You already know the channels that used to drive reliable app growth are getting less reliable. Search clicks are declining as AI answers replace them, social reach is rented, paid costs keep climbing. None of that is exactly new. What's worth talking about is what it changes about where your team should be investing.

If every acquisition channel you depend on can be repriced, throttled, or restructured by someone else's algorithm, the highest-leverage thing you can build is an audience you don't need those channels to reach.

An email list isn't an owned audience

When people hear "owned audience," they tend to think of an email list. That's part of it, but at its core it’s still a single-channel asset with a single opt-in. An owned audience is the full set of users who have opted into a direct, ongoing relationship with you across any combination of channels: push notifications, email, SMS, RCS, in-app messaging, or logged-in app sessions.

A healthy owned audience has three properties.

1. You can reach these users without paying an intermediary or hoping an algorithm surfaces your content.
2. You have behavioral data on each user that allows you to personalize what you send and when.
3. The relationship compounds over time rather than resetting every time a platform changes its rules.

That third property is the one most teams underinvest in. Every interaction a user has with your app generates signals for exactly how to retain them: what they browse, what they skip, how often they come back, which features they use. Combine that behavioral data with declared preferences (what users explicitly tell you they care about) and you have two datasets that make your messaging smarter with every cycle.

The mobile growth leaders at Surge London were emphatic on this point: user-declared preferences become first-party data you can activate across messaging, personalization, and product surfaces. The behavioral layer tells you what users actually do. Together, they form the foundation that makes an owned audience worth having.

4 ways to build an audience that sticks around

These four approaches are the ones we see working consistently across mobile teams that treat their owned audience as a strategic asset.

Let users shape their own notification experience

Users who feel ownership over what they receive from you are more likely to stay opted in and more likely to engage when a message does arrive. This sounds obvious, but most apps still treat notifications as a one-way broadcast: the marketing team decides what to send and when, and the user's only options are "allow" or "don't allow."

The better model is to give users real control. That means preference centers where they can choose topics, frequency settings they can adjust, and channel options that let them decide whether they want to hear from you via push, email, SMS/RCS, or some combination.

When users curate their own experience, two things happen: the messages they do receive feel more relevant (because they chose them), and you get a cleaner signal about what each person actually wants (because they told you). Both of those make your messaging more effective and your audience more durable.

Participation builds the kind of loyalty content can't

Content can be generated at scale now. Recommendations can be automated. What can't be replicated by AI is the feeling of contributing to something and having that contribution matter.

The most compelling example from the past year came from a news broadcasting team who ran live polls during on-air debates, asking fans for their predictions on the Premier League title race. The results were fed back to the pundits within the hour, influencing the follow-up discussion. That segment was then clipped for YouTube and delivered back to the fans who had participated. The users went from passive listeners to co-creators of the content, and the loop gave them a reason to come back for the next one.

You don't need a live broadcast to create participation. Polls, predictions, user-generated content, feedback mechanisms that visibly influence the product, community challenges with shared results. Any interaction where the user adds value and sees that value reflected back creates a form of investment that passive consumption doesn't. These are the kinds of experiences that leading mobile teams are building to differentiate in a world where content alone is increasingly commoditized.

Let segments breathe

Most teams segment their audience. Fewer teams build segments that can keep up with how quickly user intent actually changes. A user who was highly engaged last week might be in a quiet phase this week, and a segment that can't account for that shift will send the wrong message at the wrong time.

If your segments don't account for these natural rhythms, your messaging will feel tone-deaf: re-engagement campaigns hitting users who are in no danger of churning, promotional blasts reaching people who just converted yesterday, etc.

Your segments should have decay and exit logic built in. If a user hasn't taken a relevant action in seven days, move them out of the high-intent segment automatically and shift to a lower-frequency check-in cadence rather than continuing the same aggressive messaging. When they re-engage, let them flow back in. This kind of dynamic segmentation requires behavioral event triggers and journey logic that can respond to real-time signals, but it's what separates an audience that feels cared for from one that feels spammed.

Not every user needs every channel (yet)

Picture a first date where someone asks for your phone number, email, Instagram, and home address before the appetizers arrive. Talk about a red flag. That's what it feels like when an app asks a new user to opt into push, email, and SMS all in the same session.

A better approach is to layer channels based on where the relationship stands. A new push subscriber who hasn't given you their email yet gets value-first content designed to earn that next opt-in over time. A user who's on push and email gets coordinated cross-channel journeys that use the right channel for the right message type: time-sensitive alerts on push, longer-form content via email. A power user on push, email, and SMS gets the full orchestrated experience.

The progression should feel natural, not forced. Each channel you add gives you a new way to deliver value and a new data point about how the user prefers to engage. The goal is to build a cross-channel relationship where every message reaches the user through the medium that's most likely to be useful to them in that moment.

The metrics that tell you if it's working

Let’s sidestep these for a moment: open rates, click-through rates, conversion per send. Those matter, but they tell you how individual messages performed. They don't tell you whether your owned audience is getting healthier or eroding over time.

The metrics that actually track audience health are different.

• Opt-in rate trends across each channel tell you whether you're growing or shrinking your reachable base.
• Multi-channel depth (what percentage of users are subscribed to two or more channels) tells you how resilient your audience is.
• Engagement by relationship tenure tells you whether users who've been subscribed for three months are still opening messages or whether they've gone quiet.
• Churn rate segmented by opt-in status tells you the most important thing of all: whether your owned audience members actually retain better than users you can't reach directly.

If that last metric doesn't show a meaningful gap, the audience isn't owned in any practical sense. It's just another list. The whole point of building direct relationships is that those relationships translate into longer retention, higher lifetime value, and more opportunities to deliver value over time. Measure accordingly.

Build the relationship layer with OneSignal

Everything described in this post (preference-driven opt-ins, participatory experiences, dynamic segmentation with exit logic, channel layering across push, email, SMS, RCS, and in-app) requires a platform that can manage cross-channel messaging from a single system with shared user profiles and real-time behavioral data.

OneSignal is built for exactly this. Our Journeys builder lets you create automated flows that respond to user behavior across channels, with branching, wait conditions, and exit rules that adapt as users move through natural engagement cycles. Dynamic segments update in real time based on custom events and user tags, so your messaging reflects what users are doing right now, not what they did last month. And every channel is managed from one dashboard with one set of user profiles, so you can layer channels thoughtfully rather than running them in silos.

One in five new apps use OneSignal. See what they're building with us.

Mobile communication expectations for online shoppers are shifting rapidly. Plain-text alerts that performed five years ago now compete with rich, interactive experiences from native apps, social DMs, and AI-driven shopping assistants. Modern shoppers expect interactive, visual, and verified communication directly inside their native messaging apps, and the protocol you choose meaningfully impacts both conversion rates and customer trust.

When evaluating RCS vs SMS, online retailers need to weigh the universal reliability of traditional text against the dynamic capabilities of modern messaging protocols. Below, we break down the technical differences, real-world eCommerce use cases, a decision framework for choosing between the two, and how leading retailers integrate both inside a single omnichannel marketing platform.

Understanding the basics

To build an effective mobile communication strategy, it’s necessary to understand the technical foundations of both primary protocols, including a third option (MMS) that often gets confused with both.

What is SMS?

Short Message Service is the foundational text messaging protocol used by cellular networks globally. Its core function is transmitting plain text between mobile devices over voice networks. The primary advantage is universal compatibility, regardless of device type or cellular carrier, standard text messages are virtually guaranteed to reach the recipient.

The limitations are equally well known. Standard texts are restricted to 160 characters per segment, with longer messages broken into multiple segments (each charged separately). They lack native support for rich media, so any image or video must be sent via MMS (Multimedia Messaging Service) or linked externally.

A quick note on MMS: MMS allows images and short videos but is significantly more expensive than SMS per message, has inconsistent rendering across devices and carriers, and lacks RCS’s interactivity.

What is Rich Communication Services?

Rich Communication Services (RCS) is the modern upgrade to traditional cellular messaging. Unlike standard texts, RCS relies on cellular data or Wi-Fi connectivity rather than voice networks. The protocol has been available on Android via Google Messages for years, and Apple’s adoption of RCS in iOS 18 in late 2024 effectively made it a cross-platform standard. By 2026, the vast majority of US smartphones can send and receive RCS messages, though carrier rollout still varies by region internationally, and Universal Profile (the standard governing cross-carrier RCS) has not been fully implemented in every market.

Rich Communication Services unlocks several core capabilities for retail retention communication:

• Rich media: high-resolution images, video, and audio shared directly in the chat thread.
• Interactive elements: suggested replies, action buttons, product carousels, and inline maps.
• Verified sender profiles: businesses display their full company name, logo, and a verification badge — a major trust signal.
• Engagement data: read receipts, typing indicators, and message-level analytics.
• No character limits: deliver complete promotional content in a single coherent message.

The tradeoffs are real. RCS requires data connectivity, so messages can fail in poor coverage areas or for users on older devices. Cost per message is higher than SMS (though still significantly cheaper than building and maintaining a native app). And while Apple’s adoption closed the largest cross-platform gap, international coverage still depends on local carrier readiness.

RCS vs. SMS: Core differences for eCommerce

Choosing between protocols requires examining the specific features that impact retail operations. The table below summarizes the key dimensions:

The strategic point is clear: SMS wins on reach and reliability; RCS wins on engagement and trust. Treating it as either/or is the wrong frame. The right question is which protocol serves which job.

Why eCommerce businesses are adopting RCS messaging

RCS lets retailers deliver app-quality experiences without forcing users to download an app. That changes the economics of mobile engagement in a few specific ways. Below are the use cases where RCS most consistently outperforms SMS for eCommerce:

Cart abandonment with rich product display

Standard SMS cart recovery typically reads, “You left something in your cart! [link].” RCS lets you send the actual product image, name, price, and a direct “Complete Purchase” button — all inside the message thread. Retailers running RCS cart recovery sequences typically see meaningful uplift over equivalent SMS sequences, though specific numbers depend heavily on category, AOV, and timing.

Conversational commerce flows

RCS supports two-way conversational flows that SMS cannot. A customer can tap “Track my order,” “Speak to support,” or “See similar products” directly from a button — each routing to the appropriate next step inside the messaging app. For eCommerce categories with high pre-purchase consideration (furniture, electronics, luxury), this is closer to live chat than to a notification.

Loyalty program engagement

Loyalty members can browse new drops, redeem points, or RSVP to in-store events through carousels and suggested replies. Because RCS preserves brand identity through verified sender profiles, these flows feel like an app experience, without the app fatigue.

Order tracking and post-purchase

Rich order updates — tracking maps, estimated delivery windows, and action buttons like “Reschedule delivery” or “Leave with neighbor” — replace the SMS-and-link pattern with something that feels native to the messaging app. Post-purchase is often the highest-ROI place to start with RCS because the customer is already opted in and the content lends itself naturally to rich formatting.

The tradeoffs are real: cost per message is higher, poorly optimized media can load slowly on weak data connections, and rich messaging requires more design and content investment than plain SMS. Treat RCS like email marketing in terms of creative effort, not like SMS.

The continuing role of SMS in retail

Despite RCS adoption, traditional SMS is far from obsolete. It remains a critical fallback mechanism and the right tool for specific jobs:

• Two-factor authentication and security codes. Universal reach, near-instant delivery, no rendering surprises.
• Order and shipping confirmations. Simple status updates don’t need rich media — they need to arrive reliably.
• Fraud alerts and account notifications. Time-critical, low-complexity messages benefit from SMS’s reliability.
• International audiences in low-RCS regions. In markets where Universal Profile rollout is incomplete, SMS remains the default for guaranteed delivery.
• Lowest-cost broadcast. For high-volume promotional sends where rich media isn’t needed, SMS pricing wins.

Relying solely on data-dependent protocols for transactional alerts introduces unnecessary risk into critical customer experiences. SMS is the floor that catches every message RCS can’t.

When should you use RCS vs. SMS? A decision framework

A simple way to decide which protocol to use for each message type:

Pair this with automated fallback logic and the choice is rarely binary — most messages should attempt RCS first and fall back to SMS when delivery fails or the recipient lacks RCS support.

Building an integrated strategy with an omnichannel marketing platform

The most effective retail communication strategy doesn’t force a choice between the two protocols. A robust omnichannel marketing platform lets businesses run both from a single workflow, with intelligent routing between them. Three capabilities are essential to look for:

1. Automated RCS-to-SMS fallback. The platform should attempt RCS first, and if the recipient lacks RCS capability (or the message fails to deliver), automatically deliver a thoughtfully crafted SMS version. The fallback shouldn’t be the RCS message stripped of formatting — it should be designed as its own message.
2. Sender verification management. Setting up verified business sender profiles across carriers is non-trivial. The right platform handles registration, brand asset management, and ongoing compliance.
3. Unified analytics across protocols. You need to see RCS read rates next to SMS delivery rates, attribute conversions across both channels, and understand which message types perform best on which protocol.

The primary risk of running both protocols is operational complexity. Managing two formats requires careful planning to keep the customer journey consistent regardless of which message a user receives. Platforms that surface this complexity in tooling, rather than hiding it, tend to be the ones that scale well.

Implementation considerations before launching RCS

A few practical considerations worth flagging before you go live:

• Sender verification takes time. Carrier approval for verified business profiles can take weeks. Start the process early.
• Cost modeling matters. RCS pricing varies by carrier, country, and message type. Get pricing tiers in writing before scaling spend.
• Test on real devices across networks. RCS rendering is consistent in principle but can vary in practice. Validate on iOS, Android, and across major carriers before going broad.
• Design the fallback experience deliberately. A meaningful percentage of your audience will land in SMS fallback. That experience needs to stand on its own — not feel like a degraded RCS message.

Wrapping up

The decision between traditional text messages and rich communication services isn’t binary — it’s a question of which protocol serves which job inside a unified strategy. SMS provides the universal reliability required for transactional and time-critical alerts. RCS delivers the rich, interactive experiences that drive eCommerce conversions and build brand trust through verified sender identity.

A modern retail communication strategy combines both, routed intelligently by an omnichannel marketing platform that handles RCS-first delivery with automatic SMS fallback. Done well, this maximizes reach without compromising on engagement — giving eCommerce brands access to messaging experiences that previously required a native app.

If you’re evaluating where to start: pick one high-value RCS use case (cart abandonment, order tracking, or a loyalty re-engagement flow), measure it head-to-head against the SMS equivalent for 30 days, and use the result to scope a broader rollout.

Mobile users abandon applications at alarming rates. The average app loses the majority of its daily active users (DAU) within the first week of install, and the long tail of mobile churn extends well past Day 30. Marketing and product teams often try to solve this by stitching together disconnected messaging tools: a push vendor here, an email tool there, an in-app SDK from somewhere else.

The result of a fragmented retention strategy is predictable: disjointed user experiences, conflicting messages, and retention curves that don’t move.

Below is a practical framework for evaluating and selecting an omnichannel marketing platform based on actual user behavior, rather than a series of vendor feature checklists. We’ll cover how to audit your retention DNA, which capabilities actually matter, how to pressure-test vendors with use cases instead of generic RFPs, and the questions that separate good platforms from category leaders.

Platforms like OneSignal have become a category standard for this exact challenge, with one in every five new apps using it to onboard, engage, and retain customers in an increasingly AI-driven and noisy landscape.

Map your retention DNA first

Successful evaluation starts with auditing your own user behavior. Every mobile app has a unique retention “DNA”: the specific drop-off points, friction moments, and engagement triggers that define how your users actually behave. You have to map this DNA before any vendor conversation, or you’ll buy capability you don’t need and miss capability you do.

The clearest way to do this is with a problem-to-solution framework. Identify the specific business need first, then match it to the messaging solution. For example, if user activity drops sharply 3 days post-install, the business need is proactive education during the first-week activation window. The corresponding solution is an automated welcome series that introduces core features across push and in-app messaging, sequenced based on which features a user has and hasn’t engaged with.

A useful tactical exercise: pull your last 90 days of cohort retention data and identify your three biggest drop-off cliffs (typically Day 1, Day 7, and Day 30). These three cliffs are where 80% of your platform value will come from. Anything a vendor can’t directly address at those three points is, at best, a secondary consideration.

Aligning product and marketing goals

Bridging the gap between product milestones and marketing goals is critical for long-term growth. Product teams typically optimize for feature adoption and activation milestones. Marketing teams typically optimize for campaign-level metrics like open rate and conversion. These are often quietly misaligned. A marketing campaign can hit its open rate target while pushing users away from a product milestone.

The fix is shared engagement KPIs. Three that work well in practice:

• Activation rate by acquisition cohort — a shared metric between product and growth marketing.
• Time-to-first-meaningful-action, tracked across channels and surfaced in both teams’ dashboards.
• Channel-attributed retention lift — what percentage of users who received a sequence on Day 3 were still active on Day 14.

Once both teams share these metrics, procurement conversations get sharper. You stop evaluating isolated email or push tools and start evaluating unified platforms that own the entire customer lifecycle.

Shift from broadcasting to true orchestration

Legacy multi-channel marketing relies on “broadcasting,” or sending the same message to large segments through multiple channels. Modern orchestration relies on sequencing, timing, and channel-aware logic. The most common (and costly) anti-pattern is firing the same message simultaneously across every channel — a push, an email, an SMS, and an in-app banner all hitting at once. This is one of the fastest ways to drive uninstalls.

True orchestration treats each channel as part of a unified journey, not parallel broadcasts. As a rule of thumb:

• Push notifications grab attention for time-sensitive, single-action moments.
• In-app messaging educates users while they’re already engaged in your product.
• Email carries longer-form, context-heavy content that benefits from richer formatting.
• SMS and RCS are reserved for high-stakes, time-critical communication where reach matters more than nuance.

Trigger-based sequencing in practice

Your retention platform must support automated, logical pathways triggered by real-time user actions, not pre-scheduled blasts. Here’s what a well-orchestrated cart abandonment sequence actually looks like:

• Hour 0: user adds item to cart, leaves the app.
• Hour 1: push notification — “Still thinking it over? Your cart is saved.”
• Hour 24 (if no return): email with richer content — product images, related items, social proof.
• Hour 72 (if still no return): in-app message on next session — final nudge with a softer CTA.
• Suppression rule: if the user returns and converts at any point, all subsequent messages are suppressed instantly.

The critical capability here is cross-channel awareness. The platform must know what was sent on every channel, what the user did in response, and adjust subsequent messages accordingly. Without that shared context, sequencing collapses back into broadcasting.

Key capabilities of a high-performing omnichannel marketing platform

Selecting an omnichannel marketing platform requires looking past surface features to evaluate the underlying infrastructure. Five capabilities matter most:

1. Real-time data actionability. The platform must not only ingest user data but make it usable for segmentation and targeting within minutes (not hours.) If “real-time” segmentation requires a 24-hour batch refresh, you can’t run behavioral sequences that respond to in-session actions.

2. Native cross-channel orchestration. Push, email, SMS, in-app, and emerging channels like RCS and Live Activities should be orchestrated from a single workflow builder, not bolted together via integrations. Ask the vendor directly: can a single journey branch across channels based on real-time engagement signals? If not, you’re buying a multichannel tool, not an omnichannel one.

3. Tool stack integration. Your platform should read from existing analytics, CDP, and product databases without forcing engineering rebuilds. Look for native integrations with your data warehouse (Snowflake, BigQuery, Redshift), CDP (Segment, mParticle), and product analytics (Amplitude, Mixpanel).

4. Workflow building accessible to non-engineers. A drag-and-drop journey builder isn’t just a UX preference, it determines who can ship campaigns. If every change requires engineering tickets, your iteration speed drops by an order of magnitude.

5. Developer-friendly APIs and SDKs. Marketing-friendly UIs need to be backed by robust REST APIs, well-maintained mobile SDKs, and webhook support. The best platforms serve both audiences without compromise.

Omnichannel marketing platform vs. customer retention platform: What’s the difference?

These terms often get used interchangeably, but they describe different (overlapping) categories:

• An omnichannel marketing platform is defined by channel breadth and orchestration — the ability to deliver and coordinate messages across multiple channels within unified journeys.
• A customer retention platform is defined by outcome focus — explicitly optimizing for activation, engagement, and long-term retention rather than acquisition.

The best tools are both. A retention platform without omnichannel reach can’t meet users where they are. An omnichannel platform without retention-focused features (cohort analytics, behavioral segmentation, lifecycle journey templates) is just an expensive messaging tool. When evaluating vendors, score them on both dimensions independently.

How to evaluate a customer retention platform

Evaluating a customer retention platform is much easier when you build use-case scenarios rather than checking technical requirements off a list. Force vendors to demonstrate exactly how their software solves your specific funnel drop-offs, not their generic capabilities.

Map retention challenges to required capabilities

Use the following framework to translate your common retention challenges into the specific platform capabilities required to solve them.

Questions that pressure-test any vendor

A useful sanity check during vendor evaluation is forcing demonstrations of specific capabilities. The following questions tend to separate marketing claims from actual depth:

1. “Show me a journey that branches across three channels based on real-time behavior, not pre-scheduled steps.” Most platforms can run pre-scheduled multi-channel campaigns. Few can branch dynamically in real time.
2. “How quickly does a new user event become available for targeting?” Acceptable answer is single-digit minutes. Anything measured in hours is a real-time gap.
3. “What happens if a user is mid-journey and our team updates the journey logic?” This tests whether the platform handles journey versioning gracefully or breaks active sequences.
4. “What’s your data residency and compliance story for GDPR, CCPA, HIPAA, and any regional requirements we care about?” Especially critical if you operate in the EU, Brazil, or California.
5. “What does the worst-case migration off your platform look like?” Vendors who answer this directly tend to be the ones worth working with.
6. “Can you show me three customers in our segment and the retention lift they’ve seen?” Forces real customer stories, not curated case studies.

Red flags during the evaluation process

A few patterns to watch for during demos and procurement:

• “All-in-one” pitches that gloss over channel depth. Some platforms cover every channel at surface level but execute none of them well. Ask for channel-specific deliverability and engagement data.
• Demo environments that don’t use real data volumes. If a vendor can’t show you a customer-scale instance, you’ll likely run into scaling issues post-purchase.
• Reliance on a single channel’s success. A push notification specialist that “also does email” is rarely as strong on email as a true omnichannel platform.

To sum up: Choose the platform that fits your journey, not the other way around

Technology alone cannot fix a broken retention strategy. The best platform is simply the one that accurately executes your company’s unique user journey orchestration. User expectations for personalized and non-intrusive communication are at an all-time high, and AI-driven content fatigue means the bar for relevance keeps rising.

Before you book your next demo, do three things:

1. Audit your retention drop-off points. Identify your three biggest cohort cliffs from the last 90 days.
2. Define your top 3–5 operational use cases. Tie each one to a specific retention metric you’d expect to move.
3. Map your user journeys for each use case before any vendor conversation — channels, timing, suppression rules, and success criteria.

The teams that win the next phase of retention won’t be the ones with the most tools. They’ll be the ones with the clearest journeys and a platform that can execute them.

Explore how OneSignal balances depth and breadth when it comes to strengthening mobile retention.

Historically, customer engagement platforms were built for dashboards.

Users learned segmentation systems, campaign builders, analytics interfaces, channel configuration, delivery tooling, and platform-specific workflows. The software was powerful, but operating it required expertise.

AI changes that.

Instead of learning the mechanics of a platform, users can simply dictate the outcome they want and have the software execute on their behalf.

This shift is occurring across software. Developers are using natural language from the command line. AI agents are completing multi-step workflows across tools. Chat interfaces are replacing complex navigation patterns. Operational work is becoming increasingly automated.

We believe customer engagement software will evolve the same way.

Today, we’re introducing two major steps in that direction:

• The OneSignal MCP Server (available now in Open Beta)
• OneSignal AI (available today through an early access program)

Together, they make OneSignal accessible through AI-native workflows — whether inside our platform or through external AI assistants like Claude, Cursor, and ChatGPT.

Introducing the OneSignal MCP Server

The OneSignal MCP Server allows AI assistants like Claude, Cursor, and ChatGPT to securely interact with OneSignal on a user’s behalf.

Rather than switching between dashboards, documentation, APIs, and external tools, users can work directly through their preferred AI environment using natural language.

This transforms AI assistants from passive advisors into active operators inside OneSignal.

For example, an agent can:

• Retrieve campaign analytics
• Inspect segments
• Manage templates
• Look up users and subscriptions
• Export data
• Trigger messaging workflows

More importantly, MCP enables cross-platform orchestration. An external AI agent is no longer limited to operating inside OneSignal alone. It can coordinate workflows across multiple systems simultaneously.

For example:

• Pulling data from a warehouse
• Analyzing results
• Creating a segment in OneSignal
• Generating campaign copy
• Scheduling a send

All from a single conversational workflow.

Our MCP server currently supports 31 tools across the following platform features:

• Messaging (3 tools)
• Users and subscriptions (12 tools)
• Templates (4 tools)
• Segments (4 tools)
• Live Activities (2 tools)
• Exports (2 tools)
• Analytics and utility (4 tools)

You can read more about the MCP and how to set it up in our documentation.

Introducing OneSignal AI

While the MCP server is designed for external AI ecosystems and agentic workflows, OneSignal AI brings AI directly into the OneSignal platform itself.

OneSignal AI is an embedded assistant available within the OneSignal UI that allows customers to interact with the platform conversationally.

Users can ask questions like:

• “How did our push notifications perform last week?”
• “Which segments are growing the fastest?”
• “Create a highly engaged email audience.”
• “Draft a push campaign for users who abandoned checkout.”

OneSignal AI performs actions directly within the platform, reducing the need to manually navigate workflows or learn platform-specific implementation details.

How is the MCP server different from OneSignal AI?

The MCP server is how external agents like Claude, ChatGPT, Cursor can access and use OneSignal. OneSignal AI enables chat based queries and automations directly in the OneSignal platform.

The MCP server is more about making OneSignal dynamically readable and actionable to external agents.  This unlocks a powerful use case for the MCP, which is the ability for users to command agents to orchestrate cross-platform tasks across many MCP servers (not limited just to OneSignal’s).

OneSignal AI, on the other hand, offers a chat interface to query and use the product, within the OneSignal UI.  It dramatically reduces the learning curve for using OneSignal because you can ask questions and perform actions with natural language abstracting many of the idiosyncrasies of the product.  Ultimately, it’s most powerful for automating work and reducing the labor overhead of using our product.

Why we're building this

More and more of our customers are incorporating AI into their daily workflows. We think that trend is only accelerating, and we want OneSignal to work naturally inside those workflows.

Whether that’s becoming fluent in agentic workflows arbitrated from your command line or a preference for interacting with software via natural language, we’re excited to use these breakthrough technologies to improve outcomes from our software and reduce the work our customers need to do to achieve those outcomes.

Next steps

It’s important to note that this are both early stage releases. We're actively developing and iterating based on real usage and feedback. If you're someone who's already building with AI tools day-to-day, we'd love for you to try it out and tell us what works and what doesn't.

Request Early Access to OneSignal AI →

Get started with the OneSignal MCP Server →

Here's a situation that plays out more often than most teams realize. A user qualifies for your welcome sequence, triggers an abandoned cart reminder, enters a re-engagement loop, and falls into the audience for a scheduled promotional blast. All on the same Tuesday.

Nobody designed that experience on purpose. Each campaign was built independently, probably by different people, with its own audience rules and its own trigger logic. They all work correctly in isolation. The problem is they all fire on the same user at the same time, and the user gets four messages in three hours from an app that now feels like it's yelling at them.

This is the cross-channel deduplication problem. It surfaces the moment you start running multiple automated messaging flows across push, email, SMS/RCS, and in-app, and it's one of the fastest ways to turn a good engagement strategy into a retention liability. The good news: it's fixable with the right approach to priority logic, exclusion rules, and frequency controls.

Why this problem gets worse as you scale

When you're running one or two campaigns, overlap is easy to catch manually. Someone on the team can eyeball the audience, check the send time, and make sure nothing collides. But that manual check stops working the moment you have a welcome sequence, a few behavioral triggers, a couple of journeys, and a regular promotional cadence all running at once.

The root cause is almost always the same: campaigns are built in silos. Marketing sets up the weekly promo blast. Product configures the onboarding sequence. Growth builds the abandoned cart flow. Each team is working in the same platform (or worse, in separate tools), but nobody has a unified view of what any individual user is going to receive on a given day. The campaigns don't know about each other.

This is also why the problem specifically surfaces with multi-channel messaging. A user might only get one email a day but still receive two push notifications, an SMS, and an in-app message on top of it. Channel-specific limits don't prevent cross-channel pileups unless you're managing frequency globally.

What over-messaging actually costs you

Four notifications in an hour changes how a user feels about your app. It stops feeling helpful and starts feeling aggressive. The typical response is predictable: they mute your notifications, unsubscribe from email, or uninstall entirely. Any one of those actions undoes weeks or months of engagement work.

This makes deduplication one of the most underrated mobile app retention strategies available. Most teams focus their retention efforts on what to send: better copy, better timing, better personalization. Fewer teams focus on what not to send, which is equally important. A user who receives two well-timed, relevant messages per day will stay engaged longer than one who receives five messages of varying quality and relevance.

If you haven't audited this recently, try it: pull a sample of active user profiles and trace every automated message they received over a seven-day window. The overlap patterns are usually eye-opening. For a broader look at over-messaging risks and how to address them, see 7 Easy Solutions to Mitigate the Risks of Over-Messaging.

Three ways to fix it

1. Set up priority queuing

Not all messages are equal. A password reset email matters more than a promotional push. An abandoned cart reminder matters more than a weekly newsletter. But without an explicit priority system, your platform treats them all the same and sends whichever triggers first.

The fix: define a message hierarchy and configure your send logic to respect it. A simple three-tier system works for most teams:

• Tier 1 (send immediately): Transactional and security messages. Password resets, order confirmations, payment failures.
• Tier 2 (send with spacing): Behavioral triggers. Abandoned cart, milestone achievements, feature adoption nudges.
• Tier 3 (send when clear): Scheduled campaigns. Promotions, newsletters, announcements.

The rule: when a Tier 1 message fires, suppress Tier 2 and 3 for a defined window (four hours is a reasonable starting point). When a Tier 2 message fires, delay any pending Tier 3 sends. This keeps your most important messages from being buried in noise and prevents users from receiving a promotional push five minutes after a transactional alert.

2. Use tag-based exclusion logic

Tags give you a way to coordinate across campaigns in real time. The idea: when a user takes a key action or enters a specific flow, tag them. Then configure other campaigns to check for that tag before sending.

For example: a user completes a purchase. Your system tags them "purchased_today." Your promotional campaigns are configured to exclude anyone with that tag for 24 hours. Your abandoned cart flow checks for the tag and exits the user immediately if they've already converted. Without this, a user might buy something and then get an abandoned cart reminder for the product they just purchased.

This kind of tag-based coordination is especially useful when you're running multiple journeys simultaneously. Each journey can tag users as they progress, and other journeys can read those tags to decide whether to send, wait, or skip. It turns independent flows into a coordinated system. For more on how tagging works in practice with automated journeys, see our Journeys documentation.

3. Set global frequency caps

Channel-specific limits are a start, but they're not enough on their own. Capping email to one per day doesn't help if the user also gets three push notifications and an SMS. Global frequency capping limits the total number of messages a user receives across all channels within a given window.

A reasonable starting point for most apps: no more than two to three total messages per user per day, across all channels combined. Transactional messages (order confirmations, security alerts) can be exempt from this cap since users expect them, but everything promotional or behavioral should count against it.

The key is that this cap operates at the user level, not the campaign level. Each campaign might be configured to send to its qualified audience, but the platform checks the user's recent message history before delivering and suppresses the send if they've already hit the cap. This requires your messaging tools to have a unified view of cross-channel activity, which is one of the core advantages of running your channels through a single multi-channel messaging platform rather than stitching together separate tools for each channel.

Why this only works on a unified platform

All three of these approaches (priority queuing, tag-based exclusion, frequency capping) depend on one thing: a single system that knows what every user has received across every channel. If your push notifications run through one tool, email through another, and SMS/RCS through a third, there's no central place to enforce global caps or check cross-campaign tags. You end up building custom middleware to sync state between systems, and that kind of duct tape breaks the moment someone adds a new campaign.

An omnichannel marketing platform that manages push, email, SMS, and in-app messaging from a single dashboard solves this structurally. It can evaluate a user's recent message history across all channels before deciding whether to send, suppress, or delay. The coordination happens automatically because all the data lives in one place.

OneSignal handles this by bringing all channels into one platform with shared user profiles, unified tagging, and a journey builder that can check for tag-based exclusions and respect frequency logic within the same flow. If you're running more than a few concurrent campaigns and finding that users are getting caught in the overlap, this is the structural fix. See how it works in the OneSignal platform overview.

The best message is sometimes no message

The instinct in growth marketing is to send more. More campaigns, more channels, more automations. But past a certain point, each additional message competes with your own previous messages for a user's attention, and the marginal return goes negative fast.

Deduplication logic isn't glamorous. But it protects the wins your other campaigns are generating by making sure users don't mute, unsubscribe, or uninstall because they felt overwhelmed. Get the priority queuing, exclusion tags, and frequency caps right, and every other campaign you run performs better because of it.

By Aleksandra Korzynska, CMO at Survicate — a leading customer feedback platform that allows businesses to collect, analyze, and act on customer insights efficiently.

The customers you are spending the most time on are most probably not the ones that are about to churn. This is an uncomfortable truth most teams don’t want to hear.

The reality is that your NPS detractors, low scorers, and the most vocal complainers are unhappy, but they still care, and are probably not the ones about to leave.

Most of the time, when people complain about a product, they want to get it fixed because they want to keep using it. Yes, they are angry, but they are still engaged enough to respond to your survey. Even better if they left a comment! This way you will have feedback to work with. This type of complaint needs to be shared with the relevant team and let be until the issue is fixed or you need more information.

The users who are truly about to churn stop engaging: no surveys answered, no emails opened, their sessions drop. And none of it triggers your standard retention alerts because none of it fits the category "this user is unhappy."

Many teams have built their entire retention playbooks around wrong signals. It’s time to dig deeper and fix it.

Lifecycle automations based solely on usage and behavioral data aren't enough for the maximum personalization uplift anymore.

Behavioral data tells you what, but not why

Most lifecycle marketers build automations based on behavioral first-party data. While true uplift  in personalization comes from blending behavioral events with zero-party data such as survey responses.

Using a tool like OneSignal provides you with a rich stream of behavioral signals. You can see who is or isn’t opening push notifications, who has gone dormant, and which user segments are uninstalling at a higher rate. Knowing all this, you can build re-engagement journeys and segment by activity levels.

But here's the gap: behavioral data tells you what is happening, but does not tell you why.

A customer who stops opening your push notifications could be overwhelmed and just taking a break, using a competitor instead, waiting on a feature they need and didn't see it launch or frustrated with something specific that behavioral data will never surface.

Each of these situations demands a completely different reaction. If you send a promo discount to someone who left because of a missing feature, you've wasted the moment. If you send a feature education sequence to someone who churned because of price, you're pushing them further away.

And yet, most teams build their lifecycle automations entirely on behavioral triggers, because this is the data that is easiest to collect.

McKinsey found that 76% of consumers get frustrated when brands don't personalize their experience. This research predates the current wave of AI-powered behavioral triggers. Customers now notice not just when you don't personalize, but how.

When messages feel like they're based on some assumptions rather than what they actually want right now, your customers will get frustrated. And if your incredibly sophisticated behavioral trigger flow annoys them, this means something is missing from the inputs.

Personalize based on what customers say, not what they do

What’s missing is the information customers give you intentionally, in their own words. Specifically, a short, well-timed survey triggered at the moment of drift.

At Survicate, we see this play out across hundreds of B2C apps and marketplaces. The most effective retention programs we've worked with are the ones that ask a simple question at the right moment:

"You haven't been active in a while. Why?"

Answers to this question give you something you can act on.

For example, if you are a freelance job platform and you see an increase in churn or inactive users, just one simple survey sent at the 30-day inactivity mark can help you reactivate your users. Without this you are left with 3 options (and I don’t know which one is the worst): send a generic re-engagement push, do nothing, or spray a batch campaign at the whole inactive cohort.

With a short survey, you can find out that users stopped engaging with your product because:

• There are not enough jobs in their area.
• Pay rates are too low.
• They no longer want to do freelance.

This changes your reactivation plan completely: you now can send a push about new jobs in the area to the 1st group and one about increased pay rates to the 2nd one.

If you were to send a generic push about more jobs in the area to the whole inactive cohort, you would annoy 2/3 of it and push them to quit.

Where does NPS sit in a churn-prevention system

To be clear: you shouldn’t get rid of NPS surveys; they still belong in your retention stack. It simply has to be one of several signals in the system, not as the primary trigger for retention action.

CRM platform Agentor, one of Survicate’s customers, ran a routine in-product NPS survey and got back something they didn't expect: a large share of users didn't know a core part of the product existed. Not that they'd tried it and disliked it, but they'd simply never found it.

Behavioral data showed normal-looking usage patterns, nothing that would flag a problem. The survey caught what the analytics missed entirely. They fixed the discovery path, mobile installs increased, and NPS followed.

Another example from Survicate’s customers, travel app WeSki, managed to maintain a steady NPS of 55 and grew 10x after nearly collapsing during COVID. Their feedback program isn't something groundbreaking: they send post-booking and post-experience surveys and route them into their CRM. But these 2 touchpoints give them continuous data on what is working at the moments that actually matter to their customers.

In both cases, their NPS didn't improve because of some magical workflow, it improved because the team was listening continuously across the journey.

How to fit NPS into your retention strategy:

Check relationship health. Use it to understand trends, segment your most loyal customers, and identify where your experience is falling short across the lifecycle. Promoters (9–10) deserve different conversations than detractors (0–6), but so do passives (7–8), who are the most at-risk of quiet churn. This passive segment is satisfied-but-not-delighted, and that feeling is exactly the one that leads to a quiet uninstall a few months later, when a better option comes along. But most probably you are missing them, because their NPS score doesn't raise an alarm.

Understand users’ engagement. Use it to spot drift early: declining sessions, lower notification engagement, feature abandonment. These are your pre-churn signals that should trigger a question, not a campaign.

Find the why with zero-party data. Use surveys as targeted questions triggered at the moment of friction: post-trial, inactivity, screen before cancellation, a low CSAT. This is your opportunity to collect the context and launch a useful and personalized response.

One more thing to keep in mind: the right data you collect today feeds your successful strategy tomorrow. Every survey response you capture and pass to your engagement platform is a signal your tools can use to make better decisions about channel, timing, message, and offer at a level that behavioral data alone cannot support. So interpreting this data wrong will cost you a lot.

3 survey triggers that can improve your retention

Here are examples from OneSignal + Survicate integration flow for you to start with.
If you are using the integration, set up Triggers and Surveys in Survicate, and Action — in OneSignal. If not, just replicate the logic in your retention flows.

1. The inactivity reason survey

Trigger: 14–30 days of user inactivity (adjust for your app's typical engagement cadence)
Question: "We noticed you haven't been active lately. What's the main reason?"
Options: [Tailored to your product e.g., not finding value, missing a feature, too busy, using an alternative]
Action: Segment by reason, send personalized push/in-app message that directly addresses the stated barrier

2. The post-trial conversion barrier survey

Trigger: Trial ends without upgrade
Question: "What's the main reason you didn't upgrade?"
Options: [Price, missing feature, not ready, didn't see value, using a competitor]
Action: Segment by barrier: price gets a discount or lower-tier offer; missing feature gets a roadmap update or workaround; "didn't see value" gets a targeted onboarding sequence.

3. The churn reason survey

Trigger: Cancellation initiated
Question: "What's the main reason you're canceling?"
Options: [Price, missing features, low value, switching to another tool, life circumstances]
Action: Even if they churn, you now have the data to win them back intelligently later. You also have aggregate intelligence about where your product is losing the fight which no amount of session recording will tell you.

Each of these should feed directly into your OneSignal platform, so that survey responses become attributes and events you can use to trigger personalized push notifications, emails, in-app messages, and SMS campaigns in real time.

Stop guessing. Start asking.

The companies that will win at retention in 2026 are the ones who combine behavioral signals with stated intent, who know not just that a customer is drifting, but why, and can respond to the actual situation.

Your push notifications can be smarter and your re-engagement journeys can be more relevant, but only if the data flowing into them reflects what your customers are actually telling you, rather than what they're silently doing.

The customers slipping away aren't the ones sending angry NPS scores. They are the quiet ones. And the best way to reach them is to ask a simple question before they make up their minds.

Survicate integrates with OneSignal to pipe survey responses directly into your user profiles and trigger automated journeys based on stated intent. See how it works →

Healthcare marketing sits at a unique intersection: you're trying to reach patients with timely, relevant messages, but the data you're working with is some of the most protected in the world. HIPAA (the Health Insurance Portability and Accountability Act) isn't just a box for your legal team to check. The messages you write, the segments you build, and the subject lines you test are all within scope.

Any information that ties a person's identity to their health condition, healthcare visits, or health payments is considered protected health information (PHI). That includes the obvious stuff (diagnoses, prescriptions, lab results) and things that might surprise you, like the fact that someone has an appointment at a specific provider, or that they're a patient at all.

The penalties for violations are real. Civil penalties can reach hundreds of thousands of dollars per violation, with annual caps in the millions. Willful misuse of PHI can also trigger criminal liability, including substantial fines and imprisonment. More practically: if a patient feels their privacy was violated by your message, you lose their trust permanently.

The good news is that HIPAA-compliant marketing is very doable. It mostly comes down to being thoughtful about what data lives where, what you put in message content, and how you've gotten permission to use it.

The five areas where marketing teams most often slip up

1. What shows up on the lock screen
Push notifications and SMS messages display on a phone's lock screen by default, which means anyone nearby can see them without unlocking the phone. The same goes for email subject lines and preview text, which surface in notification banners and inbox previews.

If your message content reveals anything about a person's health, you've disclosed PHI to anyone who happens to glance at the screen. Even well-intentioned reminder messages can be a problem if they're specific.

⚠ Watch out: Ask yourself before sending: If someone other than this patient saw this message, would they learn anything about the patient's health? If the answer is yes, rewrite it.

2. What you name your segments and tags
Building segments based on health attributes is normal and useful. The risk is in how you label them inside your platform. A segment named "Diabetes - Type 2" or a tag like diagnosis: depression means your engagement platform now contains plaintext PHI, visible to anyone with platform access.

The fix: use internal codes. Name the segment "Condition 214" and maintain a separate, secured mapping document that only authorized team members can access. Your messaging platform should never be a readable database of health conditions.

3. Message content that reveals a health context
This is where small word choices have big compliance consequences. Here are examples of the same message, written two ways:

The pattern is consistent: keep message content generic enough that an outside observer can't infer anything about the recipient's health. Sensitive details (results, diagnoses, specialist names) belong behind a secure, authenticated portal, not in a push notification.

4. Authorization, not just opt-in
Most marketing teams are used to consent-based opt-ins. HIPAA adds a specific requirement on top of that: any use of PHI for marketing must be covered by a valid written authorization: a dedicated document that specifically describes how health information will be used for marketing purposes.

⚠ Important distinction: A generic "I agree to receive communications" checkbox, even if prominently displayed, does not meet HIPAA's authorization standard. The patient needs to explicitly agree to the specific use of their health information for marketing. When in doubt, check with Legal before building your opt-in flow.

Plain language matters here especially. Patients who aren't native English speakers or who aren't familiar with legal terminology need to be able to understand what they're signing up for.

5. What data you send to OneSignal
Not all patient data belongs in your messaging platform. Before you configure any integration, categorize your data into three buckets:

• Non-PHI (safe to send): Things like membership tier, preferred channel, language preference, and app behavior. No restrictions.
• Operational PHI (send carefully, anonymize where possible): Appointment dates, provider first names, general care status. Needed to send useful messages, but handle with care.
• Sensitive PHI (never send to OneSignal): Diagnoses, lab results, prescriptions, billing details, insurance info. Keep these in your EHR. They should never leave it.

⚠ OneSignal BAA restrictions: OneSignal's Business Associate Agreement expressly prohibits ingesting the following PHI types into the platform: medical diagnoses, medical records, medical images, patient billing information, test results, genetic information, STI information, substance abuse information, and mental health information. If you're unsure whether your use case is acceptable, contact the Security & Legal team before configuring your integration.

Setting up OneSignal for HIPAA-compliant campaigns

OneSignal has completed a HIPAA risk assessment covering both the Security Rule and Privacy Rule, and offers Business Associate Agreements (BAAs) to Enterprise Plan customers. A BAA is a legal requirement before any third-party platform can process PHI on your behalf. To request a BAA or learn more, contact the OneSignal sales team.

Once your BAA is in place, here's how to configure your setup:

Use External IDs to unify patient profiles. Assign each patient an External ID so their push subscription, email address, and phone number are all linked to a single profile. Without this, the same patient on multiple channels could receive duplicate messages, or receive a message on a channel they didn't authorize.

Code your segments and tags. Use internal codes instead of plaintext health terminology in any segment names or user tags. "Condition 214" keeps sensitive data out of your platform. Maintain the key in a separate, access-controlled document.

Configure role-based access. HIPAA's minimum necessary standard means each person on your team should only have access to the PHI they need to do their job. Use OneSignal's team member roles to restrict who can view user data, import/export records, and manage API credentials. Note: Granular roles (Editor, Composer, Finance, Operations) require a Professional or Enterprise plan. Confirm your plan includes the roles you need before using this as a compliance control.

Keep sensitive details behind a secure portal. If a message relates to something sensitive (lab results, a care plan update, a billing statement), keep the notification generic ("You have a new secure message") and direct the patient to an authenticated portal to view the actual details after logging in. The message is the nudge. The portal is where PHI lives.

Pre-launch compliance checklist

Run through this before any healthcare campaign goes live:

• BAA in place. A signed Business Associate Agreement exists with OneSignal and every other vendor that processes PHI.
• Data categorized. You've mapped patient data into non-PHI, operational PHI, and sensitive PHI. Only the first two categories should go into OneSignal.
• Segments and tags anonymized. No plaintext health conditions, diagnoses, or treatment types in your platform.
• Message content reviewed. Every push, SMS, subject line, and preview text has passed the lock screen test.
• Authorization documented. Patient written authorization for marketing use of PHI is in place: explicit, separate from other authorizations, and in plain language. A generic opt-in does not suffice.
• Access roles configured. Team members only have access to the data and functions they need. Import/export is restricted to admins.
• Secure portal in place. Any sensitive health information is accessible only through an authenticated portal, not in message content.
• Audit and training schedule set. Regular risk assessments are on the calendar. Employees with PHI access are trained at least annually. There's a process for revoking access when someone's role changes.

Have questions about OneSignal's infrastructure or HIPAA setup?

If you have specific questions about how OneSignal's platform handles PHI, what our BAA covers, or how to configure your integration for your use case, contact the OneSignal team. We're happy to walk you through the technical and operational details of our infrastructure.

For questions about your own legal obligations under HIPAA, including authorization requirements, data handling policies, and compliance programs, please consult your own legal counsel. This guide is not legal advice.

This guide covers general best practices based on HIPAA's Privacy and Security Rules. It is not legal advice and does not constitute a legal opinion on your organization's compliance obligations. Every organization has unique data flows, risk profiles, and legal requirements. Always consult your own legal and compliance counsel before making decisions about how to handle patient data. HIPAA compliance is an ongoing responsibility, not a one-time setup task.

Your Shopify store already knows a lot about your customers. What they browse, what they add to cart, what they buy, what they abandon, how much they've spent over time. The problem is that this data usually stays locked inside Shopify, disconnected from the messaging tools you use to actually communicate with those customers.

OneSignal's Shopify integration, powered by our certified partner Vendo, closes that gap. It automatically syncs your Shopify customer data, storefront browsing events, and server-side order events into OneSignal, where you can use them to trigger messages, build segments, personalize content, and power automated journeys across push, email, SMS, and in-app.

Setup takes minutes and there’s no custom development required.

What data gets synced

The integration pulls data from two sources and sends it all to OneSignal as user tags and custom events, which means it's immediately available for segmentation, personalization, and journey triggers.

Customer tags

These are properties attached to each identified customer profile in OneSignal. They update automatically every hour:

• Name, email, and phone number
• Lifetime spend and total order count
• First and last order dates
• Email marketing consent status
• Shopify customer tags (any tags you've applied in Shopify)
  

Client-side events (real-time)

These fire in real time as customers browse your storefront, captured via the Shopify web pixel:

• Page viewed, product viewed, collection viewed
• Search submitted
• Product added to cart, product removed from cart, cart viewed
• Checkout started, checkout contact/address/shipping/payment info submitted, checkout completed
  

Server-side events (payment-verified)

These come from Shopify's backend and capture the full order lifecycle:

• Order received, fulfilled, and delivered
• Order refunded or partially refunded
• Abandoned checkout (syncs every hour)

For the full list of events and properties, see the Shopify integration documentation.

How to set it up

The setup is genuinely quick. There are four steps, and none of them require writing code or editing your Shopify theme:

1. Install the Vendo app from the Shopify App Store.

2. Connect your OneSignal account by entering your OneSignal App ID and REST API key in the Vendo dashboard.

3. Choose which events and tags to sync and configure your web push settings (prompt style, notification icon, etc.).

Once enabled, data starts flowing. Vendo also runs a historical backfill, so OneSignal receives your existing customer data and recent order history right away. You don't start from zero.

For the full setup walkthrough, see the OneSignal Shopify documentation.

4. Enable the Vendo theme block in your Shopify theme settings (Online Store > Themes > Customize > App embeds > toggle Vendo on). This handles SDK initialization, push prompts, user identification, and tag syncing automatically.

How customer identity works

An important detail: the integration uses an identified-only approach to user tracking. Anonymous visitors who haven't taken an identifying action are not tracked. A customer becomes identified (and starts generating data in OneSignal) when they do one of the following:

• Subscribe to web push notifications
• Sign up for your newsletter
• Log in to their customer account
• Complete a checkout

This approach is intentional. It prevents the duplicate user profiles and ensures that every profile in OneSignal represents a real, reachable customer. If a visitor subscribes to push and later logs in, Vendo merges the two identities automatically so you don't end up with separate profiles for the same person.

What you can build with it

Once your Shopify data is flowing into OneSignal, you have the building blocks for the campaigns and automations that drive real eCommerce revenue. Here are the most common use cases:

Abandoned cart recovery. Trigger a push notification or email within an hour of a customer abandoning checkout, personalized with the products they left behind. This is typically the highest-ROI automation you can set up, and the abandoned checkout event makes it straightforward to build as a Journey in OneSignal.

Post-purchase journeys. When an order is placed, kick off an automated sequence: a thank-you message, a shipping update when the order is fulfilled, a delivery confirmation, and a follow-up a week later asking for a review or cross-selling related products. Each step maps to a server-side event that's already syncing.

Win-back campaigns for lapsed customers. Segment customers whose last order date is more than 60 or 90 days ago and re-engage them with a targeted offer across email, push, or SMS. The last_order_date tag makes this segment easy to build and it updates automatically.

VIP and loyalty targeting. Use the total_spent and order_count tags to create segments for your highest-value customers, then send them early access to sales, exclusive offers, or personalized product recommendations.

Web push for your storefront. The integration deploys OneSignal's web push SDK on your Shopify store with no code. Visitors can subscribe to push notifications directly from your storefront, and you can start re-engaging them immediately, even if they never created an account or provided an email address.

Pricing

The Vendo integration is available on three tiers, all of which include the full set of data syncing, event tracking, and web push deployment:

You can purchase through the Shopify App Store or directly from OneSignal. The pricing is the same either way. Note that the Vendo integration fee is separate from your OneSignal plan. Additional usage fees may apply if your order volume exceeds your plan's allotment.

Get started

If you're already on OneSignal and running a Shopify store, you can have data syncing within the hour. Install the Vendo app from the Shopify App Store, connect your OneSignal account, and start building the campaigns your Shopify data has been waiting to power.

If you're new to OneSignal, create a free account and explore the platform before connecting your store. And if you have questions about the integration or want help planning your first campaigns, reach out to our team.

Subscription apps have a retention problem that most engagement strategies aren't built for. A subscriber who hasn't opened your app in a week isn't necessarily churning. They might be traveling, busy with work, or just in a natural down cycle. They're still paying and they’ll probably be back.

But most messaging strategies treat that week of inactivity like a five-alarm fire. The user gets a "We miss you!" push on day three, a discount email on day five, and by day seven they're in a win-back segment receiving the same treatment as someone who actually cancelled.

You end up annoying a paying subscriber, wasting a campaign on someone who wasn't at risk, and missing the chance to do something that would have actually brought them back sooner.

The better model for subscription apps is the engagement loop: a continuous cycle designed to bring users back through value, not pressure. Instead of waiting for inactivity and reacting to it, you build repeating patterns of triggers, actions, and rewards that make the app part of the user's routine. This post covers how those loops work, two examples you can implement, and how to tell when a loop is broken.

Why the linear lifecycle doesn't fit subscription apps

The traditional lifecycle model (onboard, activate, retain, win back) assumes users move in one direction. That works for a one-time purchase. It doesn't work for a subscription where engagement naturally rises and falls over weeks and months. A fitness app user who skips a week during a vacation isn't in the same category as one who lost interest. A news subscriber who reads heavily during an election cycle and lightly in the off-season isn't disengaging. (They're just being a normal human!)

When you force cyclical behavior into a linear model, you end up running win-back campaigns on active subscribers, triggering re-engagement flows for users who just got busy, and missing the smaller, more frequent interactions that actually build habits over time. For a broader look at where lifecycle marketing strategies fit into subscription apps, we've covered that separately.

How an engagement loop works

An engagement loop is a four-part cycle that repeats. Each completed cycle makes the next one more likely to happen, because the user's investment in the app grows over time.

1. Trigger. A personalized cue that prompts the user to return. This is usually a push notification, email, or in-app message, but the key is that it's timely and relevant to something the user actually cares about. A generic "Check out what's new!" is not a trigger. "A new episode of the series you're following just dropped" is.

2. Action. The thing you want the user to do when they return. This needs to be simple and have a clear path from the trigger. If the notification says "new episode," the tap should land them on that episode, not on the app's homepage.

3. Reward. The value the user gets from completing the action. This can be the content itself (watching the episode), a progress marker (extending a streak), a social validation (seeing how they compare to peers), or a tangible benefit (unlocking a feature). The reward has to feel proportional to the effort.

4. Investment. Something the user does that makes the app more valuable for next time. Rating content, saving a preference, completing a profile, following a creator. Each investment gives you better data for the next trigger, which makes the next loop more relevant, which makes the user more likely to complete it. This is where the cycle becomes self-reinforcing.

The loop breaks when any component is weak. A trigger that feels like spam, an action that requires too many taps, a reward that doesn't feel worth it, or no investment step at all. When the loop is working, each cycle deepens the user's relationship with your product. When it's broken, your messages train users to ignore you.

Two engagement loops you can build today

The content discovery loop
Best for: media, streaming, news, and podcast apps.

The goal here is to turn passive content consumption into an active discovery habit. Instead of waiting for users to browse on their own, you surface content that's relevant to their demonstrated interests and make it effortless to engage.

• Trigger: Push notification personalized to the user's viewing/reading/listening history. "A new true-crime series just dropped. Based on your history, you'll probably like this one."
• Action: User taps and lands directly on the content page. No homepage, no search required.
• Reward: The user discovers something they genuinely enjoy. The recommendation was right.
• Investment: The user follows the series, adds it to a watchlist, or rates it. That data makes the next recommendation better.

Where this breaks: generic recommendations. If the trigger doesn't reflect what the user actually watches, it feels like noise and they stop tapping. The quality of the recommendation engine is the quality of the loop. This kind of personalized, event-driven trigger can be built as an automated flow using a visual journey builder, with user behavior data powering the content selection.

The habit formation loop
Best for: fitness, wellness, education, language learning, and productivity apps.

The goal is to embed your app into a daily or weekly routine by creating a pattern that feels rewarding to maintain and costly to break.

• Trigger: A smart reminder at the user's optimal time. "It's 7 PM. Ready for your 10-minute Spanish lesson? You're on a 15-day streak."
• Action: User opens the app and completes their daily session.
• Reward: The streak extends, progress is visible, and the user feels a sense of accomplishment.
• Investment: The completed session is logged. The streak gets longer and harder to abandon. Lesson history informs what comes next.

Where this breaks: timing and tone. A reminder that arrives at 6 AM on a Saturday or uses guilt-trip language ("You're falling behind!") creates resentment, not motivation. Send-time optimization that delivers reminders when each individual user is most likely to engage makes a measurable difference here. For more on turning messages into long-term loyalty drivers, we've written about that in depth.

How to tell when a loop is broken

A loop that looks healthy at the top (good push tap rates) but isn't improving retention is broken somewhere downstream. Here's how to diagnose where:

Triggers are working but actions aren't completing. Users tap the notification but don't finish the intended action. The problem is friction between the trigger and the payoff. Check whether the deep link lands users in the right place, whether the action requires too many steps, or whether the in-app experience doesn't deliver on what the notification promised.

Actions are completing but users aren't coming back. The reward isn't strong enough or there's no investment step. If a user watches the recommended episode but has no way to follow the series, save it, or signal their preferences, there's nothing pulling them into the next cycle. The loop completes once and then dies.

Return rates are good but frequency is declining. The trigger is losing relevance over time. This usually means the personalization is static. If you're sending the same type of recommendation week after week without incorporating new behavioral data, the novelty wears off. The investment step should be feeding new signals back into the trigger logic.

Tracking this requires visibility into each stage of the loop, not just top-line metrics like DAU or push open rates. A customer engagement platform with real-time event tracking and engagement analytics lets you measure where users are dropping off within the cycle, not just whether they're active or inactive.

What your platform needs to support engagement loops

Not every messaging tool can power engagement loops well. The difference between a basic notification sender and a platform built for loops comes down to a few specific capabilities:

• Behavioral event triggers. Can the platform fire a message the moment a user completes (or fails to complete) a specific in-app action? If triggers only support time-based scheduling, you can't build responsive loops.
• Journey automation with branching. Can you build a flow where the next message depends on what the user did after the last one? A loop needs conditional paths: if the user completed the action, reinforce it. If they didn't, try a different approach or channel.
• Cross-channel delivery. Can you reach users on push, email, SMS, and in-app from the same journey? A loop that only works on one channel breaks the moment a user's attention shifts.
• Personalization from behavioral data. Can message content dynamically reflect what the user has done recently? Generic messages kill loops. Messages that reference specific user actions sustain them.

OneSignal supports all of this through its Journeys builder, custom events, and cross-channel messaging across push, email, SMS, and in-app. One in every five new apps uses OneSignal to build and automate their engagement strategies. If you're designing loops for a subscription app, the subscription app campaigns guide is a useful companion to this post.

Build for loops, not lines

Subscription apps don't follow a straight line from onboarding to loyalty. Users cycle through periods of high and low engagement, and the apps that retain best are the ones that design for those cycles rather than panicking every time a user goes quiet for a few days.

Start with one loop. Pick the behavior you most want to reinforce (daily usage, content discovery, feature adoption), map out the trigger, action, reward, and investment steps, automate it, and measure where users drop off. Then tighten the weak link and run it again. That iterative process, loop by loop, is how subscription apps build the kind of retention that compounds over months and years.

Explore OneSignal's use cases for subscription apps to see what's possible or try OneSignal for free to get hands on.

RCS has had a long runway. The protocol has been in development for over a decade, and for most of that time it was held back by fragmented carrier support and Apple's absence from the standard. That’s since changed. Apple has adopted RCS, joining Android, and the result is that RCS messaging is now available on the vast majority of smartphones worldwide.

For brands that have been running their mobile messaging on SMS, this raises a practical question: is the upgrade worth it now, or is it still too early? Below, we break down where RCS stands in 2026, which use cases are already delivering clear ROI, what limitations still exist, and what a realistic migration looks like for a team that's currently SMS-only.

Where RCS stands in 2026

RCS upgrades the native messaging app on a user's phone into something closer to an app experience. Instead of plain text with a character limit, brands can send messages that include verified sender profiles (brand name, logo, and checkmark), high-resolution images and video, interactive product carousels, and tappable action buttons like "Track Package" or "Complete Purchase." Messages are delivered over data or Wi-Fi rather than the legacy cellular network that SMS relies on.

The tipping point was Apple's adoption of the standard, which effectively made RCS available across both major mobile platforms. That's what moved it from "promising but fragmented" to "production-ready." Our research shows that 45% of mobile teams are actively exploring or planning to adopt RCS this year, according to The 2026 State of Customer Engagement Report.

For a deeper look at the technical differences between the protocols, see our breakdown of the signs you're ready to upgrade to RCS.

Where RCS is already delivering ROI

The strongest case for RCS comes from use cases where the interactive capabilities directly translate into measurable business outcomes.

Interactive commerce

This is where the ROI case is most clear-cut. An abandoned cart message sent via RCS can include a carousel of the exact products a customer left behind, with images, prices, and a one-tap "Complete Purchase" button. Compare that to an SMS that says "You left items in your cart! Visit [link] to check out." The RCS version eliminates friction by letting the customer act without leaving the conversation, and early adopters are reporting click-through rates 3 to 7 times higher than equivalent SMS campaigns.

The same applies to product launches, flash sales, and personalized offers. Rich cards with product imagery and direct action buttons consistently outperform plain text with a shortened URL.

Transactional and service messages

RCS makes utilitarian messages more useful. An airline can send a boarding pass with a QR code and a "View Flight Status" button. A healthcare provider can send an appointment reminder with an "Add to Calendar" action. A retailer can send a delivery notification with a real-time tracking button. All of these interactions happen inside the messaging app without requiring the customer to open a browser, download an app, or navigate a website. That reduction in friction compounds across millions of messages.

Brand trust and verification

In a messaging environment increasingly cluttered with spam and phishing attempts, the verified sender profile is a meaningful differentiator. When your brand name, logo, and verification badge appear in the message thread, the recipient knows the message is legitimate before they even read it. This visual trust signal improves open rates and sets RCS apart from the anonymous short codes and phone numbers that SMS relies on.

What to know before you commit

RCS is production-ready, but there are a few practical factors to plan around.

Not every user can receive RCS yet

Adoption is broad but not universal. Older devices and some regional carriers don't support RCS, which means you need a reliable SMS fallback for any message you send. The good news is that any modern engagement platform handles this automatically: if the recipient's device doesn't support RCS, the message downgrades to SMS or MMS without manual intervention. You don't have to choose between channels. You build for RCS and fall back gracefully. Apple's initial RCS implementation also has some feature gaps compared to Android's (most notably around encryption in one-on-one conversations), but these are edge cases that don't materially affect business messaging use cases.

Per-message costs compared to SMS

RCS messages cost more to send than SMS. That's a real line item to account for. But the comparison that matters isn't cost per message. It's cost per conversion. If an RCS abandoned cart message converts at 3 to 5 times the rate of an SMS version, the higher per-message cost pays for itself several times over. The right approach is to start with your highest-value campaigns (cart recovery, promotional offers, booking confirmations), measure the lift, and expand from there based on data.

Verification takes time

Brands can't self-register for RCS. The verified sender profile requires an application process through a messaging aggregator, and approval typically takes 8 to 16 weeks. This is worth factoring into your planning timeline. If you're targeting a major campaign or seasonal push, start the verification process well in advance.

For tips on navigating this, see our guide on how to streamline the RCS approval process.

The migration path for SMS-only brands

Start with a platform that supports both

The most important decision is choosing an engagement platform that supports RCS and SMS natively, with automatic fallback built in. This means you're not ripping out your SMS infrastructure. You're adding RCS on top of it and letting the platform handle delivery logic based on each recipient's device capabilities.

OneSignal supports RCS alongside push, email, SMS, and in-app messaging, which means you can coordinate RCS campaigns with your existing channels through a single journey builder rather than managing a separate tool.

OneSignal also handles the aggregator coordination and verification process, so your team can focus on the campaign strategy rather than the carrier paperwork. For more on how OneSignal is investing in RCS, read why RCS matters now.

Phase it in with A/B tests

Most teams phase RCS in alongside SMS rather than replacing it overnight. Pick one or two high-value campaigns where interactivity will make a measurable difference (abandoned cart recovery is the obvious starting point), run A/B tests comparing RCS and SMS performance, and use the results to build the internal business case for expanding to more use cases. This is how you prove ROI with real data from your own audience rather than relying on industry benchmarks.

For a broader look at building an SMS and RCS strategy that delivers results, we've published a separate guide on that topic.

The verdict

RCS in 2026 is no longer a bet on future potential. The device support is there, the feature set is mature, and the use cases with proven ROI are well established. The remaining limitations (cost, carrier gaps, verification timelines) are real but manageable, and none of them are reasons to wait.

For brands still running on SMS-only, the practical move is to start with your highest-conversion campaigns, measure the lift RCS provides, and scale from there. The brands that adopt now will have verified sender profiles, optimized creative, and performance data while their competitors are still evaluating whether to start.If you're ready to add RCS to your messaging strategy, explore OneSignal's RCS capabilities or get in touch with our team to start the verification process.

Most teams start their search for a customer engagement platform by comparing feature lists. That makes sense on the surface, but it's often the wrong starting point. The factor that actually determines whether a platform works for your team is how well it integrates with your existing data infrastructure and how quickly your team can get value out of it.

A platform with an impressive feature set that takes months to implement, requires dedicated engineering resources to maintain, and locks key capabilities behind enterprise pricing tiers will cost you more than the subscription fee suggests. That's why more teams are evaluating Braze alternatives based on practical fit rather than feature ceiling.

This guide walks through how to match a customer engagement platform to your data stack, your team's resources, and the level of complexity you actually need.

Why your data stack should drive your CEP decision

A customer engagement platform is the system that unifies customer data and manages communication across channels: email, SMS, push notifications, in-app messages, and more. But the platform is only as useful as the data it can access and the speed at which your team can act on it.

When your CEP and data infrastructure don't connect cleanly, you end up with fragmented user profiles, siloed channel data, and an inability to build the kind of unified customer journeys that actually drive retention. And when a platform doesn't integrate easily with your existing tools, your team spends more time managing the platform than using it.

Matching a platform to your data infrastructure

Most organizations fit into one of three data architecture patterns. Each one has different requirements from a customer engagement platform, and the right choice depends on which pattern describes your team.

The data warehouse-centric stack

Some large organizations run everything through a data warehouse like Snowflake or Redshift, with dedicated data engineering teams managing the pipelines. Teams in this position sometimes gravitate toward heavyweight platforms that offer deep, native warehouse integrations.

But it's worth asking whether you actually need that level of architectural coupling. Warehouse-native CEPs tend to come with long implementation timelines (often months, not weeks), high engineering maintenance costs, and interfaces that require technical resources to operate day-to-day. For many teams, the data they actually use for messaging, behavioral events, user attributes, and channel engagement, can be piped into a more flexible platform through standard integrations or APIs without rebuilding their entire data flow around the CEP.

If your team has the budget, the headcount, and the timeline for a full enterprise implementation, that path exists. But a growing number of teams with warehouse-centric stacks are finding that a platform with strong API connectivity and out-of-the-box integrations covers their actual use cases at a fraction of the cost and implementation time.

The CDP-driven stack

Many modern teams use a Customer Data Platform or product analytics tool like Segment or Amplitude to centralize user data and event streams. For this setup, the key requirement is a CEP that integrates with your CDP in real time and lets your marketing team act on behavioral data without filing engineering tickets.

Look for platforms with native, two-way integrations that allow marketers to build segments and trigger campaigns based on behavioral data from their analytics tools. The CEP should function as the action layer for insights your CDP surfaces, not as a second data warehouse that requires its own complex configuration.

OneSignal integrates with Segment, Amplitude, Mixpanel, and other major data platforms, allowing teams to connect their existing analytics infrastructure and start building cross-channel campaigns on top of it. This is the kind of setup where measuring outcomes and proving campaign ROI becomes straightforward, because the data is already flowing.

The API-first and lean stack

Startups and mid-market companies typically run a mix of third-party services and internal tools, with limited (or no) dedicated data engineering resources. For these teams, the most important thing a CEP can offer is a fast path to value: a well-documented API, stable SDKs, and an interface that marketers can use without developer support for every campaign.

This is where the gap between high-overhead platforms and more agile alternatives becomes most obvious. A platform that gates its API behind premium tiers, requires months of engineering time to implement, or needs a dedicated admin to operate is going to slow down exactly the teams that can least afford to wait.

OneSignal was built for this profile. The API is fully accessible on every plan, SDKs are well-documented and actively maintained, and the dashboard is designed so that marketers and developers can both work in it without stepping on each other. For teams where technical flexibility and speed matter, that combination is hard to beat.

What else to evaluate

Implementation time and total cost

The subscription fee is the number you see. The implementation cost is the number you feel. Some high-overhead platforms take three to six months to fully implement, during which time your engineering team is pulled away from product work and your marketing team is waiting to launch campaigns they could have been running on a faster platform from week one.

When evaluating Braze competitors, ask how long it takes to go from signing the contract to sending your first production campaign. If the answer is measured in months, factor that delay (and the engineering hours behind it) into the real cost of the platform. OneSignal's setup process is designed to get teams live in days, not quarters.

Cross-channel orchestration

A customer retention platform should let you treat email, push, SMS, and in-app as parts of a single journey, not as separate tools with separate dashboards. The ability to build automated sequences that adapt based on real-time behavior (send a push, wait to see if it's opened, fall back to email if not) is what separates a messaging tool from an engagement platform.

If you're currently managing channels through separate systems, you're leaving coordinated, cross-channel engagement on the table. This is an area where a unified platform delivers compounding value over time.

Journey building and automation

Building complex, behavior-driven journeys used to mean committing to a platform designed for teams with dedicated engineering resources and the budget to match.. That's less true than it was a few years ago. Modern alternatives now offer visual journey builders with branching logic, behavioral triggers, wait conditions, exit rules, and cross-channel messaging steps that cover the vast majority of real-world use cases.

Before assuming you need the most complex (and expensive) journey tool on the market, map out the actual automations you'd build in your first 90 days. Most teams find that their requirements, abandoned cart recovery, onboarding sequences, re-engagement flows, post-purchase nurture, are well within the capabilities of a platform like OneSignal, which offers all of these through its Journeys builder.

How to choose the right Braze alternative

Rather than starting with a platform and working backward, start with your own team's profile:

• What does your data infrastructure look like? If you're running Segment, Amplitude, or a similar CDP, prioritize platforms with native integrations that let your marketing team act on that data directly.
• How much engineering time can you commit? If the answer is "not much," cross off any platform that requires a multi-month implementation or dedicated technical admin.
• What channels do you actually use? If mobile push, web push, and in-app messaging are core to your strategy, make sure the platform excels at those channels natively rather than treating them as add-ons.
• What's your real budget, including implementation? A platform that costs less on paper but takes twice as long to implement may not be cheaper in practice. Factor in the full cost of getting to production.
  

For teams that are mobile-first, value speed-to-value, need strong cross-channel capabilities, and want a platform that both developers and marketers can use effectively, OneSignal consistently comes out ahead in these evaluations. It's why the platform is one of the most widely adopted Braze alternatives on the market today.

Remember… the right platform is the one that fits

The best customer engagement platform is the one your team can actually use to its full potential, not the one with the longest feature list or the biggest brand name. Evaluate candidates on integration fit, implementation speed, total cost, and how quickly your team can get to production campaigns. The answers to those questions will narrow the field faster than any feature matrix.

Explore OneSignal's platform features to see if it's the right fit for your stack, or see how OneSignal compares to Braze directly.

You've been in this meeting before…

Retention is slipping, someone pulls a cohort of users who haven't opened the app in 30 days, and the team puts together a win-back campaign. A push notification, maybe an email with a discount attached, send it out, see what happens. A few users trickle back, but most don't. The campaign gets marked as complete and everyone moves on.

Elephant in the room: 30 days of inactivity is a long time, and you may need to adjust expectations if you think (essentially) cold-calling someone with a 15%-off code is enough to get them back through the door. Sometimes it is! But most of the time it buys you one session and not much else.

The teams with the strongest retention numbers treat re-engagement as an always-on system. One that’s always watching for the behavioral signals that precede churn, and always responding before the user would ever land on a 'lapsed' list.

Does this look like your win-back campaign?

A static win-back campaign has a fixed structure. A user crosses a threshold (usually days since last session), they enter a flow, they receive a predefined sequence of messages, and then the flow ends.

Everyone in the segment gets roughly the same treatment regardless of how they used the product, which channels they prefer, or what drove their disengagement in the first place.

The typical setup:

• Trigger: user inactive for 14/30/60 days
• Message 1: "We miss you" push notification
• Message 2: Follow-up email with a coupon or feature highlight
• Message 3: Maybe an SMS if the first two didn't land
• End of flow

There are two structural problems here. First, the trigger fires after the user has already disengaged, which means you're always playing catch-up. By the time someone has been inactive for 30 days, the relationship has cooled significantly and the cost of re-acquisition is high.

Second, the flow doesn't adapt. A gaming user who stopped playing because they hit a progression wall gets the same message as someone who churned because they never discovered the social features. A food delivery customer who only orders during football season gets the same "come back" email as someone who was ordering weekly until last month.

This system is structurally incapable of responding to the specific reason a user is cooling off, because it wasn't designed to detect that in the first place.

Building a re-engagement system that responds before users churn

A dynamic re-engagement system doesn't wait for the 30-day mark. It starts responding to cooling signals much earlier, and it adapts its approach based on what kind of user is cooling and why.

Here's what that looks like in practice and what your tooling needs to support it:

It monitors behavioral patterns, not strictly inactivity thresholds.

Instead of triggering on "last session was 30 days ago," it watches for changes in engagement patterns. A power user who normally logs in daily but hasn't opened the app in 3 days is a very different signal than a casual user who visits once a week and missed a week. A dynamic system treats those differently because the appropriate response is different.

To pull this off, you need segments that update dynamically based on behavioral data, not static lists. That means tagging users with properties like last session date, session frequency, and feature usage, then building segments that automatically shift users between engagement tiers as their behavior changes. Most modern engagement platforms support this. In OneSignal, segments recalculate automatically as tag and event data updates, so a user who goes quiet moves into your "cooling off" segment without anyone pulling a manual report.

It branches based on user context.

When a user enters a re-engagement flow, the system checks what they were doing before they cooled off. Were they active but not converting? Did they hit a paywall and bounce? Are they a seasonal user who always drops off this time of year? Were they engaging with the product but ignoring your messages? Each of those paths calls for a different message, a different channel, and a different cadence.

The practical requirement here is a journey builder with yes/no branching that can check segment membership or message interaction data at each step. You set up the flow once with conditional paths for each user context, and the system routes people down the right branch automatically. The entry trigger can be a custom event (like "paywall_viewed" or "cart_abandoned") or segment-based (like "was active 14 days ago, hasn't returned"), and each branch gets its own messaging sequence.

It respects channel preferences as data.

If a user hasn't opened a push notification in two weeks but clicked your last email, a static campaign might still lead with push because that's what the flow specifies. A dynamic system routes the re-engagement message through the channel the user actually responds to. It might start with a push, wait to see if it's opened, and fall back to email or in-app only if needed. The channel sequence adapts to the user rather than running the same playbook for everyone.

This requires two things from your tooling. First, all your channels need to live under one roof so the system has a unified view of how each user engages across push, email, SMS, and in-app. Second, your journey flows need wait steps and branch logic that can check "did the user open this push?" before deciding whether to send the follow-up via email instead. If your push provider and email provider are separate tools with separate dashboards, this kind of cross-channel sequencing is painful enough that most teams just don't do it.

It knows when to stop.

This might be the most underrated difference. Static campaigns end when the sequence runs out. Dynamic systems include explicit cooling-off logic: if a user completes the full flow without re-engaging, they're tagged and suppressed from promotional sends for a defined period. No more stacking three different re-engagement campaigns on top of each other because each one operates independently. The system recognizes when further outreach is doing more harm than good and backs off before it pushes the user from "inactive" to "unsubscribed."

The mechanics here are straightforward but easy to skip: tag users at key points in the journey (e.g., "winback_attempted"), set exit conditions that pull users out when they re-engage or when the sequence completes, and use re-entry rules to prevent someone from cycling back through the same flow too soon. The tagging piece is especially important because it's how you coordinate across multiple journeys. Without it, your abandoned cart flow, your feature adoption flow, and your win-back flow can all fire on the same user in the same week without any of them knowing about the others.

What changes when you build this kind of system

The shift from static campaigns to dynamic systems changes the shape of your entire retention strategy, not just your re-engagement metrics.

When you have a system that responds to early cooling signals, your win-back campaign doesn't need to do as much heavy lifting because fewer users reach the point of full disengagement. The users who do churn are the ones your system already tried to reach through multiple channels and approaches, which means the remaining churn is more likely to be genuine product-fit issues rather than messaging failures you could have prevented.

Your channel health improves too. A dynamic system that monitors opt-out velocity and routes users to preference centers before they unsubscribe protects your reachable audience in a way that static campaigns can't. You stop losing channels you can't get back.

And your team spends less time building one-off campaigns. Instead of launching a new win-back flow every quarter with slightly different copy, you're maintaining a system that runs continuously, learns from engagement data, and routes users through the appropriate path automatically. The upfront investment is higher and the ongoing operational cost is lower.

Ready to put this into practice?

The concepts above are deliberately framework-level. The harder (and more valuable) work is translating them into specific flows for specific retention scenarios: the gaming app whose D30 retention is falling apart, the eCommerce brand whose users opt out faster than they can grow the list, the subscription app that loses people at the paywall, the food delivery service whose seasonal users never become year-round customers.

We've been building out exactly those flows in detail. Our upcoming Breaking the Churn Loop webinar brings together leaders from Crocs and Ringier to share the exact journey flows, channel timing, and branching logic behind retention strategies that actually compound. If the patterns in this post feel familiar, the webinar is where we get into the specifics of how to build the system.

Save your seat and prepare to break the churn loop →

When your product is invisible until it breaks, how do you build loyalty?
At a recent Surge event in Singapore, Terry Williams, who leads My Republic's consumer business across sales, marketing, products, and customer service, sat down for a fireside chat about how a broadband provider in one of the world's most price-competitive markets is rethinking customer engagement from the ground up.

The conversation was packed with real examples, honest failures, and a refreshingly unconventional playbook. Here are the key takeaways.

1. If your product is invisible, create new reasons to show up

Broadband is the definition of a "set it and forget it" service. Customers sign up, and the next time they hear from their provider is usually a renewal notice, or an outage. Terry was blunt about this: most telcos leave a blank space between installation and recontract. "They're going to be like, 'Okay, you're in'... and then blank space… and then 'please give us your money again.'”

My Republic decided to fill that gap with customer touchpoints that actually provide value: community events, a loyalty program, in-app experiences, and even a physical retail store. The goal is to make the brand part of a customer's life, not just their monthly bill.

2. Start with a niche, then expand adjacently

My Republic built their initial identity as "the gamer ISP", a positioning so strong that 85% of gaming cafes in Singapore now run on their network. But rather than staying put, they expanded into gamer-adjacent spaces: trading card games, esports events, and collectibles.

Singapore is the second-largest market in the world for trading card games per capita, behind only Japan. My Republic validated that overlap through surveys, customer calls, and conversations at gaming cafes and events, and found that their gamer customers were deeply into TCGs as well. That insight led to Card Arena by My Republic, a trading card game hub at Suntec City that's become one of the most successful card shops in Singapore, and Card Con, a convention that drew 15,000 attendees across a single weekend.

3. The installation is your first date, don't blow it

Terry compared the broadband installation experience to a first date: "If you have a bad installation experience, you're probably not going to stay married for long."

Most providers treat installation as a logistical step. My Republic treats it as the most important moment in the customer relationship. That means pre-installation education so customers know what to expect, chatbot support and video guides for self-troubleshooting, and fail-safes, like providing a temporary Wi-Fi unit if the connection doesn't come up on time.

4. Know your customer… or end up with 350 unsold hats

The most memorable moment in the conversation was a candid failure story. At Card Con 2, My Republic invested in branded merchandise: custom shirts, tote bags, mouse pads, and 350 hats designed by a professional artist. Almost nobody bought them.

"I have 350 hats. If anyone wants one, come by the office. Seriously, you can have a hat."

The lesson? You can't skip the customer research step. Data and surveys are useful, but they're not enough. You have to go to the events. You have to call customers. You have to get your product managers talking to real people. "If you're just living inside your own office and asking all your co-workers what they like, they're going to be very different demographics."

5. Build engagement that's a value exchange, not a giveaway

Terry drew a sharp contrast between generic perks ("here's a free ticket to a movie or a free thing for Sephora") and engagement that's actually tied to what your customers care about. The former is expensive and only resonates with maybe one in ten people. The latter builds real loyalty.

My Republic's approach: if you're a broadband or mobile customer, you jump to level seven in their new tiered loyalty system. That unlocks discounts on trading cards at Card Arena, free spins in their mobile games, early access to events, and other benefits their core audience genuinely values. They even built Paka Rocket Adventures, a mobile game that now has 10,000 daily active users, with in-game rewards tied to their loyalty ecosystem.

The result is that customers don't mind paying a small premium for My Republic because they're getting tangible value back in a space they care about.

6. Community is the moat when price can't be

Here's the competitive reality: in Singapore, you can get 1 Gbps broadband for around $30 a month. In Canada, that same speed costs roughly $100. Singapore is one of the cheapest broadband markets in the world, and mobile pricing follows a similar pattern.

When everyone is competing on price, the only path to survival is differentiation. Terry was direct: "If we didn't do this, we would die. There's no way around it. If everybody's competing on price, there's just going to be consolidation."

The community strategy is paying off. My Republic is one of the only non-new-entrant broadband companies in Singapore that's still growing. Sales are trending up, churn has come down significantly, and Card Arena's revenue is contributing meaningfully to their P&L. They're shifting ad spend into community investment, bringing down cost per acquisition in the process.

7. If you can only fix one thing, fix how well you know your core customer

When asked for a single piece of advice for anyone running a subscription business, Terry's answer was simple: know your loyal core.

"Figure out who your loyal core is. The people who are interacting, giving you feedback, adding comments on social media, recommending your company. Get in touch with them, find out what they really like. Then when you're building engagement campaigns, make sure whatever it is is good for that core person. Don't do something that's vanilla for everybody."

The payoff? "You won't be staying up at night because you're worrying about churn. You'll always have them to carry you through."

The bottom line

My Republic's story is a case study in what happens when a company stops treating customer engagement as a retention tactic and starts treating it as the product itself. They've gone from broadband provider to community builder, to the point where new customers are discovering them through trading cards, not the other way around.

As Terry put it: "People are already starting to see us. They're like, 'Oh, you guys are the trading card company.' And we're like, 'No, we're actually a broadband company.' They're like, 'Oh, we didn't know you had broadband.' And we're like, 'Well, do you want to sign up?'"

It's working exactly how it's supposed to.

Whether you're filling the gap between signup and renewal or building loyalty touchpoints your customers actually look forward to, it starts with being able to reach them on the right channel at the right time. OneSignal makes that part straightforward.

The first hour on any new customer engagement platform is usually wasted. You poke around the dashboard, open a few settings screens, maybe start configuring something before realizing you needed to do something else first. Here's a faster path: the specific setup steps that matter most, in the order you should do them.

This guide provides a focused, 60-minute action plan to cut through the noise. We will prioritize the critical first steps to configure your platform correctly and launch your first automated message. Most teams complete this initial setup in under an hour, building a solid foundation for the sophisticated strategies that follow.

Successful implementation of a customer engagement platform hinges on deliberate, prioritized actions. By focusing on data integrity and a single automated workflow first, you can demonstrate value immediately and build momentum for your team.

The first 15 minutes: Foundational setup & data integrity

This phase is non-negotiable. Every future act of segmentation, personalization, and messaging will be built upon the foundation you lay down right now. Get this right, and you set your entire engagement strategy up for success.

Action 1: Configure your user identity system

Your first and most critical action is to tell the platform how to recognize a unique user across every device and channel. A true customer engagement platform is designed to manage and optimize the end-to-end customer journey across all touchpoints. This process starts with a unified user profile, which prevents duplicate data and ensures every interaction contributes to a single source of truth for each customer.

To do this, map your internal user ID from your database to the platform's "External ID." A detailed quickstart guide can walk you through the precise process. Using a stable, non-personally identifiable unique ID is a firm best practice that will save you from major data cleanup headaches down the road.

Action 2: Install the SDK & connect your first channel

Next, you must install the Software Development Kit (SDK) to build the bridge that allows data to flow from your digital properties into the platform. This is what enables message delivery.

For a website, begin with the Web SDK; for a mobile app, choose the appropriate mobile SDK. To gain immediate momentum, focus on configuring just one or two primary channels, such as push notifications or in-app messaging. The goal is to centralize conversations from your most important channel first. You can easily integrate other channels like email and SMS later, once this core setup is validated.

Minutes 15-45: Activating your first automated journey

With your foundation secure, it is time to shift from passive configuration to active engagement. You will now build and launch a simple, high-value automated workflow that showcases the platform's power and delivers an immediate return on your time.

Action 3: Choose a high-impact welcome journey

Start with a new user welcome journey. This is a classic low-effort, high-impact workflow that engages every new user at the peak of their interest. It improves onboarding, sets the tone for future communications, and delivers measurable results.

According to the 2026 State of Customer Engagement Report, push notifications in the first 30 days after an install drive a 75.8% impact on long-term retention. Furthermore, behavior-triggered messages like a welcome journey outperform standard broadcasts by up to 9x, making them essential for cutting through the noise.

The journey's logic is beautifully simple:

• Trigger: A user signs up or opens the app for the first time.
  
• Action: Send a single, welcoming push notification or in-app message.

Action 4: Build the workflow

Modern platforms provide visual journey builders that let you map out this entire workflow without writing a single line of code. For now, resist the temptation of complex, multi-branch logic. The goal is to establish a clear baseline you can optimize later.

Create a new journey, select the "new user" trigger, and drag a message block like "Send Push Notification" onto the canvas. Your objective here is to drive value with a simple workflow and achieve a quick, motivating win.

Action 5: Write your message (With an assist from AI)

Staring at a blank message composer can be intimidating. This is where an AI message composer becomes your creative co-pilot. It is one of the most powerful user engagement tools for apps available today, helping you craft clear, compelling copy in seconds.

Use the AI tool to generate ideas with a simple prompt, such as: "Write a friendly 15-word push notification welcoming a new user to our app and hinting at the value they'll find." Review the generated options, pick the one that best captures your brand's voice, and give it a final polish.

Minutes 45-60: Test, launch, and plan for day two

In this final stretch, you will validate your work, activate the journey for real users, and define a clear path forward. This ensures you leave your first hour with a tangible accomplishment and a roadmap for what comes next.

Action 6: Test and verify

Never launch blind. A broken link or poorly rendered message can sour a customer's first impression. Before activating the journey for everyone, it is crucial to send and test your notification by targeting a test device or a user segment containing only your own ID. Take a moment to see how the message looks in a real-world notification center.

Run through this simple pre-flight checklist:

• Did the message deliver successfully?
  
• Does it look and read exactly as intended on the device?
  
• Does the click-through action lead to the correct destination?

Action 7: What to defer for later

To protect your focus, knowing what not to do in your first hour is just as important as knowing what to do. Conclude your session by parking advanced features for another day. Mastering the fundamentals is the key to unlocking more powerful tactics down the line.

Save these items for after your initial setup is live and validated:

• Complex multi-channel messaging platform workflows, such as the sophisticated sequences needed for creating a strong messaging campaign across multiple channels.
  
• Intricate A/B testing with multiple message variants and control groups.
  
• Advanced segmentation based on granular user behaviors.
  
• Setting up webhooks or custom API integrations.
  
• Configuring multi-language or localized messaging.

To wrap up

In just one hour, you have transformed an empty dashboard into a fully configured platform running an automated engagement campaign. It’s this focused approach that explains why one in every five new apps uses a platform like OneSignal to onboard, engage, and retain customers. By following this prioritized plan—securing your data foundation, activating a single journey, and testing rigorously—you have built an essential base for future success.

This initial, focused effort is the key to unlocking the long-term power of a customer engagement platform. It creates the momentum you need to build a sophisticated messaging strategy that drives lasting retention and growth for your business.

Here’s a quick exercise. Install a clean build of your app on a test device, and use it the way a real customer would for a couple of days. Let the notifications pile up without clearing them. Then open that notification center and look at it with fresh eyes. What you see is an unfiltered record of the experience you’re creating, message by message. And for a lot of teams, the gap between the strategy they think they’re running and the reality sitting in that tray is humbling.

How to run a notification center audit

The goal is simple: experience your mobile messaging strategy the way an actual user does. Not in a dashboard, not in a journey builder, not in a Figma mock. On a real device, in real time, with real context.

• Install a clean build of your app on a test device. Don’t use a staging environment with debug flags. Use the same build your customers download.
• Trigger a range of notifications. Browse products, abandon a cart, complete a purchase, go inactive for a day, come back. Hit as many of your automated flows as you can.
• Let them accumulate. Don’t clear your notification tray for 48 hours. Let the full picture build up the way it does for someone who checks their phone a few times a day but doesn’t obsessively swipe away every alert.
• Review with honest eyes. Open the tray and ask: does this feel helpful or noisy? Is the timing right? Would I tap any of these? Would I mute this app?

What you find will fall into a few predictable categories. Let’s walk through each one.

Common anti-patterns (and how to fix them)

Notification clustering

This is the one where three messages land in a 15-minute burst: a promotional offer, a cart reminder, and a shipping update, all stacked together like the app is having a panic attack. Each message might be reasonable on its own. Together, they train the user to ignore everything from you, because the cognitive cost of sorting through the pile isn’t worth it.

The fix: Frequency capping and Intelligent Delivery. A good notification management system lets you set limits on how many messages a single user can receive in a given window, and spaces non-urgent sends to arrive when the user is most likely to engage rather than whenever the campaign happens to trigger. Batch low-priority updates into summary notifications where it makes sense. Not every event in your system needs its own push.

Redundant multi-channel messages

You know the one. A push notification says “Don’t miss our summer sale!” and then an email arrives 10 minutes later with the exact same copy and the exact same hero image. Then maybe an SMS for good measure. The user hasn’t even had time to read the first message before channels two and three pile on.

The fix: Sequenced, cross-channel journeys with fallback logic. A multi-channel messaging platform should let you define a primary channel and escalate only if the first message goes unread. For example: send a push. If it’s not opened within two hours, follow up with an email. If the email goes unread for a day, try SMS as a last resort. Each channel has a distinct role in the sequence, not a redundant one. OneSignal’s re-engagement Journeys builder walks through exactly how to set this up.

Wrong message, wrong channel

Sending a generic 20%-off promotional blast via SMS or RCS is a fast way to get users to block your number. A text message carries an implied urgency. When someone sees a text from a brand, they expect it to matter: a delivery is arriving, an appointment is in an hour, a security code is needed right now. Using that channel for a promotion you could have sent via email or push violates the unspoken contract users have with each channel.

The fix: Define a clear role for each channel and stick to it.

• Push notifications: Timely, contextual, high-value. Cart reminders, price drops, breaking updates.
• Email: Longer-form content, transactional receipts, newsletters, onboarding sequences.
• In-app messages: Contextual prompts while the user is active. Feature announcements, feedback requests, onboarding tips.
• SMS and RCS: Genuinely urgent. Authentication codes, delivery-day alerts, time-sensitive confirmations.

For more on setting up each channel properly, OneSignal’s channel setup documentation covers the technical side, and our multichannel marketing glossary entry is a useful reference for the strategic framing.

Stale and irrelevant messages

The abandoned cart notification for an item you already purchased on your laptop. The “we miss you” push that arrives while you’re literally using the app. The event promotion that lands two hours after the event started. These are all symptoms of a messaging system that’s operating on stale data, or no behavioral data at all.

The fix: Real-time behavioral triggers and proper segmentation. Your notification system should know that a user already converted before it sends a conversion nudge. It should know that a user is currently active before it sends a win-back message. This is table stakes for any customer engagement platform in 2026, and if your current setup can’t do it, that’s the gap to close first.

Building a healthier strategy

Personalize at the behavioral level

"Personalization" that starts and ends with a first name merge tag is not going to move your metrics. The notifications that actually get tapped are the ones that reflect what a specific user did (or didn’t do) recently: the category they browsed, the feature they haven’t tried, the subscription that’s about to renew. Behavioral data is the difference between a message that feels relevant and one that feels like it was addressed to a mailing list.

Measure, learn, adjust

Run your audit once and you’ll find problems. Run it quarterly and you’ll track whether you’re actually fixing them. Pair the qualitative audit (looking at your notification tray) with quantitative monitoring: delivery rates, open rates, click-through rates, and conversion events per channel. If your push CTR is climbing but your email open rate is dropping, that’s a signal about channel preference you can act on.

OneSignal’s engagement trends dashboard and engagement analytics make this kind of ongoing measurement practical without building custom reporting.

Your notification center is a mirror

A cluttered, repetitive, poorly timed notification tray isn’t a minor UX issue. It’s a direct readout of strategic gaps: missing coordination between channels, absent frequency controls, stale data powering automated flows, or simply no clear philosophy about what each channel is for.

A clean tray, on the other hand, one where every message feels timely, relevant, and distinct from the last, is the output of a team that has done the hard work of defining channel roles, investing in behavioral data, and building journeys that adapt to what users actually do.

The audit takes 20 minutes. The patterns it reveals will tell you more about the state of your engagement program than a month of dashboard reviews. Start there, fix the worst offender first, and work forward.

For a broader look at where engagement strategy is headed this year, the 2026 State of Customer Engagement Report is worth reading alongside this exercise.