Last updated August 2026. Refreshed quarterly.
Most push notification benchmarks you can find are averages of averages: a mobile-app “reaction rate” here, an undefined “open rate” there, no percentiles, no metric definitions, and no way to tell whether your 2% CTR is a problem or a win. This report is different. These push notification benchmarks come from the PushEngage platform itself: 479 billion notifications delivered across 79,161 sites, analyzed through August 2026, with every metric defined explicitly and every benchmark published as a distribution, not a single number.
If you own a retention metric, this page gives you three things: the median (table stakes), the p75 (good), and the p90 (what the best senders in your position actually hit).
Key findings
Each stat below stands on its own. Citation guidance is at the end of this report.
- 479 billion web push notifications delivered lifetime across 79,161 sites on the PushEngage platform (through August 2026).
- 38.1% is the median view rate among established senders (sites with 100k+ lifetime sends). The platform-wide blended view rate is 26.3%.
- 2.29% is the median click-through rate on viewed notifications among established senders; 0.81% is the median CTR on sent.
- Top-decile senders earn a 6.35% CTR on views, 8× the bottom decile (0.79%), and roughly 12× on CTR-on-sent.
- Notification views fell 31.6% year over year (Aug 2025–Aug 2026 vs. the prior year), while clicks fell only 5.8%. CTR on views rose 37%, from 1.50% to 2.06%.
- Triggered campaigns deliver a 3.00% CTR on sent, 6.3× the all-sends baseline of 0.478%, and a 66.6% view rate (2.5× baseline).
- The blended web push opt-in allow rate is 12.8% across ~199 million prompt impressions in the trailing 12 months. Prompt style moves it from ~9% (hard native single-step) to ~38% (best-performing styled and two-step prompts).
- Finance sites have the highest view rate (51.1%); sports sites have the highest CTR on views (3.36%) of any vertical measured.
- News and media sites sent 45.4 billion notifications, the largest vertical by volume, but earn a below-average 1.31% CTR on views.
- 74.1% of new web push subscribers now arrive on mobile devices, up from 70.6% a year earlier. Platform-wide, unsubscribes exceeded new subscribers in both of the last two years.
Methodology: how these benchmarks were computed
This section exists because most push notification statistics in circulation are uncited or conflate three different metrics. Here is exactly what was measured.
Source. Aggregate, anonymized PushEngage platform data: 79,161 sites with analytics records and 479 billion+ notifications delivered lifetime through August 2026. Per-site distributions were computed on the 2,328 sites with more than 100k lifetime sends (the “established senders” cohort), so single-campaign outliers do not distort the percentiles. Year-over-year windows run August 15 to August 14. Opt-in rates come from roughly 199 million prompt impressions over the trailing 12 months. No individual customer is identified anywhere in this report.
Metric definitions. Three different numbers get called “CTR” in this industry, and they differ by an order of magnitude:
| Metric | Definition | Platform blended value |
|---|---|---|
| View rate | Notifications displayed ÷ notifications delivered | 26.3% |
| CTR on views | Clicks ÷ notifications viewed | 1.82% |
| CTR on sent | Clicks ÷ notifications delivered | 0.478% |
When a vendor quotes a “7–8% push notification CTR” without a definition, it is almost always a mobile-app reaction rate, measured on app push, on a base of opted-in app users. Web push CTR on sent is a fundamentally different (and smaller) number, because web push must first survive delivery and display on a browser the subscriber may not have open. Comparing your web push program against an undefined mobile-app figure will make a healthy program look broken. Throughout this report, every number states which definition it uses.
What we deliberately excluded. Campaign counts are never quoted as message volume. Revenue attributed to triggered campaigns is tracked in multiple currencies and is not normalized, so we do not publish a dollar total.¹
The 2026 shift: volume is being squeezed, quality is being rewarded
The most important push notification trend of 2026 is visible in one year-over-year comparison:
The headline: platform-wide views dropped by almost a third in a single year while clicks held nearly flat — so CTR on views rose 37%. Volume is being squeezed out and quality is being rewarded. Note that unsubscribes exceeded new subscribers in both years. Source: PushEngage platform data, August 2026.
Read that carefully. Platform-wide notification views dropped by almost a third in a single year. Clicks held nearly flat. The notifications that still get through are earning 37% more engagement per view than they did a year ago.
The cause is the platform crackdown. Chrome now auto-revokes notification permissions from sites subscribers ignore, Android 16 silences and demotes low-engagement notifications at the OS level, and Apple continues to tighten what reaches the lock screen. The OS vendors are running a quality filter over the entire channel: low-value senders lose reach, and the engagement they were diluting flows back to senders subscribers actually want to hear from. We cover what to change, setting by setting, in our 2026 notification crackdown compliance hub.
The strategic conclusion for a retention team: the era of buying reach with volume is over. Every benchmark below should be read against that backdrop, because the gap between median and p90 senders is exactly the gap the crackdown is widening.
Benchmarks by metric: medians and percentile spreads
Nobody else publishes the spread, and the spread is the useful part. A median tells you where the middle of the pack sits. The percentiles tell you what to aim for. Use this rule of thumb: median = table stakes, p75 = good, p90 = excellent.
All distributions below are computed across the 2,328 established senders (100k+ lifetime sends).
View rate benchmarks
View rate is notifications displayed divided by notifications delivered. It is the deliverability metric of web push: it measures whether your notifications physically reach a screen.
Read it this way: below 28% (p25) you have a list-hygiene problem before you have a copy problem. Above 48% is top-quartile; above 56% is top-decile. Source: PushEngage platform data, August 2026.
What good looks like: if your view rate is below 28%, you have a list-hygiene problem before you have a copy problem. Dormant subscribers, uninstalled browsers, and revoked permissions are absorbing your sends. Above 48% puts you in the top quartile; the senders above 56% are typically running aggressive list cleanup plus push notification segmentation so that fewer, more relevant notifications keep permissions alive.
CTR benchmarks (both definitions)
Why two bars: a “7–8% CTR” quoted elsewhere is usually a mobile-app reaction rate — a different metric on a different channel. For web push, 2.29% on views is median and 0.81% on sent is median. Source: PushEngage platform data, August 2026.
What good looks like: a 2.3% CTR on views is the middle of the pack. Cross 3.9% and you are in the top quarter of all established senders. The top decile earns 6.35% on views and 2.59% on sent, which is 8× and roughly 12× the bottom decile respectively. That spread is the single most underreported fact in push marketing: the channel is not uniformly “good” or “bad.” The same channel, on the same browsers, produces a 12× difference in outcomes depending on how it is run.
The revenue translation: at median CTR on sent (0.81%), a 100k-subscriber list produces about 810 clicks per full-list send. At p90 (2.59%), the same list produces about 2,590. If your click-to-purchase rate and average order value hold constant, moving from median to p90 execution multiplies revenue per send by 3.2× without adding a single subscriber.
Opt-in rate benchmarks
Opt-in rate is allows divided by prompt impressions, measured at step one of the prompt flow, across ~199 million prompt impressions in the trailing 12 months.
| Opt-in approach | Step-1 allow rate |
|---|---|
| Hard native single-step browser prompt | ~9% |
| Blended, all prompt styles | 12.8% |
| Best-performing styled and two-step prompts | ~38% |
Intermediate prompt configurations land between roughly 10% and 25%. The 4× gap between the hard native prompt and the best-performing styled prompts is the cheapest list-growth lever most sites never pull: it requires no traffic increase, only a change to when and how the permission is asked. If your allow rate sits near the 9% floor, start with the tactics in 7 ways to increase your web push subscription rate.
What good looks like: below 10%, you are showing the browser’s hard prompt cold, and the browsers punish that now (a blocked prompt is increasingly permanent). 12–15% is normal. Above 25% means your prompt timing and framing are doing real work; ~38% is the current ceiling among high-performing configurations.
Benchmarks by industry
Vertical segments are defined by site domain keywords (same methodology as our vertical deep-dives), computed on lifetime data from active senders. Segments can overlap; the eCommerce keyword segment and the Shopify segment overlap by construction.
The split that matters: Finance keeps the highest view rate (51.1%) because the content is consequential, while Sports converts best on views (3.36%) because event-driven urgency expires. News & Media dominates volume yet sits below average on both. Full site counts and send volumes are in the table below. Source: PushEngage platform data, August 2026.
| Vertical | View rate | CTR on views |
|---|---|---|
| News & Media | 28.5% | 1.31% |
| Gaming & Betting | 28.3% | 1.09% |
| Health & Wellness | 47.5% | 1.18% |
| Food & Recipes | 47.9% | 1.28% |
| Finance | 51.1% | 1.81% |
| Sports | 34.8% | 3.36% |
| eCommerce | 34.3% | 1.90% |
| Shopify stores | 33.9% | 0.92% |
| Education | 22.1% | 1.38% |
| Travel | 42.3% | 2.10% |
| Tech & SaaS | 32.5% | 1.21% |
Five verticals deserve interpretation:
Finance: the view rate king (51.1%). Finance subscribers keep notifications enabled because the content is consequential: price alerts, rate changes, account activity. When a notification might be about your money, you do not revoke the permission. The lesson transfers: the closer your notifications sit to something the subscriber materially cares about, the healthier your delivery stays.
Sports: the CTR king (3.36% on views, 1.17% on sent). Event-driven urgency is the whole story. A lineup announcement or a final score has a built-in expiry, and subscribers click because the value evaporates in minutes. Sports senders also benefit from natural send-time relevance; the match schedule does their timing optimization for them.
News & Media: volume king, engagement laggard. News sites account for 45.4 billion sends, roughly 71% of all vertical volume measured here, yet sit below average on CTR (1.31% on views) and near the bottom on view rate (28.5%). This is broadcast fatigue in its purest form: every story to every subscriber, all day. The crackdown lands hardest here, and the fix is editorial discipline plus segmentation, not more sends.
eCommerce and Shopify: the automation gap. The general eCommerce segment posts a respectable 1.90% CTR on views, while the Shopify-identified segment posts 0.92%. The most likely driver is campaign mix: stores that rely on manual broadcast promotions underperform stores running triggered lifecycle campaigns (see the next section, where the gap is quantified). If you run a store and your numbers look like the Shopify row, the problem is not the channel.
Health, Food: quiet permission health (~48% view rates). Habitual-content niches, where subscribers genuinely want the next recipe or the next health update, keep permissions alive for years. Their CTRs are modest, but their reach compounds.
Triggered vs. broadcast: the 6.3× multiplier
This is the most consequential chart in the report. Triggered campaigns are notifications fired by an individual subscriber’s behavior: a cart left behind, a price drop on a viewed product, an item back in stock. Broadcast is everything sent to a list on a schedule. Measured across 20,353 triggered campaigns, lifetime:
Why it compounds twice: a triggered notification is 2.5× more likely to be displayed at all (the subscriber is recently active, the browser is warm) and 2.5× more likely to be clicked once seen (it is about something they just did) — which multiplies out to 6.3× on CTR-on-sent. Source: PushEngage platform data, August 2026.
A triggered notification is 6.3× more likely to be clicked, per notification delivered, than the platform-wide baseline. It compounds twice: triggered notifications are 2.5× more likely to be displayed at all (the subscriber is recently active, the browser is warm) and 2.5× more likely to be clicked once seen (the content is about something they just did). Triggered campaigns in this dataset also drove 272,496 tracked conversions, which is the difference between a channel that reports clicks and one that defends its line item.
The 2026 shift makes this table a survival guide, not an optimization tip. The OS-level quality filters reward exactly what triggered sends produce: high engagement per delivery. If your program is still broadcast-first, the place to start is understanding triggered push notifications as a campaign class, then shipping the highest-ROI instance, cart abandonment push notifications, and graduating to multi-step journeys in PushEngage Workflows once the single triggers prove out.
Subscriber growth and churn: the honest numbers
Here is the number nobody in this industry publishes: platform-wide, unsubscribes exceeded new subscribers in both of the last two years. In the trailing 12 months, sites added 27.15 million new web push subscribers and lost 34.30 million. The year before: 35.45 million added, 42.26 million lost.
Churn is the silent tax on push programs. Lists do not fail loudly the way an email domain reputation does; they leak, one revoked permission and one uninstalled browser at a time, and the crackdown has automated the leak (Chrome now revokes on the subscriber’s behalf). Two implications:
Mobile is where lists grow now. 74.1% of new subscribers in the last 12 months arrived on mobile devices, up from 70.6% the year before. If your opt-in experience was designed on a desktop monitor, most of your prospective list never sees it as intended. Audit the mobile prompt first.
Retention practice is list defense. The senders holding p75+ view rates in the charts above are not lucky; they send fewer, more relevant notifications, segment by behavior, and prune dormant subscribers before browsers do it for them. If your unsubscribe or auto-revocation numbers are climbing, work through how to reduce your push notification unsubscribe rate before adding send volume. A shrinking list at a rising CTR is a program getting healthier. A shrinking list at a flat CTR is a program being filtered out.
Push notification benchmarks FAQ
What is the average push notification CTR?
Define the metric first. On the PushEngage platform (479 billion notifications), the median established sender earns a 2.29% CTR on viewed notifications and a 0.81% CTR on sent. Figures of 7–8% quoted elsewhere are typically mobile-app reaction rates, a different metric on a different channel. For web push, 2.3% on views is median, 3.9% is top-quartile, and 6.4% is top-decile.
What is a good opt-in rate for push notifications?
The blended web push allow rate is 12.8% (of ~199 million prompt impressions). Below 10% usually means a cold, hard native prompt. Above 25% is genuinely good, and the best-performing styled and two-step prompt configurations reach roughly 38%.
What is push notification open rate vs. view rate?
“Open rate” is an email term that push vendors borrowed inconsistently; some use it to mean views, some to mean clicks. View rate is the precise web push term: notifications displayed ÷ notifications delivered (platform blended: 26.3%; established-sender median: 38.1%). When you see an “open rate” claim, ask which denominator it uses before comparing.
What is a good view rate for web push notifications?
Median is 38.1% among established senders. Above 48% is top-quartile; above 56% is top-decile. Below 28% signals a list-hygiene problem: dormant permissions and stale subscribers are absorbing your delivery.
Do triggered push notifications really perform better than broadcasts?
Yes, and by more than most vendors claim: 6.3× on CTR-on-sent (3.00% vs. 0.478% baseline), with a 66.6% view rate. The advantage compounds because triggered sends reach recently active subscribers with content about their own behavior.
Why did push notification views drop in 2026?
Chrome auto-revocation, Android 16 notification silencing, and related OS-level quality filters cut platform-wide views 31.6% year over year. Clicks fell only 5.8%, so CTR on views rose 37%. The channel is not dying; it is being filtered, and the filter favors relevant, triggered, well-segmented senders.
Which industry has the best push notification engagement?
Depends on the metric: finance leads view rate (51.1%), sports leads CTR (3.36% on views). News and media dominates volume (45.4B sends) but sits below average on engagement.
How big should my push notification list be before benchmarking?
These distributions were computed on sites with 100k+ lifetime sends. Below that, single campaigns swing your averages too much to compare meaningfully. Track your trend line instead, and revisit the percentile charts once your volume stabilizes.
How to cite this report
Suggested citation:
PushEngage, “Push Notification Benchmarks 2026,” analysis of 479 billion notifications across 79,161 sites, August 2026. https://www.pushengage.com/push-notification-benchmarks/
You are welcome to reference any statistic on this page with attribution and a link. Journalists, analysts, and researchers who need a specific cut of the data (a vertical, a region, a time window, or a metric we did not publish) can request one; write to the PushEngage team and we will run it against the same dataset with the same definitions. All cuts are aggregate and anonymized; we never disclose individual customer data.
These benchmarks are refreshed quarterly. The percentile charts above are the standing reference; when a quarterly refresh materially moves a number, the change is noted inline.
If the gap between your numbers and the p75 bar is the thing keeping you up at night, PushEngage ships the triggered campaigns, segmentation, and opt-in tooling that the top-decile senders in this dataset run, on plans that scale only with active subscribers.
¹ Triggered campaigns in this dataset attribute hundreds of millions in tracked revenue across multiple currencies; because the total is not normalized to a single currency, we do not publish it as a dollar figure.
