3 ms·
This is true for everyone from janitors to CEOs. The big question is, what jobs are harmless, which cause improvements being “over-worked”, and which have the g
by AJ007 8y ago
This is true for everyone from janitors to CEOs. The big question is, what jobs are harmless, which cause improvements being “over-worked”, and which have the greatest odds of causing substantial destruction?
UI / UX redesigns should be tested with small amounts of users, with all key metrics measured, and see what happens to those metrics and user retention after a few months. Otherwise, UI redesigns has a chance of causing a lot of destruction both by loss of users and bugs. Fortunately for software this isn’t that difficult to do. Other tasks, less so or impossible.
- sizzle 8y agoAbsolutely, and testing with a diverse population with a sufficiently large sample size will yield the highest quality signal for measuring redesign success e.g. time on task, task completion, discoverability/usability, etc.