How to Write Effective Push Notifications That Guide Users to Their Goal (Best Practices + Examples)
You have a maximum of 7 seconds. That’s how long it takes a user to decide whether a notification from your app is worth their attention.
Spoiler: usually it isn’t.
The problem is that most push notifications are sent for the sender, not the recipient. A marketer wants to remind users about a promotion — so they send it. They want to drive traffic — they send it. They want to “maintain engagement” — they send it. Meanwhile, the average user receives around 46 notifications per day. From their bank, from friends, from Netflix, from a shop they visited two weeks ago. The brain quickly learns to sort — what to tap, what to ignore. And if your app keeps sending things the user skips, they’ll eventually turn them off entirely.
And why is that a problem?
Because push notifications are one of the few marketing channels that reach the user’s screen directly — with no cost per message, no reliance on ad algorithms, and no competing in a feed. The average opt-in rate for mobile apps is 67.5% globally. Apps with active push permission see 3x higher retention than those without it. This is not a marginal communication channel.
In this article, you’ll learn how to write effective push notifications that users actually want to tap… and why it’s worth starting with changes you can make today.
Why people turn off notifications — and what that says about your communication
The opt-out rate — the percentage of users who have turned off your notifications — is one of the most honest indicators of communication quality. People rarely write to tell you your pushes are useless. They just turn them off.
Three causes almost always lead to this: too many, too irrelevant, at the wrong time. Each of these can be fixed. But to know where exactly the problem lies, you need to look at the data — when precisely are users opting out? In the first few days after installation? After a specific campaign? Gradually, over several weeks?
Specifically: you’re looking for the moment when users drop off en masse. That’s where the data will tell you what’s wrong.
- If opt-out rate spikes in the first 3–7 days after installation — the problem likely lies in the frequency or relevance of onboarding notifications. New users don’t yet understand the app’s value and have very low tolerance for communication that makes no sense to them.
- If opt-outs spike after a specific campaign — that campaign was probably too aggressive, too broad, or reached a segment it shouldn’t have. Go back to that campaign and check: who received the message, what it promised, and whether the deep link led where it should have.
- If opt-out rate grows gradually over several weeks — you have a frequency or monotony problem. Users don’t turn you off after one bad notification. They turn you off after the tenth one that brought nothing new.
Three metrics worth keeping at hand before you start looking for the cause:
- Opt-out rate over time — is it growing linearly, spiking, or stable? The pattern says more than the number itself.
- Opt-out rate per campaign — which specific sends generate the most churn? This quickly points to the culprit.
- Time to opt-out from installation — the median will tell you at what stage of the relationship with the app users lose patience.
💡 One rule worth remembering — and it’s not obvious: every notification that adds no value isn’t neutral. It actively erodes trust in the entire app. Not in that specific notification — in the entire app.
Permission to notify: you get one shot. Don’t waste it.
On iOS, the system allows you to ask for notification permission exactly once. If the user taps “Don’t Allow” — you cannot automatically ask again. The only path to a second chance runs through manual system settings… and unfortunately, most users won’t take that path.
That’s why the so-called soft ask matters so much. Its only job is to explain to the user why it’s worth saying yes:
- What exactly will they receive?
- How often will you be in touch?
- Will it be something valuable to them, or just ads?
The difference between a bad and a good soft ask is straightforward:
❌ “Enable notifications to stay up to date!”
✅ “Your parcel is on its way! Enable notifications — we’ll let you know about important status updates.”
The first version is an empty statement. The second is a concrete value and a clear scope. Someone who reads it knows exactly what they’re agreeing to — and that’s precisely why they’re far more likely to tap: “Yes, I want this!”
Good timing: when should you ask for permission?
The best moment is right after the user’s first “success” in the app. In e-commerce — after adding something to the cart. In a fitness app — after completing the first workout. In SaaS — after the user has genuinely used the core feature. At that point, they understand what the app does for them and have a reason to want to be kept informed.
And the worst moment? The first launch. A user who just installed the app doesn’t yet know whether it’ll be useful to them. They have no reason to say yes.
A few practical examples:
What a good push notification looks like on the inside
A push notification consists of several elements, and each one does a specific job. It’s worth understanding them, because individual mistakes translate directly into lost clicks.
- The title is the first — and often only — thing the user sees. In 5–6 words, it should tell them why it’s worth tapping right now. Instead of: “We have something for you!”, something like “Your parcel leaves today at 2 PM” or “You have £230 remaining” will perform better. In short: specifics, not teaser copy.
- The body expands the title with one — exactly one — additional piece of information. If the body adds nothing new and simply restates the title in different words, leave it empty. Seriously.
- The deep link is an element that has an enormous impact on conversion, yet many apps neglect it. Every notification should lead to a specific place within the app — not the home screen. A notification about order status — to that specific order view. A notification about a message from a friend — to that specific conversation. A user who taps and lands somewhere unexpected is disappointed. And next time, they’ll tap less often.
What about length?
iOS on a locked screen shows the title up to around 50 characters and the body up to around 110. Android is slightly more generous, but the principle remains the same — write so that the message works at minimum visibility. Everything beyond that is a bonus for those who expand the notification.
It’s worth keeping in mind that personalised mobile push notifications achieve an average CTR of 5%–25% (depending on the industry and targeting precision). That’s roughly 4–7 times higher than mass sends to the entire user base (according to research conducted by OneSignal over the past three years).
But personalisation doesn’t stop at inserting a first name in the title.
That stopped impressing anyone a long time ago. It’s about relevance: the right content, the right moment, the right context for that specific person.
Because not all users need the same thing.
Sending the same message to everyone is like sending a company-wide email with the subject line “Important information for you.” It technically reaches everyone. But honestly — who’s really going to take that message to heart?
That’s why it’s so important to understand that a user who installed the app three days ago needs something completely different from someone who has been using it for a year. Someone who just abandoned their cart needs a different notification from someone who hasn’t logged in for a month.
5 segments to start with
- New users (first week) need onboarding and reassurance that they made the right choice — not promotions or upsells. Example: “Complete your profile — it takes 2 minutes and unlocks 3 premium features.”
- Regular, active users want to know about things that will genuinely improve their experience — new features, personal alerts, changes that affect them. They don’t need a reminder that the app exists. They know.
- Lapsed users (7–30 days without a session) need a concrete reason to return, not a generic “We miss you.” Example: “Your training plan for this week is waiting — you last worked out 12 days ago.”
- Users mid-flow — such as those who abandoned a cart — need a notification tied directly to the action they started. Example: “Your Nike Air Max are still in your cart. Last pair in your size.”
- VIP and high-value users appreciate exclusive treatment: early access, first-to-know updates, communication that sets them apart from the crowd.
When is the best time to send notifications?
On timing — there’s no single golden hour for everyone, but data suggests that for most categories, Wednesday and Thursday between 10:00–12:00 and 19:00–21:00 local time deliver the highest CTR. E-commerce tends to perform better Thursday–Friday evening; fitness apps work best in the morning and early evening; fintech in the morning and after work hours. Please treat this as a starting point for your own tests, not as gospel truth.
Behavioural triggers — the most effective format most teams aren’t using
Notifications triggered by specific user actions — rather than a marketer’s calendar — are the most effective form of push notification. The user browsed trainers for three minutes? The next day they receive a notification with a discount on that category. They dropped off at step 3 of registration? A reminder goes out two hours later.
This requires a platform that supports event-based triggers (Braze, CleverTap, MoEngage), but the ROI is clearly higher than with mass sends. For those with the right infrastructure — this is the thing worth implementing first.
4 metrics that will tell you whether your push notifications are working
- Delivery rate — how many notifications actually reached the device. Below 85% is a signal that you have a large number of inactive tokens: users have uninstalled the app or changed their phone. Clean your database regularly.
- Opt-in and opt-out rate are two sides of the same coin. An opt-in rate below 40% on iOS suggests an issue with the soft ask or poor timing of the permission request. An opt-out rate above 2% per month indicates you’re sending too much or too inaccurately.
- CTR — how many people who received the notification actually tapped it. Benchmarks vary by industry: for e-commerce, 3–5% is a solid result; for utility apps with well-segmented sends, 5–10% is achievable. A CTR below 1% on mass campaigns is a signal to improve.
- Post-click conversion rate is a metric many teams ignore — wrongly so. CTR tells you whether the notification was interesting. Conversion rate tells you whether the promise was consistent with what the user found after tapping. A high CTR with a low conversion rate almost always means either a bad deep link or content that promises more than it delivers.
A/B testing: where to start
Push notifications can be tested very quickly. Most clicks appear within 1–2 hours of sending, so results are visible almost immediately. That’s why A/B testing is a good practice. The rule is the same as with any other test: change one thing at a time, because otherwise you won’t know what caused the difference.
- The title has the biggest impact on CTR and is the easiest thing to test — start there. Test specific vs. vague, question vs. statement, with emoji vs. without.
- Timing is the second priority. Send the same message to two groups at different times. Even a 2-hour difference can shift the result by 30–50%.
- Personalisation — a version with specific user context versus a generic version — typically produces the biggest CTR lift, but requires proper data infrastructure. If you have it, test this as soon as possible.
- Rich notifications with images are worth testing especially in visual categories — e-commerce, food, travel. In utility apps, the difference is often minimal.
3 things you can start doing today
- First: check your current opt-out rate. If you’re not measuring it — that’s your starting point. The number will tell you whether you have a problem and how big it is.
- Second: if you don’t have a soft ask, add one. It’s one of the highest-impact changes you can make to opt-in rate, and it can be implemented relatively quickly.
- Third: divide your user base into at least three segments — new, active, inactive — and write a different message for each on your next campaign. You’ll see the difference in CTR.
Everything else — behavioural triggers, advanced testing, rich notifications — comes next. But the foundation is right here.
Summary
Effective push communication starts with permission: on iOS you have exactly one chance for the soft ask — a screen that clearly explains what the user will receive and how often. The best moment to ask is right after the user’s first success in the app, never at first launch. Apps with active push permission see 3x higher retention than those without it, and personalised pushes achieve a CTR 4–7x higher than mass sends.
A well-constructed push notification consists of a title (max. 50 characters, specifics over teasers), a body (one additional piece of information or left empty) and a deep link leading to a specific place within the app.
The key to strong results is segmentation: new users, active users, lapsed users, users mid-flow, and VIPs all need different messages. The highest ROI comes from behavioural triggers — notifications triggered by specific user actions.
Four metrics worth monitoring: delivery rate, opt-in/opt-out rate, CTR, and post-click conversion rate. A high CTR with a low conversion rate almost always means a bad deep link or content that promises more than it delivers.
Want to implement push notifications that genuinely engage?
If you’re building a mobile app and want your pushes to actually work… or you already have notifications but the results are disappointing — let’s talk. We’ll be happy to help you plan segmentation, set up behavioural triggers, and write content your users will actually want to read.
Data cited in this article comes from publicly available reports (Notix 2023, OneSignal 2023–2025, Braze, Localytics) or is marked as market observations.
