Skip to content
← All articlesTesting & QA

Mobile Device Testing Strategy: Real Devices vs Emulators

Plan a mobile device testing strategy that balances real devices, emulators and cloud device farms, so your app works for users without a bloated test budget.

Every mobile team eventually asks the same question: which devices do we actually need to test on? Android alone runs on thousands of hardware and OS combinations, and iOS adds its own spread of screen sizes and versions. Testing everything is impossible. Testing only on the phones in the office is risky. A good device testing strategy sits between the two, and it is driven by data rather than guesswork.

Why device coverage matters

Bugs that appear only on certain devices are the ones that reach the app store reviews. Typical causes include:

  • Different screen sizes, aspect ratios and notches or camera cut-outs
  • Manufacturer-specific Android behaviour, such as aggressive battery management that kills background tasks
  • Older OS versions missing newer APIs
  • Low memory and slow storage on budget handsets
  • Differences in keyboards, fonts, accessibility settings and system dark mode

None of these show up in a desktop browser or in unit tests. They need a deliberate device plan.

Step 1: Build a device matrix from real usage

Start with your own analytics, not a generic “top 10 phones” list. If the app is live, look at the OS versions, device models, screen sizes and countries of your active users. If it is not live yet, use the target market as a guide: a product for Gulf markets and one for the UK will not share the same device mix.

A practical matrix has three tiers:

Tier Purpose Typical share of devices
Core Devices your largest user groups use. Every release is fully tested here. Few, high-usage models
Extended Different screen sizes, OS versions and manufacturers. Tested on smoke and key flows. A wider, rotating set
Edge Oldest supported OS, smallest screen, low-memory device. Tested before major releases. Handful of worst-case devices

Review the matrix every quarter or after a major OS release. Decide a minimum supported OS version explicitly, and write it down so product, design and QA agree.

Step 2: Know what each option is good at

Option Strengths Limits
Emulators and simulators Free, fast, easy to reset, good for automation in CI Do not reproduce real hardware, sensors, battery, thermal or network behaviour well
In-house real devices True behaviour, camera, GPS, Bluetooth, push notifications, biometrics Purchase and upkeep cost, limited variety
Cloud device farms Wide variety on demand, remote access, parallel runs Per-minute or subscription cost, some hardware features limited, latency

The honest answer is that you need a mix. Emulators give quick feedback during development. Real devices catch the problems customers actually experience. Cloud farms extend your reach without buying forty phones.

Step 3: Decide what to test where

A sensible split looks like this:

  • Emulators or simulators: unit and UI component tests, layout checks across screen sizes, automated smoke tests on every pull request.
  • Core real devices: full regression of critical journeys such as sign-up, login, payment, checkout and onboarding before each release.
  • Cloud device farm: the extended matrix, run in parallel against automated end-to-end suites.
  • Edge devices: performance, memory, startup time and behaviour on slow networks.

Features that depend on hardware should always be checked on physical devices. That includes camera and QR scanning, biometric login, location, Bluetooth, push notifications and deep links from other apps.

Step 4: Test conditions, not only devices

Device is only one variable. Add these to your test plans:

  • Slow, flaky and offline networks, and switching between Wi-Fi and mobile data
  • Interruptions such as incoming calls, notifications and app backgrounding
  • Low battery and low storage states
  • Different languages, text sizes and right-to-left layouts for Middle East audiences
  • Fresh install versus upgrade from the previous version
  • Permission denied or revoked scenarios

If your app is offline-first or relies on push, these conditions matter more than the model number.

Step 5: Automate wisely

Automating every device combination is expensive and often brittle. Automate stable, high-value flows and run them across the core and extended matrix. Keep exploratory testing manual, especially for new features and visual polish. Our regression testing guide and test automation strategy cover how to choose what to automate.

A quick checklist before release

  • Device matrix reviewed against current analytics
  • Critical journeys passed on all core real devices
  • Smoke suite passed on the extended matrix
  • Oldest supported OS and smallest screen checked
  • Network, interruption and permission scenarios exercised
  • Crash and performance monitoring ready for the first days after launch

Working with an outsourced QA team

If you outsource testing, share your analytics, supported OS policy and critical journeys up front, and agree who owns device procurement or cloud farm access. A dedicated QA team that knows your product can maintain the matrix and report device-specific issues clearly, with logs, screen recordings and exact steps.

AkMo Technologies provides software testing for web and mobile products, working alongside our mobile app development team or your own. To discuss a device strategy for your app, get in touch.

Let’s build what’s next.

Tell us about your platform, your product idea or the systems you want to modernise. Someone from our Chennai engineering team will get back to you personally.