Test Your Design on Actual Devices
Planning how your design will scale for different sizes is an important part of designing for those devices, but you cannot stop at that theoretical approach.
Being able to interact with your design on real devices will show you how it really works, either confirming the decisions you made or showing you where improvements may need to be made.
In addition to your own testing, running usability tests with actual users is another important part of working with your designs in a real world setting.
First check existing analytics to see the percentage breakdown for users based on devices. Prioritize your tests based upon the breakpoint information. If you’re moderating the tests on-site, it’s also worth purchasing commonly used devices
If you can spare the resources, it’s worth testing different devices with different users. While this multiplies the number of users required for testing, it is the safest way to prevent bias due to design familiarity. For example, if you’ve designed three breakpoints, testing with 15 users (five per breakpoint) will help you uncover most usability issues. Services likeUserTesting help scale usability testing affordably since they handle the recruitment, task creation, moderation, and analysis.
If you’re particularly tight on timing or money, you can test multiple breakpoints with the same users. Just randomize the order of devices presented to minimize the bias from design familiarity. For example, if you test mobile, tablet, desktop with User A, switch up the order for User B.