I’ve found myself becoming increasingly dissatisfied with calendar applications. Fundamentally, I think they’re solving the wrong problem.
With all calendar apps you manage exactly when you do what, and add those events to your calendar. You decide when to go to the movies, and fill out the time, and title, and perhaps the location if you want traffic alerts. This has gotten a lot easier due to calendar apps accepting Natural Language additions to the calendar, and then parsing those results.
However, Calendar apps aren’t the best answer to the questions people ask their calendars. People ask their calendars:
- What do I have to do in the immediate future?
- What do I have to do now, in order to prepare for the future?
The core of these questions gets at the notion of how to best spend your time. What’s the most optimal arrangement of these events for my productivity and well-being?
How do Calendar Apps stack up?
Before looking at alternatives, lets see how current calendar apps address the questions outlined above.
No surprise that all of them offer both a view over upcoming events (Temporal View) and tasks to be completed (Agenda View). Some of the other distinctions are smart event adding, if you receive an email with ticket confirmations it will automatically be added. If you get an email with a date in plain text, you can generally also add that fairly seamlessly by clicking on the date.
Google Calendar has recently announced a feature called Google goals. The idea is to make it easier to plan time for habits that you want to work towards. For example you might set a goal to meditate in the morning, or go for runs in the evenings. Goals is the closest any product has come to selecting the time for an event, but it could go much further.
Fantastical and Google Calendar both allow for Natural Language additions to the calendar, while the Apple stock apps only do so for desktop. These additions are excellent, but we’re still missing something.
The primary driver of these features, is that people don’t really add things to their calendar. This makes a whole lot of sense, when you add something to the calendar you have to fill out all sorts of fields in order to make the most of your calendar. If you want Travel Alerts, you have to fill out the location information and match it to an actual location. The same thing occurs if you want to send an invite to a friend, you have to fill out that field too. To this end Natural Language Interaction has helped users enter that information.
I think that people have difficulty entering this information for other reasons. It’s not that they don’t want to spend a few more seconds to enter more content, but that they don’t know exactly when to complete a specific task. How do you enter an event if you don’t know when it starts? Google’s Calendar has an interesting solution to this problem. It allows you to view reminders added from inbox or calendar while looking at your daily agenda. But this solution doesn’t help people manage their time, it you to see both a to-do list, and your calendar at the same time, but this is a sub-optimal solution. Items on a to-do list take time, they don’t exist in their own isolated world.
These calendars do an excellent job of showing us what we have to do next. They manage fixed events — Meetings, Appointments, Classes — very well, but do not do well with general tasks or items in which the time allocation is flexible. They fundamentally don’t answer the question of how I can best spend my time.
What would a better calendar look like?
In order to empower people to make the most of their time, we have to allow them to make flexible events that rely upon other constraints. I’d like to be able to tell my calendar that I want to complete an paper by March 22nd, and that I like writing in the morning. It’ll then allocate time in the morning to complete the paper.
Of course, this solution requires the system to know how much time to allocate in total. You could approach this from two directions, the first being letting people define a total duration, the second being looking at past events and assuming a specific duration among other settings. The calendar would then allocate that time in the most efficient way according to my preferences.
The key shift here is that the user is no longer required to plan everything out specifically, the calendar becomes a place in which you can specify some things — appointments and meetings — and leave the other outside variables up to the system to allocate. People can add a whole bunch of tasks to the calendar and not have to worry about scheduling them, they just need to provide due dates.
It would also be beneficial to be able to include more complex constraints: On Friday, I’m going to go see a movie, then get food afterwards with John, Katie and Jessica. The system ought to be able to add those events in order, at times that everyone is available, and there is a showing. This is really a constraint problem, and constraint problems are ones we can solve computationally.
Most tasks don’t exist in isolation, they depend on each other. Only after you select a guest list can you mail invitations. But modern calendar apps don’t account for such events, the user has to explicitly decide the order, and when everything should be completed. If the user provides an ordering of dependencies then the calendar ought to be able to schedule things effectively.Instead people use task apps, which isn’t really a full solution either.
Allowing the calendar to schedule tasks opens up more time for us work instead of plan. The parts needed aren’t outside the realm of possibility either, constraint solvers are very well understood, as well as the machine learning to find your preferences for specific types of tasks.
Such a system helps us assess what we need to do right now, as it dynamically allocates time for the most pressing tasks, and works around your other events. Allowing us to more organically plan how to spend our time in the future. It answers the right question: Whats the best way to spend my time?