There’s no one perfect user interface design, and you can’t get good usability by guessing. You have to try (and test) multiple design ideas.
Combine these three methods, in this order, for the best results
- Competitive testing – Compare usability testing of your design with 3-4 competitor’s designs.
- Parallel design – Create 3 variations of your new design, test them, and bring the best elements of each into your new design.
- Iterative design – Throughout development of the final design, plan for at least 3 iterations (up to 10) of testing to ensure your interface is as usable as possible.
Competitive usability studies test your own design and 3–4 other companies’ designs. The process model looks the same as for parallel design, except that the original design alternatives are pre-existing sites or apps as opposed to your own wireframe designs.
- Gain insight to user needs before attempting to design for those needs
- Build off of what your competitors are already doing well
After user testing, create a single merged design, taking the best ideas from each of the competitors versions. The usability study is not a competition to identify “a winner” from the designs. Each design always has some good parts and some that don’t hold up to the harsh light of user testing.
In a parallel design process, you create multiple alternative designs at the same time. To stay within budget, all parallel versions should be created quickly and cheaply. For websites, design 10 key pages. For applications, design only the top features. Spend only a few days on rough versions and stick to wireframes.
Ideally, create 2-3 different design alternatives using your best ideas. Any more and users get jaded and can’t articulate the differences. After user testing, merge the best functionality into a single design.
Iteration simply means to step through one design version after another. For each version, you conduct a usability evaluation (such as user testing or heuristic evaluation) and revise the next version based on these usability findings.
More iterations are better, as budget allows. At least 3 is ideal. NN/g research shows that each iteration improves usability by 38% on average.