Push notifications are the most powerful re-engagement tool a mobile app gives you — and the most dangerous one to misuse. Used well, they are the reason app-owning customers buy more often than email-only customers. Used poorly, they are the reason someone uninstalls your app and turns off notifications from the whole category. This guide covers what works, what does not, and the mechanics of running push campaigns that your customers actually want to receive.

Why push outperforms email for re-engagement

Email is a pull medium in most inboxes. It arrives in a place people visit when they choose to, and it competes with dozens of other emails for attention. Push is an interrupt medium — it appears on the lock screen or notification center when the phone is in the shopper's hand. That is not an advantage you want to abuse, but it is a structural one when the message is timely and relevant.

Average open rates for push in e-commerce sit in the 40 to 60% range. Average email open rates are 20 to 25%. The difference is not because push is inherently better — it is because the cost of ignoring a push is lower (a swipe dismisses it), so people engage rather than leaving it to accumulate. The message needs to earn that open rate by being genuinely worth the interrupt.

The message types that convert

Not all push campaigns work equally well. These four categories consistently drive results across e-commerce categories:

Back-in-stock alerts

A shopper viewed a sold-out product and opted in to an alert. When it restocks, a push reaches them immediately. This is the highest-converting push type because the purchase intent was already established — the only thing missing was availability. A well-timed back-in-stock alert often converts at 20 to 30%.

Abandoned cart recovery

A shopper added items to cart and left without purchasing. A push 30 to 45 minutes later catches them while the intent is still warm, before the session memory fades. This performs better than email for same-session recovery because it reaches them on the device they were already using. Keep the message simple: what is in the cart, and a clear path back to it.

New arrival announcements

Your app installs are your most engaged customers — people who cared enough to download and keep your app. A new collection launch or product drop push to this audience is essentially a warm announcement to your most interested segment. Keep the message specific: "The fall collection is live" with a direct link to the new arrivals, not a generic "new things are here" that requires clicking and navigating.

Price drop alerts

A shopper viewed a product at full price and did not buy. When that product goes on sale, a targeted push tells them exactly what changed. This converts well because the barrier was likely price, not interest. The specificity matters — "The jacket you looked at is 20% off" performs much better than a blanket sale announcement.

The messages that destroy opt-in rates

Every push you send that does not feel valuable erodes the opt-in rate. Enough poor pushes and the shopper turns off notifications or uninstalls. The fastest ways to burn through goodwill:

  • Generic promotional blasts on a predictable schedule. "Weekend sale! 20% off everything!" sent every Friday teaches shoppers to ignore you. It adds no information they could not have found themselves.
  • More than one non-transactional push per week. More than one promotional message per week is the threshold where opt-out rates start rising noticeably. Most stores should aim for one or fewer.
  • Pushes with no clear action. "Check out our blog!" is not a reason to open the app. Every push should have a specific destination and a reason that benefits the recipient, not just the sender.
  • Sending to everyone when targeting is possible. A back-in-stock alert for a women's dress sent to your full install base including men creates irrelevance. When you can segment — by past browse history, by category affinity, by purchase history — do it.

Cadence and timing

For most e-commerce stores, one to two non-transactional push notifications per week is the ceiling before opt-out rates increase. One per week with high relevance performs better than two per week with mixed relevance.

Transactional pushes — order confirmations, shipping updates, back-in-stock alerts the user opted into — are not subject to the same fatigue. Shoppers want those and have a high tolerance for them because they requested them.

Time of day matters. Pushes in the early morning (6 to 8am) or evening (7 to 9pm) in the user's local time zone consistently outperform daytime sends. Daytime sends often arrive when people are at work and dismiss notifications quickly. Evening sends catch people in low-intent browsing mode where a well-timed offer can shift behavior.

Measuring what works

The metrics to watch are open rate (did the message earn an open), click-through rate (did they engage with the destination), and conversion rate (did they buy). High open rate with low conversion suggests the message was interesting but the landing experience did not close the sale. Low open rate suggests the message copy was not compelling enough or the timing was wrong.

The opt-out rate over time is the health metric — if it is climbing, something in your push strategy is creating friction. A flat or declining opt-out rate means your audience finds your messages valuable enough to keep.