12 onboarding patterns from top apps and what to A/B test on each
A pattern library connecting UI inspiration to experiment ideas so your next onboarding test is informed by what works in the market, not guesswork.
Design galleries show what top apps look like. Growth blogs explain tactics in the abstract. What is missing is the bridge: this pattern exists, here is when it works, and here is what to test first.
This library covers twelve onboarding patterns seen across leading consumer and subscription apps. For each pattern you get fit criteria, implementation notes, and two concrete A/B test ideas you can run as remote flows in Rheo without rebuilding native screens every iteration.
How to use this library
- Pick the pattern closest to your current onboarding, not the flashiest reference.
- Read the "when it works" guardrails honestly.
- Choose one test idea with a clear primary metric (activation, trial start, or completion).
- Build control and variant as Rheo channels or experiments.
- Measure step-level drop-off, not only final conversion.
Patterns are not mutually exclusive. Many apps combine progressive disclosure with a single paywall beat. Avoid stacking three patterns at once in your first test.
1. Progressive disclosure
What it is: Reveal features across multiple beats instead of one dense screen.
When it works: Complex products with more than one core capability (finance, health, productivity).
Test ideas:
- A: Four beats, one feature each vs two beats with paired features. Metric: activation rate.
- B: Same content, but beat three moved before account signup vs after. Metric: signup completion without activation drop.
2. Quiz onboarding
What it is: Questions about goals, experience, or preferences that shape the rest of the flow.
When it works: Fitness, learning, wellness, personalization-heavy subscriptions.
Test ideas:
- A: Three questions vs six questions. Metric: flow completion and day-7 retention.
- B: Quiz results shown as a "custom plan" screen vs skipped straight to product. Metric: trial start rate.
3. Stories-style sequential flow
What it is: Full-screen beats with tap-to-advance and progress indicators.
When it works: Emotional categories where narrative beats feature lists.
Test ideas:
- A: Five beats vs three beats before permissions. Metric: ATT grant rate and activation.
- B: Final story CTA "Continue" vs "Start my plan". Metric: paywall view rate.
See the dedicated post on Stories-style onboarding for deeper structure guidance.
4. Paywall at end of onboarding
What it is: Value demonstration first, monetization after the user understands the product.
When it works: Categories with strong trial value and low same-day churn risk.
Test ideas:
- A: Paywall immediately after last value beat vs after first activation action. Metric: trial starts per install vs activation rate.
- B: Soft paywall (dismiss allowed) vs hard paywall on first session. Metric: day-7 paid conversion.
5. Paywall early (day zero)
What it is: Monetization before or alongside minimal setup.
When it works: Content libraries, utilities with obvious premium gate, high-intent paid traffic.
Test ideas:
- A: Paywall on screen two vs screen five. Metric: revenue per install.
- B: Free trial emphasis vs annual discount emphasis on same placement. Metric: ARPU at day 30.
Run paywall pricing tests in RevenueCat Experiments; run placement tests in Rheo.
6. Checklist setup
What it is: Explicit tasks (enable notifications, connect account, complete profile) with checkmarks.
When it works: B2B-ish mobile tools, banking, apps requiring configuration.
Test ideas:
- A: Checklist visible on home vs dedicated onboarding checklist screen. Metric: tasks completed per user.
- B: Three tasks vs five tasks on first day. Metric: activation without day-3 overwhelm (retention guardrail).
7. Social proof and trust
What it is: Reviews, press logos, user counts, security badges during onboarding.
When it works: New brands, finance, health, categories with skepticism.
Test ideas:
- A: Trust screen before account creation vs omitted. Metric: signup completion.
- B: Star rating widget vs "X million users" stat. Metric: trial start on following paywall step.
8. Permission priming
What it is: Educational beat immediately before ATT, push, or location request.
When it works: Any app where permissions gate core value.
Test ideas:
- A: Priming copy focused on benefit vs focused on privacy control. Metric: grant rate.
- B: System prompt on beat two vs beat four. Metric: grant rate and activation.
9. Template / sample content onboarding
What it is: Pre-filled example project, workout, or document users can edit.
When it works: Creative tools, planners, note apps, meal planners.
Test ideas:
- A: User starts from template vs blank canvas. Metric: activation (first save or export).
- B: Template chooser vs auto-assigned template from quiz. Metric: time to activate.
10. Account last (guest first)
What it is: Let users experience value before mandatory signup.
When it works: High top-of-funnel drop at signup, products with tangible instant value.
Test ideas:
- A: Guest mode until activation vs signup on screen two. Metric: activation and account creation rate.
- B: Signup prompted with saved progress vs empty state. Metric: account completion after prompt.
11. Personalization by attribution
What it is: Different onboarding paths based on campaign, creative, or deep link.
When it works: Paid UA at scale with varied ad promises.
Test ideas:
- A: Generic path vs campaign-matched first beat. Metric: completion by source.
- B: Short path for returning reinstalls vs full path for new installs. Metric: trial start without support tickets (guardrail).
Requires acquisition context in SDK attributes (for example AppsFlyer via Rheo integrations).
12. Single-screen value gate
What it is: One strong value proposition screen, then straight into product or paywall.
When it works: Simple utilities, apps with known intent keywords, narrow jobs-to-be-done.
Test ideas:
- A: Video hero vs static illustration on the single screen. Metric: continue rate.
- B: Primary CTA "Try free" vs "See how it works" (extra micro-demo). Metric: activation within session.
Implement without rebuilding every test
Historically, each pattern change meant engineering tickets and App Store review. Remote flows invert that:
| Pattern change | Rheo approach |
|---|---|
| Reorder beats | Drag screen nodes, publish |
| Add quiz branch | Decision nodes on answers |
| Swap paywall placement | Move RevenueCat node or custom surface |
| New attribution path | Decision on acquisition.* traits |
| Win-back variant | Separate channel, same SDK |
Native rendering keeps gestures and performance consistent while copy, order, and branches change from the dashboard.
Checklist before you ship a test
- Primary metric named (activation, trial, completion)
- Guardrail metric named (uninstall, dismiss rate, support contacts)
- Control and variant documented on the canvas
- Sample size rough check (avoid 95/5 splits on low traffic)
- Runtime long enough to cover weekday/weekend behavior (often 7–14 days)
- Fallback edges wired on paywall and permission surfaces
From inspiration to action
Mobbin and similar tools answer "what does it look like?" This library answers "what do I try next?" Pick one pattern, run one experiment, read step funnels, then compound.
Over time, your pattern stack becomes a playbook: quiz plus progressive disclosure for fitness, guest first plus paywall at end for utilities, attribution paths for paid social campaigns.
Pattern selection matrix
| If your top problem is... | Start with pattern |
|---|---|
| Users never understand value | Progressive disclosure or stories |
| Wrong plan for user type | Quiz onboarding |
| Drop at signup | Account last or shorter checklist |
| Low trial after education | Paywall at end vs early test |
| Paid traffic mismatch | Personalization by attribution |
| Skeptical new users | Social proof and trust |
| Permission denial blocks aha | Permission priming |
| High intent, simple product | Single-screen value gate |
Revisit the matrix quarterly. As activation improves, the bottleneck moves (often from onboarding to retention or paywall).
Downloadable checklist (copy into your doc)
Use this before each experiment:
Pattern under test:
Control description:
Variant description:
Primary metric:
Guardrail metric:
Start date / review date:
Decision rule (ship / iterate / stop):
Step with highest expected impact:Rheo experiments map cleanly to this checklist: control and variant are channels or experiment arms, metrics come from step funnels and billing events, and ship decisions happen via publish without app store delay.
Anti-patterns when copying top apps
- Copying beat count without copy depth: five shallow beats lose to three strong ones
- Copying paywall timing from a different monetization model: content apps are not utilities
- Adding quiz questions for "engagement": extra taps without personalization hurt
- Trust screens on established brands: users already trust you; skip to value
- Running twelve tests at once: pick one pattern, one hypothesis
Summary
Twelve common onboarding patterns each suit different categories and risks. Two test ideas per pattern give you a backlog grounded in market practice, not random UI tweaks.
Build control and variant flows in Rheo, measure step-level outcomes, and promote winners without waiting on store review.
Start for free and turn your next onboarding idea into a live experiment.