The digital note-taking software, Evernote, lets users set reminders for themselves. Our heuristic evaluation found using this feature to be un-intuitive, with multiple redundant steps and lack of feedback
Identify specific usability issues with the feature by conducting a set of cognitive walkthrough usability tests, and analyze these for common patterns.
I took a course called "Usability Testing". Being students, my group and I decided to test the usability of the note-taking app, Evernote. We noticed there was a reminder feature, and that it was pretty confusing to use. We used Neilsen's 10 interaction design principles.
After using the Reminder feature, we realized we were confused. Mainly, we were confused about:
1. Whether or not a reminder had been set,
2. What time we had set the reminder for,
3. How exactly reminders were related to notes, and...
4. Why setting a reminder took so many frickin' steps.
How does using the reminder feature differ from expected functionality?
What are the most confusing steps in using the feature to remind yourself to do something?
Is this feature easily navigable across machine states?
Does it allow users to efficiently set reminders?
Is the reminder status visible across machine states?
There were 3 ways to set reminder, which all attached reminders to notes in slightly different ways
Setting a reminder took two three steps: Clicking the icon, clicking "Notify Me", and entering the time. Once you clicked a day on the calendar, the reminder window disappeared automatically.
To test how well people understood the relationship between notes and reminders, we gave users two scenarios to test our research questions, one that would force users to attach reminders to existing notes (or so we thought...), and one that didn't.
We recorded video and audio from the users interactions with Evernote using a software called Morae. This allowed us all to observe participants in real time and return to footage after the study based on notes we took. We evaluated the following measurements:
Number of Errors
We organized our observations into conceptual groups, separating observations from each task. We then looked for patterns that were common among all participants. These groups together described usability issues that we had observed in our study.
5 of 6 participants thought of reminders as a separate function from notes. When participants created reminders separate from existing notes, they were surprised that Evernote automatically made a new empty note, and said it confused them. For example, they could end up with multiple notes of the same title if they tried to create a new reminder with the same name as an existing note.
Clearly defined fields in sidebar for notes and reminders
Don’t auto-generate new notes with new reminders
Recognize duplicate reminders (give option to add or not)
On a positive note, the alarm icon was easily recognized as the "reminder" icon. But it wasn't clear that pressing "Notify Me" let participants set the date and time. There also wasn't much feedback about whether or not the reminder was set. 4 of 6 participants had to go back to select a time because the calendar window closed out as soon as they clicked on a date.
Change "Notify Me" to "Set Date/time"
Add "Set Reminder" button to reminder calendar popup to help affirm that date and time were set
After the reminder is set, display the date and time in the sidebar
Participants struggled with setting a reminder from the "Notes" sidebar because the '+' icon in the 'add reminder' field didn't look clickable
"Add reminder to..." field should strongly indicate that it is a text input field, maybe with a border, underline, so users know it is clickable.
The field reminder field should not disappear when no reminders are set, since setting just one reminder is a common workflow.
I was especially surprised at how differently each participant approached and completed the scenarios we laid out for them. We would expect, for example, that a user would want to clear a reminder when they were done with it, but some people just marked it as "done" and left it there. Also, some tasks that were easy for some people felt impossible to others. It's true that everyone interacts with an interface a little differently.
If I could do this again, I'd thing more thoroughly about the range of ways users could navigate a scenario while I write it.