Push notifications. Ecommerce strategy.
When to ask for push permission, how to segment on day one, the three message types that drive revenue, and the frequency ceiling that keeps the channel alive past month two.
Timed prompts, segmented sends.
Push notifications for ecommerce in 2026 work when three discipline points are in place. One, ask for permission after a positive engagement moment (not first app launch); acceptance climbs from 25 percent to 50-plus percent. Two, segment into six minimum audiences (new, active, dormant, high-value, subscription, post-purchase) from day one. Three, hold frequency to 2 to 4 messages per week in the first 30 days and 4 to 8 per week for engaged users after. Three message types drive 70 to 80 percent of push revenue: abandoned-cart sequences, back-in-stock alerts, and order-update notifications. Klaviyo is the default tool for Shopify merchants 1M to 20M; OneSignal for stores not on Klaviyo; Braze for 20M-plus. iOS anti-tracking changes barely affect push because push runs on a direct APNS channel rather than cookies or pixels; delivery rates remain 95-plus percent.
Timing is the channel.
On iOS the push permission prompt is a one-time decision; if a user denies it, the app must direct them to Settings to change the choice. On Android the prompt was optional until Android 13 (2022); since then a similar one-time prompt applies. The user's decision determines whether push is available for the entire lifecycle of their relationship with the app - get it wrong and the channel is gone.
First-launch prompts get 25 to 35 percent acceptance on iOS, 35 to 45 percent on Android. Users have no context for what the app will send; the rational default is decline. Post-engagement prompts get 40 to 55 percent iOS, 50 to 65 percent Android. Users who have completed a meaningful action (added to cart, completed profile, viewed three products) have reason to want ongoing communication. The right moment is often at checkout: "Want shipping updates and special offers? Enable notifications." Contextualize the ask.
The pre-permission prompt pattern: show a custom in-app screen explaining what the app will send before triggering the system-level prompt. If the user taps "Not now" on the custom screen, do not trigger the system prompt at all. This preserves the ability to ask again later (the system prompt can only be triggered once); the custom screen can be re-shown weekly with new engagement context. Well-executed pre-permission flows push acceptance rates toward 60 percent combined.
Segmented from day one.
The six minimum segments any DTC commerce app should operate. New (installed within 7 days): lighter messages focused on onboarding and first-purchase nudge. Active (opened within 7 days): full mix including promotional and recommendations. Dormant (no open in 30 days): re-engagement sequence plus winback offer. High-value (above a lifetime-value threshold, typically top 10 percent of customers): earlier access to new products, exclusive sales, VIP-tier messaging. Subscription (on a recurring order via Recharge or similar): pre-shipment reminders, modify-order prompts, subscription anniversary recognition. Post-purchase (within 14 days of last order): review request, cross-sell recommendations, shipping updates.
Overlapping segments are fine - a new installer who purchases day 3 is both New and Post-purchase. Klaviyo, Braze, and OneSignal all handle overlapping segments with priority rules (which message wins when multiple qualify). The operating practice: no user receives more than one push in any 4-hour window regardless of how many segments they belong to.
For fashion, beauty, and lifestyle brands, add category-based segmentation on top of the six base segments. Users who have browsed dresses, users who have browsed skincare, users who have browsed accessories. This is where push CTR jumps from 1-to-3 percent average to 5-to-10 percent for well-targeted segments. The cost is operational; maintaining 20-plus segments requires tooling like Klaviyo segments that auto-update based on browse and purchase behavior.
Abandon, restock, update.
Abandoned cart sequences. The highest-converting push message type for commerce apps. Three-touch sequence works best: push 1 at 1 hour post-abandon ("Still thinking about [product]?"), push 2 at 24 hours ("Your cart is waiting"), push 3 at 72 hours (often with a modest incentive like free shipping over a threshold). Typical recovery rate: 8 to 15 percent of abandoned carts. Above 3 touches the fatigue cost exceeds the recovery lift.
Back-in-stock notifications. When a user has browsed a product that was out of stock and it returns to stock, a push notification drives 12 to 20 percent conversion because intent is unambiguous - this user wanted this exact product. Shopify's native inventory tracking makes back-in-stock triggers reliable; the push tool subscribes to the product restock event and fires to users who have browsed that product or added it to wishlist within the last 60 days.
Order update notifications. Transactional messages (order placed, payment received, shipped, out for delivery, delivered) land with 70 to 85 percent open rate because users expect them. Beyond the transactional value, order-update notifications drive app re-engagement: the user opens the app to check status, sees recommended products, and occasionally buys again. Pair order-shipped notifications with a related-product recommendation for a 3 to 7 percent incremental conversion rate. Skip this on the delivered notification - at that point the user is focused on the arrived product, not browsing.
Two to four, then climb.
First 30 days of a user's lifecycle: 2 to 4 messages per week. The user has just installed and is still deciding whether the app is worth keeping. Over-messaging pushes them toward disabling notifications or uninstalling. Most of the 2 to 4 weekly messages should be valuable to the user: onboarding education, order-related, genuinely relevant product recommendations.
After 30 days of engagement: 4 to 8 messages per week for active users, 1 to 2 for dormant users in re-engagement sequences. The cap is still lower than most email send-cadence norms because push is more interruptive. A promotional email sits in the inbox waiting; a push notification pings the user in the moment. The same message weight feels heavier on push.
Transactional messages do not count against the cap. Order-placed, shipped, delivered, password-reset - these are expected and welcomed. Promotional and retention messages are where frequency discipline matters. Tools like Klaviyo offer frequency capping rules that automatically suppress promotional sends to users who have already received N messages in the last 7 days; turn these on from day one. The common failure mode: a brand has a great launch, ramps frequency to 10 to 14 messages per week in month two, and watches push opt-in rates drop 20 percent within 60 days.
Klaviyo by default. Braze at scale.
Klaviyo is the default recommendation for Shopify merchants. Unified email, SMS, and push in one flow builder; deep Shopify integration for product, order, and customer events; pricing 60 to 2,000 dollars per month based on contact count. The unification is the advantage: a user on an abandoned cart path gets push 1, email if push 1 does not drive action, SMS if email does not drive action - all orchestrated in one journey instead of three tools.
OneSignal for brands not on Klaviyo. Push-only (no email or SMS), generous free tier up to 10,000 subscribers, paid from 99 to 1,000+ per month. Good for brands that already have email elsewhere and just need a clean push-specific tool. The tradeoff is orchestration: push sequences sit in OneSignal, email in Mailchimp or similar, and building cross-channel journeys requires integration work.
Braze for brands above 20M revenue with in-house growth teams. Enterprise pricing (30K to 500K+ annually), deepest segmentation and journey builder in the category, but overkill for smaller brands. The builder tooling is more powerful than Klaviyo's; the tradeoff is implementation cost, ongoing management complexity, and pricing. Below 20M revenue, Klaviyo covers 90 percent of what Braze offers at 10 percent of the cost. Related reading: mobile app launch checklist covers day-zero push permission strategy; Shopify mobile app builders comparison covers push capabilities in app-builder platforms.
Six answers.
When should I ask users for push notification permission?
After a positive engagement moment, never on first app launch. First-launch permission prompts on iOS get rejected by 60 to 75 percent of users; the pre-permission prompt context matters more than any other factor. Good moments to prompt: right after a user completes their first meaningful action (added to cart, viewed 3 products, completed profile setup), after a successful transaction, or when the user would obviously want updates (order placed, ask if they want shipping notifications). Well-timed prompts hit 40 to 55 percent acceptance rate; that is the difference between a dead push channel and a live one.
What is the right frequency for push notifications?
2 to 4 messages per week for the first 30 days, 4 to 8 per week for engaged users after that. Above those ceilings, users start disabling push notifications or uninstalling the app. Below them, the channel underperforms because users forget the brand. The exceptions: transactional messages (order confirmation, shipping updates, delivery) do not count against the weekly cap because users expect them. Promotional and retention messages (abandoned cart, product recommendations, sale announcements) are the ones that burn permission if over-sent. Segment by engagement level and vary frequency accordingly; a new installer gets fewer messages than a daily-app-opener.
What push notification types drive the most revenue?
Three types account for 70 to 80 percent of push-driven revenue for DTC commerce apps. One, abandoned-cart sequences sent 1 hour, 24 hours, and 72 hours after cart abandon; typical 8 to 15 percent recovery rate. Two, back-in-stock notifications for products the user has browsed but not purchased, typical 12 to 20 percent conversion because intent is high. Three, order-update notifications (shipped, out for delivery, delivered) which are transactional but convert on delight and drive app re-engagement. The messages that do not drive revenue: generic sale announcements to every user ('20 percent off sitewide'), re-engagement pings ('we miss you'), and new-collection announcements to people who have not purchased the category.
How should I segment push audiences?
Six segments that most apps should operate from day one. New (installed within 7 days, sent lighter-touch onboarding messages). Active (opened app within last 7 days, full message mix). Dormant (no app open within 30 days, re-engagement push series). High-value (above certain LTV threshold, earlier access and exclusive offers). Subscription (on a recurring product, pre-shipment reminders and modify-order prompts). Post-purchase (within 14 days of last order, review request and cross-sell). Beyond these six, add category-based segments (new dress browsers vs jewelry browsers) for DTC fashion or lifestyle brands. Tools like Klaviyo, Braze, and OneSignal all support these segmentations natively.
What tools should I use for push on Shopify?
Four tiers of tools. Tier one, Klaviyo: best-in-class for Shopify merchants because it unifies email, SMS, and push in one flow builder; 60 to 2,000 dollars per month based on contact count. Tier two, OneSignal: standalone push, free tier up to 10,000 subscribers, paid tiers from 99 to 1,000 dollars per month; strong for Shopify stores not using Klaviyo. Tier three, Braze: enterprise tier for brands above 20M revenue, 30K to 500K+ annual contract, deepest segmentation and journey builder. Tier four, Shopify app-builder native push (Tapcart, Vajro): push notifications included in the app subscription, handling basic segmentation but not the depth of a dedicated tool. Most 1M to 20M DTC brands run Klaviyo plus the app-builder native for the best of both worlds.
Do push notifications still work in 2026 with all the anti-tracking changes?
Yes, push is largely unaffected by iOS anti-tracking because push delivery is a direct APNS relationship between your app and Apple, not a cookie or pixel. Push delivery rates remain 95-plus percent on both platforms for subscribed users. What has changed: attribution of push-driven revenue to push specifically has become harder because cross-device and web-to-app attribution is degraded by iOS privacy changes. For attribution modeling, see the related post on Northbeam vs Triple Whale. The push channel itself - as a direct line to opted-in users - is one of the few channels that has strengthened relative to email and paid ads since 2022.
Push is a direct channel.
Our growth-strategy engagements include a full push-channel audit: permission flow, segment design, message library, frequency cap, and 90-day retention metrics. Scoped quote in 48 hours.