Ok, here’s a quick update on how Usaura is doing. Some stats if you’re interested: after a day of running there’s around 100 tests, with over 3800 clicks. This averages at 38 clicks per test, which is more than I expected. I’m guessing a lot of people are trying out the app right now to see how it works.
When I launched the app yesterday there was a Firefox bug I left in which prevented clicks from registering. I fixed it soon after but as half the people who tried the app are on Firefox I’m guessing it left a lot of people confused as to why the app didn’t respond. Apologies if that happened to you, I hope you give it another shot.
Thank you to everyone who left feedback and suggestions in the previous post. Here are 3 updates:
1. More customization
I’ve added a couple of important features to the test creation page. Firstly, there’s now an option to make the test private to keep it out of the public test pool. This can be useful if you don’t want to reveal unreleased product interfaces, or more important, if you want a more targeted audience as in the case of foreign languages.
Secondly, you can now customize the maximum number of clicks a tester can make before the test ends. By default I used 4 clicks to allow people to set up tests that ask the user to click on multiple areas. Now you can trim that down to 1 if you just want one location, or more if you like. This will default to 1 from now on as this is what the majority of the tests focus on.
2. Improving test quality
A couple of people uploaded inappropriate stuff that’s clearly not testing anything useful (there was even one ad!). This is to be expected with an app that doesn’t have a registration barrier. I do want people who take tests to get relevant material so they can actually get value out of them by seeing the results. The public test pool needs to be cleaned up a little.
For this reason there’s now a “Flag abuse” button on every public test. If a few people click this button, the test is taken out of the public pool circulation. It’s a soft measure, so the test still remains up and can keep collecting results from the testers the user brings themselves, it’s just no longer part of the public test pool.
One thing I want to note about the public test pool is that it focuses on the more recent tests, so over time the older tests are also taken out of circulation. This ensures the inflow of testers are not spread out too thin.
3. Improving the flow
I’ve added more buttons to make it easier to take multiple tests. On the test page there’s now a skip button, which lets you go right ahead to another test if the one you’re on doesn’t make sense (e.g. it’s in another language—another problem I’ll need to try and fix). This button is hidden on private tests as they are more targeted.
When you complete a test you’re directed to a thank you page. Here again there’s a new button letting you take another test. Results page now has a link to the home page so you can jump to another test or the new test page quickly. So if you like taking click tests—and it can be fun—it’s now much easier to do.
One of the main things I still need to add is the ability to chain several tests together so people can test multiple screens or multiple versions in one go. This is a priority feature and will be added soon.
Also need to implement a better way to display time. This should be visual rather than just text, i.e. a diagram of a clock with every click as a little arrow so you can instantly get a feel for how long the test took most people to complete. Finally, will need show the drop-out out rate (i.e. people who loaded the test but didn’t click anywhere), which will be even more important now that there’s a “Skip” button.