In the early Floatask prototypes we crammed in everything we'd seen
in the serious task managers: tags,
projects, areas, contexts,
energy, estimated time. It was a joy to
look at on the settings screen. It was a nightmare to use.
What the testers did
They kept repeating a very specific pattern: they'd jot down a task, open the detail, look at the fields, close them without touching anything, and go back to the list. When we asked why they didn't use the labels, the answer was always some version of:
"I just don't feel like thinking about that right now. I only wanted to write it down."
Later, reviewing at the end of the week, they couldn't find something. And they blamed the app. The conclusion took us two weeks to swallow: the problem wasn't them. The problem was us, asking them to decide things they had no reason to decide.
What we decided
- Tags out. A list with a color covers 95% of cases.
- Contexts out. The vague date is already a context ("today" = now; "this week" = soon).
- Numeric "energy" and "time" fields out. Nobody fills them in correctly.
What was left: title, date (at whatever level of detail you want), importance (4 levels), assigned person, list, notes. All optional. Enough.