Ecommerce App Analytics: KPIs That Actually Predict Revenue

Ecommerce app analytics KPIs separate brands that profit from their app investment from brands that built an expensive second website nobody uses. Mobile accounts for roughly 57% of holiday ecommerce sales in 2026, and shoppers who buy through apps add 40% more to their baskets than equivalent mobile web shoppers — but only when the app is properly measured, iterated on, and tied to real revenue behavior. The mistake most brands make is tracking the wrong signals.

Vanity vs. Revenue Metrics: What Most Brands Track Wrong

Every ecommerce app generates impressive-looking numbers at launch. Downloads climb. Daily active users tick up. Session duration looks decent. Then, six months later, the team realizes the app is generating a fraction of the revenue they projected — and nobody can explain why, because none of the metrics they were watching predicted it.

The core problem is that most teams inherit their analytics framework from mobile gaming or social apps, where engagement is the product. In ecommerce, engagement is only valuable when it converts to purchase behavior. A user who opens your app 12 times without buying generates exactly zero revenue and costs you push notification infrastructure to maintain.

Here is the practical distinction:

Vanity Metric	Why It Misleads	Revenue Metric to Track Instead
Total downloads	Includes installs that never open a second time	D1 / D7 / D30 retention rate
Daily active users (DAU)	Active sessions without purchases = zero revenue	App-initiated purchase rate
Average session duration	Long sessions can mean poor UX, not high intent	Sessions per purchase (lower = better)
App Store rating	Ratings reflect satisfaction, not spend	App AOV vs. web AOV
Push notification send volume	Sends mean nothing without opens and clicks	Push opt-in rate + push-to-purchase rate

The shift from vanity to revenue metrics is not a dashboard cleanup exercise — it requires custom instrumentation. Standard analytics SDK installs do not capture the events you actually need. Plan for a dedicated instrumentation sprint before you can generate meaningful data.

Tier 1 KPIs: The Metrics That Directly Predict Revenue

These five metrics have the strongest direct correlation to app revenue. If any of them are unhealthy, fix them before investing in acquisition or new feature development.

1. D30 Retention Rate

D30 retention measures the percentage of users still active 30 days after install. Industry benchmarks for ecommerce apps: D30 retention above 15% is considered strong; the average sits around 5–8%. This is the single most predictive leading indicator for long-term app revenue.

Why D30 specifically? By day 30, users who install out of curiosity have churned. What remains is genuine repeat behavior — the users who have integrated the app into their shopping habits. Brands with D30 retention above 20% consistently outperform on app LTV because each retained user represents a compounding revenue stream, not a one-time purchase.

2. App-Initiated Purchase Rate

This measures what percentage of app sessions result in a completed purchase. For healthy ecommerce apps, this typically runs 2–5% of sessions (higher than mobile web's 1–3% average, lower than the best desktop checkout flows). If your purchase rate is below 1.5%, you likely have a checkout flow problem, a product discovery problem, or a friction point that instrumentation will surface.

3. Push Notification Opt-In Rate and Push-to-Purchase Rate

Push notifications are one of the primary ROI drivers that justify app investment over mobile web. But they only deliver that ROI if users opt in and if those notifications drive purchases, not just opens.

Healthy push opt-in rates for ecommerce apps run 45–65%. If yours is below 35%, the timing or framing of your opt-in prompt needs work — most brands make the mistake of asking for notification permissions on the first screen, before the user has any reason to say yes. Target push-to-purchase rate: 3–8% of push opens resulting in a purchase within 24 hours.

4. App AOV vs. Web AOV

This comparison tells you whether your app users are your best customers or just convenience shoppers. Ecommerce apps with strong personalization and loyalty integration typically show app AOV running 15–30% above web AOV for comparable products. If your app AOV is below or equal to web AOV, your app is not creating incremental value — it is just providing an alternate channel for the same transactions.

5. 90-Day Repeat Purchase Rate for App Users

The 90-day repeat purchase rate captures how many app users make a second purchase within three months. This is the metric that justifies the entire LTV premium attributed to app users. A healthy benchmark: 30–45% of app users who make a first purchase should make a second within 90 days. Brands below 25% are typically missing personalized re-engagement flows or have weak loyalty mechanics.

Tier 2 KPIs: Engagement Indicators That Signal Future Churn

Tier 2 metrics do not directly measure revenue, but they are early warning systems. When these shift, revenue KPIs follow 30–60 days later — which gives you a meaningful window to intervene before the decline shows up in your actual numbers.

Wishlist Add Rate

Wishlist adds are one of the strongest purchase-intent signals in ecommerce apps. Users who add to a wishlist are 3–5x more likely to purchase within 30 days than users who browse without saving. A declining wishlist add rate signals weakening product discovery or catalog relevance — worth investigating before it hits purchase rate.

In-App Search Usage Rate

Brands with robust search analytics know that what users search for — and fail to find — is a direct product and inventory feedback loop. If 20%+ of searches return zero results, you have a catalog gap or a search quality issue. Track search-to-purchase rate separately from browse-to-purchase rate; the two user paths have meaningfully different conversion rates and require different optimization strategies.

Session Frequency Trend (7-Day Rolling)

Track how often your retained users are opening the app on a 7-day rolling basis. For healthy ecommerce apps, retained users typically open 2–4 times per week (higher during sale events, lower during lull periods). A downward trend in session frequency among retained users — before any change in D30 retention — is a churn leading indicator. By the time retention drops, the behavior shift has been happening for weeks.

Checkout Abandonment Rate (App-Specific)

Checkout abandonment in apps should be meaningfully lower than on mobile web, because apps can persist payment methods, pre-fill addresses, and reduce friction at the point of purchase. If your in-app checkout abandonment rate is above 65%, you are not capturing the experience advantage that justifies the build cost. Compare abandonment at each checkout step to identify the specific friction point.

App vs. Web Benchmarks: Are Your App Customers Actually More Valuable?

Before investing in app growth — whether through paid acquisition, ASO, or new features — confirm that your app customers are actually generating higher value than web customers. This comparison should be part of your monthly analytics review.

Metric	Mobile Web Benchmark	Healthy App Benchmark	What the Gap Means
Conversion rate	1.5–3%	2.5–5%	App reduces friction vs. web
AOV	Baseline	+15–30%	Personalization drives larger baskets
90-day repeat purchase rate	18–25%	30–45%	Push + loyalty create return behavior
12-month LTV	Baseline	+20–40%	Combined retention + AOV lift compounds
Return rate	Industry avg.	Typically lower	App users are higher-intent buyers

If your data shows app users performing below these benchmarks relative to web, the problem is almost always one of three things: weak onboarding (users never experience the personalization that drives the premium), missing push notification strategy (the retention engine is not running), or a checkout experience that is not actually better than mobile web.

Our team at Atlas has worked with brands where the app had been live for 18 months with zero intentional analytics instrumentation. The app was technically functioning — it had thousands of installs — but the team was making product decisions based on session counts that had no relationship to revenue. Fixing the instrumentation and spending 60 days on data collection revealed that three specific product categories had zero app-to-purchase conversion, while two others were generating app AOV 38% above web. That data reshaped the entire content and merchandising strategy for the app. For context on what an instrumented app architecture looks like from the ground up, our guide on whether your ecommerce brand should build a mobile app covers the full decision framework and build requirements.

Setting Up Your Analytics Stack: Firebase, Amplitude, Mixpanel

The three analytics platforms that dominate ecommerce app measurement each have a specific strength profile. Choosing the wrong one does not break your analytics, but it does mean you will hit limitations when you try to answer the questions that matter most.

Firebase (Google Analytics for Firebase)

Firebase is the strongest choice when crash reporting and behavior analytics need to be tightly correlated. If a checkout bug is causing abandonment on a specific device type, Firebase surfaces this connection in ways that pure behavioral tools do not. It is also the default choice for teams already in the Google ecosystem — BigQuery integration is seamless, and the cost (free at most ecommerce scales) makes it difficult to argue against as a baseline layer. The limitation: revenue attribution and funnel analysis are weaker than Amplitude or Mixpanel without significant custom configuration.

Amplitude

Amplitude is the strongest standalone choice for ecommerce brands that need revenue attribution alongside behavioral analytics. Its cohort analysis, revenue correlation tooling, and Shopify integration make it possible to connect in-app behavior to actual purchase outcomes without building a custom data pipeline. For brands running multi-channel commerce (app + web + email), Amplitude's cross-platform identity resolution helps de-duplicate users across touchpoints — which is critical for accurate LTV calculations.

Mixpanel

Mixpanel excels at funnel analysis and feature adoption tracking. If your primary analytical question is "where are users dropping out of the purchase path and which features are driving conversion," Mixpanel's event-based model and funnel visualization tools are best-in-class. It integrates cleanly with custom mobile app development stacks, including both React Native and Flutter builds, and its pricing scales reasonably for mid-market brands generating under 10 million monthly events.

Recommended Stack Configuration

For most ecommerce brands building their first analytics layer:

• Firebase as the crash reporting and baseline event logging layer
• Amplitude as the revenue and cohort analytics layer
• Custom event taxonomy agreed upon before any SDK installation
• Monthly data review cadence with cross-functional ownership (product + marketing + engineering)

Do not add Mixpanel on top of these two unless you have a dedicated analytics resource to manage three platforms. Tool sprawl is a bigger risk than tool gaps at this stage.

A 90-Day Analytics Roadmap for New Ecommerce Apps

The most common mistake with new ecommerce app analytics is trying to optimize before you have a baseline. Thirty days of clean, properly instrumented data is worth more than six months of metrics collected from an incomplete SDK install. Here is the roadmap we run with new app clients.

Days 1–14: Instrumentation Sprint

Define your complete event taxonomy before touching any SDK. Minimum required events for an ecommerce app: app install, account create, product view, add to cart, checkout start, checkout complete, push permission prompt shown, push opt-in, push notification open, wishlist add, search query submitted, search result click, and first repeat purchase (separate event from standard checkout complete). Implement these events with consistent naming conventions across iOS and Android. Test against a staging environment before going to production.

Days 15–44: Baseline Collection

Run the app with no major product changes for 30 days. Collect clean baseline data for all Tier 1 KPIs. Resist the urge to ship new features during this window — every change introduces a confounding variable that makes it harder to understand what your baseline actually is. Document the baseline numbers before day 45.

Days 45–60: Diagnosis and Prioritization

With 30 days of clean data, map each Tier 1 KPI against the benchmarks in this post. Identify your weakest metric — the one with the largest gap from the healthy benchmark. That gap is your first optimization target. For most new apps, D30 retention and push opt-in rate are the first two problems surfaced, because they are the metrics teams invest least in during the initial build phase. Working with your Shopify ecommerce team during this phase ensures that fixes to the app experience are coordinated with any web-side merchandising or loyalty program changes.

Days 61–90: First Optimization Cycle

Run one focused optimization experiment against your weakest metric. For D30 retention: test a structured onboarding sequence with a loyalty reward trigger at day 3. For push opt-in: test permission prompt timing (post-first-purchase versus post-first-browse). For app AOV: test personalized product recommendations on the home screen versus a curated best-seller list. Measure against the baseline collected in days 15–44. At day 90, you should have your first data-driven insight about what actually moves your app's revenue needle.

FAQ

What are the most important ecommerce app analytics KPIs to track?

The KPIs with the strongest correlation to revenue are D30 retention rate (target: above 15%), app-initiated purchase rate, push notification opt-in rate, app AOV versus web AOV, and 90-day repeat purchase rate for app users. These five metrics tell you whether your app is generating incremental revenue or simply providing an alternate screen for existing web shoppers. Downloads, daily active users, and session counts are secondary signals — they only matter if the revenue KPIs are healthy.

What is a good D30 retention rate for an ecommerce app?

D30 retention above 15% is considered strong for ecommerce apps, meaning 15 out of 100 users who install your app are still active 30 days later. The industry average sits closer to 5–8%. Brands consistently above 15% at D30 typically have strong push notification programs, a loyalty or subscription mechanic that creates return behavior, and personalized product discovery that improves with each session. If your D30 retention is below 10%, fixing that gap delivers more revenue than any amount of new user acquisition.

Which analytics platform should I use for my ecommerce app?

For most Shopify-based ecommerce brands, Amplitude is the strongest choice because it handles both behavioral event tracking and revenue attribution in a single platform, with direct Shopify integration. Firebase is the right choice if you need tightly coupled mobile crash reporting alongside analytics and your team is comfortable with Google's ecosystem. Mixpanel is a solid alternative when the primary need is funnel analysis and feature adoption tracking rather than revenue attribution. Whichever platform you choose, the critical requirement is custom event instrumentation — standard SDK installs only capture session counts, not the purchase-behavior signals that actually matter.

Why do app customers spend more than web customers?

App customers typically show 20–40% higher LTV than mobile web users for three compounding reasons. First, app installation is a self-selection signal — customers who invest effort in downloading, installing, and keeping an app on their phone are already higher-intent. Second, apps enable personalization that mobile web cannot match: persistent wishlists, loyalty point visibility, and browsing history all create a stickier experience. Third, push notifications drive re-engagement at near-zero marginal cost, while retargeting mobile web visitors through paid ads is increasingly expensive and limited by cookie deprecation.

How do I set up ecommerce app analytics from scratch?

Start with a one-week instrumentation sprint before doing anything else. Install your analytics SDK (Amplitude, Firebase, or Mixpanel), then define and implement a minimum of 12 custom events: app install, account create, product view, add to cart, checkout start, checkout complete, push opt-in, push open, wishlist add, search query, app background after purchase, and first repeat purchase. Map these events to your revenue funnel and set baseline benchmarks for each. Once you have 30 days of data, segment by acquisition channel, device type, and user cohort to identify where users are dropping out of the purchase path. The entire setup takes 4–6 weeks to produce actionable data.