Free resource · use it on this page
Notification Frequency Planner
A per-segment worksheet for setting frequency caps that protect opt-in and engagement without muting your program.
How to use this planner
Work through this page top to bottom. Step 1 captures your program baseline, Step 2 classifies every message you send into a tier, Step 3 is the worksheet you complete once per audience segment, Step 4 is the guardrail checklist you verify before the caps go live, and Step 5 is the review loop that keeps them honest. Print the page or copy the prompts into your own document; by the end you will have a complete cap sheet you can hand to whoever operates your sends.
The reason caps deserve a worksheet: the cost of one more message is invisible on any single send and compounding across a subscription. Opt-out is mostly a one-way door. When someone revokes notification permission, unsubscribes, or deletes a pass, you rarely get that permission back. A frequency cap is how you spend a finite attention budget deliberately instead of letting whichever team scheduled last spend it for you.
One rule holds everything together: caps are counted per recipient, across all channels and all campaigns combined. A cap that only counts one channel, or only one team's calendar, is not a cap.
Step 1 — Program baseline
Fill this in once for the whole program before setting any per-segment numbers. Every answer should come from logs or settings you can point at, not from what the calendar intended.
Channels in scope
List every channel that can interrupt this audience: app push, SMS, email, in-app messages, wallet lock-screen updates. Caps that ignore a channel undercount real interruptions.
Actual volume today
Messages received per subscriber per week right now, across all campaigns and channels. Pull it from send logs. If different teams send independently, sum them; the recipient does.
Exit metrics you can measure
Name each way someone leaves and where the number lives: unsubscribe, notification permission revoked, app uninstalled, pass deleted. If you cannot measure it, note that gap now.
Quiet hours
The local-time window when nothing promotional sends, for example 21:00 to 09:00. Decide whether it applies in each recipient's time zone or one fixed zone, and write which.
Transactional exemptions
Message types that bypass caps: receipts, delivery updates, security alerts. Keep the list short and literal. Every exemption is a loophole someone will eventually market through.
Global weekly ceiling
The absolute maximum interruptions one recipient can get per week, all channels and segments combined. Set this first so per-segment caps have something to sum against.
Cap owner and change log
One named person approves cap changes. One shared document records every change with date, old value, new value, and reason. Caps nobody owns drift upward.
Step 2 — Classify every message into a tier
Go through your active and planned campaigns and assign each one a tier. The tier decides how a message counts against caps and which message wins when a cap forces a choice. Cadences listed are starting points to test against your own exit metrics, not targets.
| Tier | What belongs here | Starting cadence to test | How it counts against caps |
|---|---|---|---|
| Transactional | Receipts, order and appointment confirmations, security alerts | Triggered by the recipient's own action; no calendar | Exempt from caps, but still counted in the totals you monitor |
| High-value alerts | Expiring rewards, booked-appointment reminders, restock alerts the recipient explicitly requested | As triggered; typically a few per month | Counts against caps and wins arbitration over promotional sends |
| Promotional | Offers, sales, launches, seasonal pushes | Around one per recipient per week; adjust from evidence | Counts fully; first tier to be dropped when a cap is reached |
| Lifecycle and educational | Onboarding sequences, usage tips, milestone messages | Front-loaded in the first week or two, then tapering | Counts fully; pause promotional sends to a recipient who is mid-onboarding |
| Re-engagement | Win-back series for recipients inactive past your threshold | A short, finite series with weeks between messages | Counts fully and hard-stops when the series ends; no indefinite retries |
| Passive updates | Content that changes without an alert, such as a points balance or current offer displayed on a wallet pass | Continuous | Does not count against interruption caps because nothing buzzes; wallet platforms such as PushNotice separate silent pass updates from lock-screen pushes, and only the pushes belong under your caps |
Step 3 — Per-segment cap worksheet
Copy this block once for every segment you message differently. If two segments would get identical answers, merge them; a segment that does not change a sending decision is not a segment.
Segment name and definition
A name plus the rule that builds it, precise enough that two people would put the same person in the same segment. Example: lapsing = no visit or open in 30 to 90 days.
Relationship stage
New, active, lapsing, or dormant. Stage drives tolerance: new subscribers are forming expectations, dormant ones are one annoyance from leaving. Cap accordingly.
Weekly cap
Maximum interrupting messages this segment receives per recipient per week, all channels combined. Verify that your segment caps summed for one person never exceed the global ceiling from Step 1.
Monthly cap
A separate monthly maximum stops a legal weekly pace from compounding into a heavy month. A recipient at four quiet weeks and one loud one experiences the loud one.
Minimum gap between messages
The shortest allowed interval between two interruptions to the same recipient, for example no two promotional messages within 48 hours, regardless of channel.
Tier priority for this segment
When the cap leaves room for one message and two are queued, which tier wins. Write the order from Step 2 so the sending system, not send time, decides.
Cool-down triggers
Events that pause promotional sends to a recipient: a purchase or redemption in the last few days, a support complaint, an ignored streak of consecutive sends.
Segment transition rule
When a recipient moves between segments, which cap applies and when the counter resets. Decide it here, or every transition becomes an accidental loophole.
Hold-out cell
Reserve a slice of this segment at a deliberately different frequency, higher or lower. It is the only way to learn what your cap is costing or saving you.
Step 4 — Guardrails to verify before the caps go live
Check every item. The first four are about enforcement; the middle four are the fatigue signals to watch; the last four are the habits that keep the system trustworthy.
Caps are enforced by the sending system, not a calendar
A spreadsheet agreement fails the first time two teams schedule the same Tuesday. The tool must refuse the send.
Caps count per recipient across all channels
Per-campaign or per-channel counting lets one person absorb every team's full allowance at once.
Cross-channel dedupe is in place
The same offer should not arrive by push, SMS, and email on the same day. One message, one best channel.
Arbitration follows written tier priority
When messages compete for the last slot, the Step 3 priority order decides, not whichever job ran first.
Quiet hours verified in recipient local time
Send a test to accounts in your two most distant time zones and confirm delivery timestamps before trusting the setting.
Opt-outs tracked by send day
A spike in unsubscribes or permission revocations aligned to a specific send is the clearest fatigue signal you will get.
Engagement decay watched within segments
Tap or open rate falling across consecutive sends to the same segment is fatigue arriving before the opt-out does.
Silent churn monitored
Permission revocations, uninstalls, and pass deletions are people who left without unsubscribing. Count them with the loud exits.
New subscribers get their own cap
The first two weeks set expectations for the whole relationship. Onboarding volume should displace promotions, not stack on them.
Transactional exemptions audited on a schedule
Marketing dressed as a transactional message erodes trust faster than any cadence mistake. Reread the exempt list quarterly.
One variable changes at a time
If cap, timing, and content all change together, the exit metrics cannot tell you which one moved them.
Overrides are possible but logged
Name who can break the cap for a genuine emergency, and make every override leave a record with a reason.
Step 5 — The monthly review loop
Caps set once and never revisited become folklore. Run this loop monthly; it takes under an hour once the baseline exists.
- 1
Pull actuals from logs
Messages received per recipient per segment over the last 30 days, from send logs, not planning documents. The gap between planned and actual is the first finding.
- 2
Compare actuals against caps
Flag segments running over their cap, which is an enforcement failure, and segments far under it, which may be a missed opportunity or a dead segment.
- 3
Read the exit metrics per segment
Opt-outs, permission revocations, uninstalls, and pass deletions by segment, aligned against send days. Note which sends the exits cluster after.
- 4
Adjust one cap by one notch
Raise or lower a single segment's cap and write the hypothesis: what signal you expect to move, in which direction, by roughly when.
- 5
Record the change in the log
Date, segment, old value, new value, reason, expected signal. Future you, or your successor, should be able to reconstruct why every cap is what it is.
- 6
Close the loop next cycle
Check the hypothesis against the hold-out cell. If exits worsened, revert and note it; a reverted change with a documented reason is a finding, not a failure.
From the guide: Best Push Notification Tools (2026)
This resource accompanies the full article — worth reading before you commit to a tool.