Quick answer: To test on real devices: test on a representative spread of real hardware, prioritize the devices your players use, and capture crashes from the full real device range in production.
Testing on real devices catches issues your machine and emulators miss. These are the steps.
Step 1: Test on a Representative Spread of Hardware
Start by testing on a representative spread of real hardware, not just your development machine: include different performance tiers, OS versions, and screen sizes, especially the low-end hardware that defines your performance floor. Real devices behave differently from your machine and from emulators, surfacing device-specific issues.
Bugnet extends your device testing to the full real population: since you can only test a handful of devices, Bugnet captures crashes from whatever devices your real players use, tagged with device and OS, so the device coverage your hands-on testing cannot reach is covered by real-world capture.
Step 2: Prioritize the Devices Your Players Use
Next, prioritize testing on the devices your players actually use: you cannot test every device, so focus your hands-on testing on the hardware most common among your players (and the low-end devices most likely to struggle). Testing the devices that matter most to your audience gets the best return on limited testing.
Bugnet tells you which devices to prioritize: it shows the actual distribution of devices your players use and which ones crash, so you can focus your device testing on the hardware your players actually have and that actually has problems, rather than guessing which devices to test.
Step 3: Capture Crashes From the Full Real Device Range
Finally, complement device testing by capturing crashes from the full real device range in production: no amount of hands-on testing covers every device players use, so capturing crashes from real players across all their devices catches the device-specific issues testing missed. This is the most complete device coverage.
Bugnet provides exactly this: it captures crashes from real players tagged with device and OS and groups them by signature, so device-specific issues surface from the complete real device population (far more than you could test), with the context to fix them, completing what hands-on device testing starts.
To test on real devices: test on a representative spread of real hardware, prioritize the devices your players use, and capture crashes from the full real device range in production, real devices catch issues your machine and emulators miss.