Push notification metrics: How to measure performance 

Most marketers open their push dashboard, see a 4% open rate, and panic. Nothing is broken. Push notifications don't behave like email, so the metrics you trust for email will quietly mislead you here.

Push is one of the fastest-growing channels in mobile app marketing. In Customer.io's 2026 customer engagement trends report, 30% of teams said they use mobile push to reach customers, and engagement and retention ranked as the number one goal those teams are messaging toward. Push earns that spot because it reaches people on the device they check most, in the moment that matters.

The catch is measurement. A push notification can succeed without ever being "opened," which means a low open rate often signals the opposite of underperformance. Here's a strong opinion worth internalizing: open rate is a vanity metric for push, and reporting it to leadership as a health signal will make your best campaigns look like your worst.

This guide covers the push notification metrics that actually matter, how to read each one correctly, and how to improve performance once you're measuring the right things.

• Push open rates average around 4% in Customer.io platform data, and a low number is normal because people act on a notification without formally opening it.
• Delivery rate is an imperfect signal because mobile operating systems don't always report delivery back reliably, and iOS and Android report it differently.
• Conversion rate measured against a defined goal is the most reliable way to judge push performance, and you should gauge behavior within 48 hours of send.
• Customer.io's mobile SDK automatically tracks delivered and opened for push originating from the platform, and conversion goals are fully customizable to your own events and timeframes.
• Lugg triggers push on real booking behavior instead of scheduled blasts and saw a 2.5% monthly revenue lift from recovered bookings, with conversion rates reaching up to 20% during peak demand.

A low push open rate is normal and usually not a problem. The industry average open rate for push sits around 4% in Customer.io platform data, and that number reflects how people use their phones, not how well your message performed.

Picture a language-learning app sending a campaign to customers who finished the Spanish 101 course. The push reads: "Way to ace Spanish 101, Susan! Visit the app for an exclusive 40% off Spanish 102." To register as an open, Susan has to tap the notification or launch the app. Unlike email, she gets the full message just by glancing at her lock screen.

Say she's in a meeting when it arrives, so she glances down, makes a mental note, and dismisses it. At lunch she opens the app and signs up for Spanish 102. The push converted even though she never opened it. If open rate is your headline metric, you'll record that campaign as a failure. Conversion tells the real story.

Delivery rate is the share of messages that reach a device, calculated as delivered divided by sent. It tells you whether notifications are landing, but it's an imperfect signal because mobile devices don't always expose delivery information.

Delivery rate = (delivered / sent) x 100

If a message shows as delivered, someone very likely saw it and may have acted on it even without opening. The bigger issue is what delivery rate hides. On iOS, delivery is only reported when a user has opted into notifications and the message actually appears. On Android, delivery is reported regardless of opt-in status, which can make Android look like it delivers more than iOS when the real gap is smaller.

One current lever helps close part of that gap. Customer.io lets you set message priority in the push composer without touching your SDK. High priority tells Android devices to wake up and report delivery metrics properly, so standby devices don't leave you guessing whether a message landed.

Treat delivery rate as a diagnostic, not a scorecard. Aim for a high delivery rate, but read open and conversion rates for the truth.

Open rate is the share of people who open a push after receiving it, calculated as opens divided by sent. Treat it as a secondary signal, not your headline metric.

Open rate = (opens / sent) x 100

The 4% average is a feature of the channel, not a flaw in your campaign. People extract the value of a push straight from the notification, so an unopened push can still do its job. If your open rate looks low next to email, that's expected, and it's the wrong comparison to make.

For push, an open and a click-through are effectively the same action. When someone taps a push that contains a deep link, they land on the exact screen you linked to inside your app, so the tap counts as both the open and the click.

This is another place push diverges from email, where opens and clicks are distinct events. For push, an open signals that a person viewed the notification and actively engaged with it. Design your deep links so that tap sends people somewhere useful, because the tap is the click.

The main difference is how each operating system displays and reports notifications, not how well your message performs. iOS notifications appear once and then move to the notification center, and people can see them without touching the screen. Android notifications persist until someone swipes them away, which gives more chances to engage but also produces accidental opens when people clear their trays. Because of this, iOS and Android open rates aren't directly comparable.

Bottom line: Don’t put too much stock into the difference between iOS and Android open rates.

Unlike email marketing, push notification CTR is measured almost like an open.

Let’s say you’ve included a deep link to your app within the push notification. When a user taps your push notification, they’ll automatically be dropped to the specific screen you’ve linked to within the message. So, in this sense, push open rate and push click-through rate are the same thing.

Bottom line: An open signifies viewing the push notification and actively engaging with it via a click.

Conversion rate is the metric that actually tells you whether push is working. It measures the share of people who complete a desired action, like opening your app, finishing onboarding, or making a purchase, after a notification is delivered.

Conversion rate = (conversions / delivered) x 100

Return to Susan. She receives the push at 10:00 a.m. and signs up for Spanish 102 at 12:45 p.m. You can reasonably attribute that conversion to the push. A useful rule of thumb: gauge push performance by how customers behave within 48 hours of receiving a notification. As you watch behavior over time, you may find your audience responds inside a shorter or longer window.

This is where Customer.io's conversion goals do the heavy lifting. You define the behavior and the timeframe that matter for your customers, and the platform reports how many people hit that goal rather than just how many messages sent. That's a far better measure of success than raw sent and delivered counts.

Three moves reliably lift push performance: personalize every message, add rich media, and coordinate push with your other channels. Customer.io is used to send more than 5M push notifications a month from its own platform, and these are the patterns that consistently work.

Relevance is what keeps people opted in. Customers are quick to mute push, or delete an app, when messaging misses the mark. Behavioral triggers are the highest-return form of personalization, because responding to what someone just did beats sending on a fixed schedule.

Lugg, an on-demand moving platform, is a clear example. Instead of scheduled push blasts, Lugg triggers notifications on real behavior during the booking flow, so a stalled booking gets a timely, context-aware nudge rather than generic urgency. That approach delivered a 2.5% monthly revenue lift from recovered bookings, with conversion rates up to 20% during peak demand. Use Liquid to personalize the copy itself with your customer data, and deep links to send people back to the exact screen they left.

Rich push turns a plain text alert into something people want to tap. A simple push includes a title, text, and your logo, while rich push lets you add images, video, and deep links that give context and room to experiment.

Say you're announcing a new feature. Include an image of it in the notification and a deep link straight to the screen where a customer can try it. The visual earns attention, and the deep link removes a step, which usually lifts conversions.

Push works best as part of a coordinated, omnichannel messaging strategy, not in isolation. A single channel rarely carries a full customer journey, so the strongest programs move people between push, email, in-app, and SMS based on where they are in their lifecycle.

Take a win-back campaign for customers at risk of churn. You might send a "we miss you" push to guide them back into the app, follow with an email showing what's new since they last logged in, then greet them with an in-app message touring recent updates when they return. Coordinating those touches creates a consistent experience and lifts engagement across the board.

Respect frequency and consent, and give people control. Push is hard to ignore, which is exactly why overuse pushes customers to opt out or delete the app. A few Customer.io features make restraint easier to operationalize.

Set frequency limits and send-time logic so people don't get duplicate nudges across a phone and a tablet. Let customers subscribe or unsubscribe from individual channels like push, email, or SMS independent of topic preferences, so someone can keep your emails while muting push. And use the notification inbox to keep a copy of a push in the app, so a dismissed message isn't a lost one.

Around 4% is the industry average in Customer.io platform data, so anything in that range is normal. Because people often act on a push without opening it, open rate is a weak measure of success. Judge push by conversions instead.

That's expected behavior. Many customers read everything they need from the lock screen and never tap the notification, especially on iOS, where the message appears once and moves to the notification center. A delivered-but-unopened push can still convert.

Measure behavior within 48 hours of send as a starting rule of thumb. Most attributable conversions happen inside a day or two, though your audience may respond in a shorter or longer window. Set a conversion goal with a timeframe that fits your product.

Yes. The Customer.io mobile SDK automatically tracks delivered and opened for push notifications that originate from the platform. If you need to report metrics for messages sent another way, you can log them with the SDK's trackMetric method.

No. Email opens and clicks are distinct, trackable events, while a push open and click are effectively the same tap, and push open rates run far lower by design. Use conversion rate against a defined goal as your primary push metric.

Push notifications are capped at 4KB (4096 bytes) on both APNs (iOS) and FCM (Android). That limit covers the entire payload, including any custom data, so keep media and custom fields lean.

Push performance begins with measuring the right things. Track delivery, open, and conversion rates together, weight conversions most heavily, and read open rate as context rather than a verdict. From there, personalization, rich media, and coordinated cross-channel messaging are what move the numbers that matter.

Customer.io brings push, email, SMS, and in-app messaging into one place so you can build journeys around real behavior and measure them against goals you define. See how it works with a live demo, or explore the push notifications feature page to go deeper.