segmentation compound: cohorts decide who a campaign targets, tenant caps decide how much any one contact in the target set can absorb. Neither claims compliance with Meta's threshold — both make the recipient's budget something you budget, not something you discover exhausted.

Lever 4 — Pace release through quiet hours

Two campaigns with identical volume behave differently if one releases inside one recipient-local evening and the other spreads across a week of sendable windows. The cap's counter is recipient intake inside a rolling window, so spreading a campaign across days spends less of each recipient's ceiling per day than concentrating it.

Quiet hours are the pacing mechanism: outbound sends hold during each recipient's local off-hours and release inside the sendable window, as the send-time and quiet-hours post works through — including the DST failure mode that punishes any send pinned to UTC instead of recipient-local time. For a marketing program the same windows double as a pacing dial: narrower daily sendable windows stretch a campaign across more days; off-peak hours concentrate volume where attention, and therefore engagement resets, are likelier.

The secondary effect is quality-side protection. The sends you hold to a recipient's waking, willing hours are the sends a recipient is least likely to report — and reports are what drive the quality-rating and template-pause gates that run alongside the cap. Pacing spends less per-user budget and feeds the feedback gates less at the same time.

Lever 5 — Keep categories honest

Meta counts what it classifies as marketing — its classification, not your intent. A utility template carrying promotional copy gets re-classified, and then it counts toward the recipient's cap, bills at the marketing rate, and feeds the quality gates as marketing: three treatments for one send. The category discipline is a budget discipline.

The split is mechanical. Order updates, appointment reminders, and account alerts go out as clean utility templates — uncapped per recipient, so they never spend the budget your promotions need. Offers go out as correctly-tagged marketing templates, throttled by levers 1 through 4. Promotional copy does not ride along inside the utility send, because the re-classification converts an uncapped send into budget spend at the worst possible time: mid-lifecycle, attached to a transaction the recipient actually wanted.

Under Channels → WhatsApp → Templates each template's category is part of authoring and submission, which makes category honesty enforceable at template review rather than discovered at delivery. The per-message pricing explainer prices the same classification — the category you declare is the category you pay at and the category the cap counts.

Lever 6 — Graduate dormant contacts to the uncapped lanes

The cap's boundaries define the recovery lanes, and they are wide:

  • Utility and authentication sends deliver without spending the marketing budget.
  • Service-window replies — anything inside the 24 hours after a customer messages you — are neither capped nor charged.
  • Free entry points — a conversation the customer starts from a Click-to-WhatsApp ad or a Facebook Page call-to-action — open a free window, as the entry-point post details.

A dormant cohort does not respond to more marketing templates; that is what dormant means, and it is also the cohort Meta tightens the ceiling on first. So the graduation path runs off the capped lane entirely: re-acquire dormant names through entry points they choose, and let the volitional inbound message open the service window where your follow-up travels free. A dormant contact who clicks a Click-to-WhatsApp ad has told you more than the last three unopened templates combined — and their reply resets their standing with you on Meta's side, which the templates never could.

The reframing worth taking: under a per-recipient cap, the uncapped lanes are not a loophole, they are the majority of a durable program. Marketing templates become the minority spend — the high-value moments — and conversations carry the rest.

The cadence worksheet

A concrete picture ties the levers together. A weekly program under the cap, as an operator would set it:

  • Full-fan cohort. Up to the tenant cap of two marketing sends per week per contact. Campaign release paced through quiet hours across three recipient-local evenings, not one. Lifecycle series enrolled, but every stage interrupts on engagement — an opened stage advances the series, a reply releases the contact into a service window and out of the template queue.
  • Lapsed cohort. One marketing send per week, capped by segment. A single win-back stage rather than a series — one attempt, then demotion on silence.
  • Dormant cohort. Zero marketing templates. Re-entry through a monthly Click-to-WhatsApp acquisition burst and any utility sends their transactions generate.
  • All cohorts. Utility and authentication sends run independently of the caps — they never spend the budget. Frequency-cap suppressions reviewed weekly on the send log; a rising suppression count in the full-fan cohort means the cadence outran tolerance, and the levers move before Meta does.

The rhythm is a budget review, not a send schedule: what did each cohort absorb, what did suppression save, what did engagement reset, and where does next week's spend go.

Frequently asked questions

How many marketing messages can I send one WhatsApp user per day?

Meta does not publish the threshold, and the recipient's ceiling varies with their engagement — chronically disengaged recipients get a tighter one. There is no number to plan against, and any vendor quoting one is guessing. The workable posture is your own per-contact frequency cap, set below what your engaged recipients tolerate, so over-cadence sends suppress on your side of the line instead of failing on Meta's.

What counts as a marketing message under the cap?

What Meta classifies as a marketing-category template — its classification, which is also the classification that prices the send. A nominally utility template with promotional copy gets re-classified as marketing, and then it counts toward the cap alongside the explicitly marketing sends. Utility and authentication templates as Meta classifies them, inbound service-window replies, and free entry-point conversations are outside the cap.

What happens when a user hits their cap?

Further marketing template sends from your business to that recipient fail at delivery instead of landing, until the recipient's window resets or the recipient engages — reads, replies, and inbound messages restore their standing. Nothing is wrong with your number or your template; that one recipient is closed to your marketing until then. It is a per-recipient limit, independent of the per-number quality rating and per-template pause gates.

How is Meta's per-user cap different from quality rating and template pauses?

Three separate counters. The quality rating grades your number on recipient feedback — blocks and reports. A template pause removes one template from your sendable set on per-template feedback. The per-user cap limits marketing deliveries per recipient regardless of both: a green-rated number sending an unpaused template still fails to a recipient whose ceiling is spent. All three punish the same underlying pattern — marketing volume recipients did not want — but they measure different things and reset on different events.

Should I just split big campaigns across more recipients instead?

More recipients dilute per-recipient spend, but the recipients absorbing your highest cadence are the same engaged minority in every campaign, and the cap is exactly per-recipient. Splitting volume fixes a per-account problem that is not the problem. The fix is allocating cadence by engagement so your best budget goes to recipients who extend it — and moving the cohorts that do not engage onto the uncapped utility, service-window, and entry-point lanes.

Resources

Published 11 October 2026.

Keep reading

Tailoring WhatsApp Send Cadence Under Meta's Marketing-Messages-per-User Cap — Orbit by Devotel