Most push programs don’t fail because the copy is bad. They fail because teams optimize the wrong thing first.
In mid market orgs, a Growth or Performance Marketer usually inherits a push setup that is already “working” in the sense that messages go out. But when you try to prove impact, you hit familiar friction. Attribution is fuzzy, segments are too broad, reporting is fragmented, and experiments take too long to launch. That is why application push notification performance needs a tight metric stack. One that quickly tells you whether you have a delivery problem, a relevance problem, or a conversion problem.
Push is also one of the few channels where tiny changes compound fast. A one point lift in CTR can mean thousands of extra sessions. A one point lift in conversion rate can change CAC payback. The trick is knowing which metric to look at first. Then connecting it to an experiment you can run this week.
The practical order of operations: fix reach, then relevance, then revenue
When a dashboard is noisy, it helps to read it like a funnel.
Start with whether notifications arrived and were seen. Then check if users were motivated enough to open and click. Only after that should you judge whether the experience behind the tap converts and makes money.
In practice, this order prevents wasted work. If delivery is unstable, A/B testing copy is mostly pointless. If view rate is low, personalization won’t save you. If CTR is healthy but conversion rate is weak, the issue is usually deep link destination, landing friction, or offer mismatch.
See real-time delivery and failure diagnostics in action with SashiDo - Push Notification Platform.
Application push notification metrics you should track as a system (not a list)
The fastest teams track these metrics together, because they explain each other.
A simple example you can spot in the wild. A campaign has strong open rate, but CTR is flat. That typically means users are curious enough to expand or tap the notification preview, but the value proposition does not justify the next step. Another pattern. CTR improves, but revenue per notification drops. That usually means targeting got broader, or you optimized for clicks rather than high intent users.
That’s why the goal is not “better metrics”. The goal is clear diagnosis. You want each campaign to answer one question. Did we reach the user. Did we earn attention. Did we drive the intended action. Did that action produce value.
Below are the ten metrics that consistently give that diagnosis. Use the formulas exactly, and treat the benchmarks as context, not a grade.
1) Push Notification Click-Through Rate (CTR)
CTR is the clearest signal that your message and targeting created intent. In performance terms, CTR is your ad like relevance score. If it’s low, you either sent the wrong message, to the wrong segment, at the wrong time.
When CTR is low across the board, marketers often overreact by adding urgency or discounts. The faster fix is usually segmentation. For example, split “active in last 7 days” from “active in last 30 days”. A reactivation message that works on warm users can feel irrelevant to cold users, and that mismatch drags CTR down.
Formula
CTR = (Number of clicks / Number of push notifications delivered) * 100
Example math stays simple on purpose. If you delivered 100,000 notifications and got 3,000 clicks, CTR is 3%.
For benchmark context, some industry rollups put average CTR around 2.25% across verticals, which is useful as a sanity check when you are trying to detect tracking bugs versus true underperformance.
External source: CleverTap benchmark report for average CTR (2.25%).
https://clevertap.com/blog/push-notification-report/
2) Push Notification Conversion Rate
CTR tells you the notification was compelling. Conversion rate tells you whether the tap was worth anything.
Conversion rate is where attribution arguments usually start. A push click is not a conversion. A push click is a visit. In most apps and subscription products, conversions depend on what happens after the tap. That makes conversion rate the metric that forces alignment between marketing and product.
A real pattern you will see. CTR looks fine. Conversion rate is weak. That is usually a deep link and destination issue. The notification promises one thing, and the screen after the tap shows another, or it forces login, or it loads slowly. Fixing that often beats rewriting the message.
Formula
Conversion Rate = (Number of users who completed the desired action / Number of users who clicked the notification) * 100
For a more holistic view, you can also measure:
Overall conversion rate = (Number of conversions / Number of notifications delivered) * 100
Use the click based conversion rate to diagnose the post tap experience. Use overall conversion rate to compare campaigns end to end.
3) Push Notification Delivery Rate
Delivery rate is the part everyone assumes is fine. Until it isn’t.
A mid market team can lose 10 to 20% of reachable audience just from stale tokens, uninstalls, permission changes, or a misconfigured iOS push notification integration. When that happens, your test results get noisy because the segment you think you are sending to is not the segment that receives.
Delivery rate is also where platform complexity shows up. iOS and Android have different behaviors, and web adds another layer. If you run a javascript push notification backend for browsers, you can also see delivery issues tied to service workers, browser throttling, or users clearing site data.
Formula
Delivery Rate = (Number of push notifications delivered / Number of push notifications sent) * 100
If delivery rate drops, you usually investigate in this order. Token hygiene and invalid tokens, app uninstall handling, credential and certificate changes for APNs, throttling and quota behavior for FCM, then any provider side failures.
External sources you can use for technical grounding:
Apple APNs overview and payload guidelines.
Firebase Cloud Messaging delivery optimization docs.
https://firebase.google.com/docs/cloud-messaging/doc-revamp/optimize-delivery/understand-delivery
4) Push Notification Open Rate
Open rate is about winning the first second. The lock screen moment.
This is the metric most affected by timing and preview relevance. And it can move dramatically when you shift from generic blasts to contextual triggers.
You will often see a misleading interpretation here. People assume low open rate means bad copy. In practice, low open rate often means the user saw it at the wrong moment, or the preview did not match their current context.
Formula
Open Rate = (Number of notifications opened / Number of notifications delivered) * 100
For benchmark context, contextual push can outperform generic messaging by a wide margin. Batch’s benchmarks show contextual notifications with significantly higher open rates than generic ones.
External source: Batch CRM and push benchmark.
https://batch.com/ressources/etudes/benchmark-notifications-push-crm-mobile
5) Push Notification Opt-In Rate
Opt-in rate determines how big your push channel can get.
Performance marketers feel this as a ceiling. If opt-in is low, you can have a perfect CTR and still not move the business because your addressable audience is too small. Opt-in is also where user trust shows up. If users don’t understand why you want permission, they say no.
On iOS, opt-in is famously harder because the system prompt is explicit and permanent unless the user changes settings. That makes when you ask almost as important as how you ask. Asking on first launch often underperforms. Asking right after a user takes a meaningful action usually performs better because the value exchange is clearer.
Formula
Opt-In Rate = (Number of users who opted in / Total number of users prompted) * 100
For benchmark context, recent mobile benchmark reports frequently cite iOS opt-in around the mid 40% range across categories.
External source: OneSignal mobile app benchmarks.
https://onesignal.com/mobile-app-benchmarks-2024
6) Push Notification Revenue Rate (Revenue per Notification)
Revenue rate is the metric that ends debates.
If you can attribute revenue to a campaign, you can compare push to paid social, email, and in app messaging. You can also justify investment in richer segmentation, better experimentation, or a more capable customer communications platform.
A practical way to use this metric is to pair it with frequency. If you increase send volume and revenue per notification drops, you may be saturating low intent users. That is the moment to tighten targeting or cap frequency by segment.
Formula
Revenue Rate = Total revenue generated from push notification campaign / Number of notifications delivered
You can still break it down internally as revenue per engaged user and ROI, but keep this base number in your core dashboard because it is the easiest to compare across campaigns.
7) Push Notification Success Rate
Success rate is the metric you define per goal. That makes it powerful, but easy to misuse.
For lifecycle programs, success might be “completed onboarding step”. For content apps, it might be “watched a video”. For commerce, it might be “added to cart” rather than purchase. The key is to pick an action that is both meaningful and frequent enough to optimize.
Success rate also helps when your conversion event is delayed. If purchase happens days later, you can still optimize for nearer term success signals that correlate with purchase.
Formula
Success Rate = (Number of successful outcomes / Number of push notifications delivered) * 100
If you use success rate, document the success event clearly so teams don’t compare incompatible campaigns.
8) Push Notification Opt-Out Rate
Opt-out rate is the cost of being annoying.
Opt-outs are not just a messaging problem. They are also a segmentation and frequency problem. A small set of highly active users can tolerate more messages. Dormant users often cannot, because any notification feels like an interruption.
Watch opt-out rate by segment, not just overall. A rising opt-out rate in one segment is an early warning that your personalization is not landing there, or your frequency is miscalibrated.
Formula
Opt-Out Rate = (Number of users who opted out / Total number of users who received notifications) * 100
The trade off is real. Increasing frequency can lift short term clicks while accelerating long term opt-outs. Treat opt-out rate as a guardrail metric in your experiments, not an afterthought.
9) Average Time to Open a Push Notification
Time to open tells you whether your message creates urgency, and whether your timing matches user routines.
In many products, a push click is time sensitive. If users open hours later, the offer might have expired or the content is stale. That turns into low conversion rate even if open rate is decent.
Time to open is also how you discover that “best time to send” is not one time. Different segments have different rhythms.
Formula
Average Time to Open = Sum of time taken by all users to open notifications / Number of opened notifications
One of the most consistent performance levers here is tailoring send times to user behavior. Research summaries often cite sizable lifts in reaction rates when send time is personalized.
External source: Business of Apps push notification statistics (send time personalization lift).
10) Push Notification View Rate
View rate is your “had a chance to work” metric.
A notification can be delivered but never actually seen, especially on mobile where lock screen grouping and do not disturb modes can hide notifications. View rate helps you separate “low engagement because people didn’t see it” from “low engagement because they saw it and ignored it”.
If view rate is low, the fix is usually timing, relevance, and sometimes technical presentation. Rich formats can help, but only after you know users are actually seeing messages.
Formula
View Rate = (Number of notifications viewed / Number of notifications delivered) * 100
External source: PushEngage KPI overview that includes view rate framing.
https://www.pushengage.com/kpi-measure-effectiveness-push-notifications/
Turning metrics into decisions: what to fix when a number drops
Metrics are only useful when they change what you do next.
If delivery rate drops, treat it like an incident. Pause big sends until you understand whether the problem is credentials, token hygiene, platform throttling, or provider issues. Your marketing calendar is not worth corrupting your measurement.
If view rate drops but delivery rate stays stable, you are likely sending at the wrong time for that segment, or you are competing with too many other notifications on device. This is where local time zone sending and user level send time optimization matter.
If open rate drops, check preview relevance first. Your headline, first line, and icon context are doing the work. Generic copy tends to decay quickly as your audience grows.
If CTR drops while open rate holds, you have a message to value mismatch. Tighten the segment. Adjust the offer. Improve the deep link destination so it matches the promise.
If conversion rate drops while CTR holds, your post tap experience is leaking. That is landing screen friction, slow load, forced login, broken deep links, or an offer that disappears after tap.
If revenue per notification drops while conversions hold, your conversions may be lower value. That is where targeting high intent users and adding purchase value thresholds into segmentation tends to pay off.
Instrumentation that avoids the attribution trap
Performance marketers care about ROI, not vanity taps. But push attribution can get messy because users often don’t convert immediately.
A practical approach is to track two layers.
First, track interaction attribution. Every notification should have campaign ID, variant ID, send timestamp, and destination deep link parameters. This gives you clean CTR and open rate by test cell.
Second, track business attribution with a defined window. Decide what counts as influenced conversion, for example within 24 hours of click or within 72 hours of delivery. The exact window depends on your product cycle. Short for on demand actions. Longer for considered purchases.
This is where a platform that is built for engineers and marketers working together helps. With SashiDo - Push Notification Platform, teams typically want two things at once. Reliable delivery across iOS, Android, and web, plus diagnostics that tell you what happened when delivery or engagement drops, without building and maintaining new infrastructure.
If you are managing a stack that includes a digital experience platform or cross channel marketing platforms, push metrics become even more important because push often looks like the “cheap channel” that can be scaled endlessly. The numbers above are how you prevent that from turning into opt-outs and wasted impressions.
A 30/60/90 day metric-first experiment plan (minimal, but realistic)
The goal of this plan is to shorten your iteration cycle while keeping measurement clean.
First 30 days. Make the data trustworthy.
- Instrument delivered, viewed, opened, clicked, and converted events with campaign and variant IDs.
- Establish baselines for delivery rate, view rate, open rate, CTR, conversion rate, opt-out rate, and revenue rate.
- Create two core segments that you will reuse. “Active last 7 days” and “Dormant 30 plus days” is a common start because behaviors differ.
Next 60 days. Run tests that isolate one lever at a time.
- Timing tests. Same copy, different send time windows. Measure open rate and average time to open.
- Targeting tests. Same offer, narrower versus broader segments. Measure CTR and opt-out rate as a guardrail.
- Destination tests. Same notification, different deep link destinations or flows. Measure conversion rate.
By 90 days. Optimize for revenue and scale safely.
- Add value based segmentation. High intent users, high LTV users, or users with cart value above a threshold.
- Introduce frequency caps by segment. Use opt-out rate to set guardrails.
- Shift reporting from “campaign performance” to “program performance”. Revenue per notification by lifecycle stream.
This is also the point where teams often revisit tooling. If your current setup makes it hard to run A/B and segmentation fast, consider whether you need a more developer friendly customer communications platform.
Where platform choice shows up in your metrics
Most teams don’t change push providers because of features. They change because measurement, control, or deliverability becomes a blocker.
If you are planning a parse migration from OneSignal or you are comparing vendors like OneSignal, it helps to evaluate them through the metric lens. Can you diagnose delivery failures quickly. Can you segment precisely without exporting data to three places. Can you ship changes without weeks of engineering time.
If you do compare, keep it practical and specific. Here is a direct comparison page if OneSignal is in your shortlist.
https://www.sashido.io/en/sashido-vs-onesignal
The best practice layer: the three habits that lift every metric
Segmentation is the habit that lifts open rate, CTR, conversion rate, and opt-out rate all at once, because it reduces generic messaging. The easiest win is to segment by recency and behavior. Then grow into dynamic segments that update based on real time events.
Timing and frequency discipline is the habit that protects your addressable audience. When you see opt-outs climb, do not just send less. Send smarter. Cap frequency for low intent segments and earn the right to message more by sending fewer, higher value notifications.
Experiment design is the habit that prevents false wins. Keep tests focused. One variable at a time. And always use guardrails like opt-out rate and revenue per notification so you don’t optimize yourself into short term clicks and long term audience loss.
Conclusion: tie every application push notification metric to a next action
A strong push program reads like a control system. Delivery rate and view rate tell you whether you have reach. Open rate and CTR tell you whether you earned attention. Conversion rate, success rate, and revenue per notification tell you whether attention turned into business value. Opt-in and opt-out rates tell you whether your program is sustainable. Average time to open tells you whether timing and urgency fit the moment.
If you run this as a Growth or Performance Marketer, the fastest path to ROI is not guessing new copy. It’s building a metric stack that makes problems obvious, then running small experiments that move one metric without breaking another. That is how application push notification becomes a reliable lever instead of a noisy channel.
Measure, target, and scale your push programs with SashiDo - Push Notification Platform. Explore unified push, advanced segmentation, real-time delivery, and enterprise-grade diagnostics at https://www.sashido.io/en/products/push-notifications-platform.
