Q: What is an auto-scheduler?
A: It’s a kind of advanced calendar which has the power to schedule itself.
Slowly but surely, these AI-powered calendar apps are making their way into the lives of early adopters. They are looking for a way to boost the number of tasks they can manage effectively.
Taking over the grunt work of managing a calendar full of tasks, these apps do more than blindly shuffle around tasks and appointments. Instead, they act as intelligent agents, responding to the actions users take each day. This interplay, they report, is quite game-like: it can be immersive, challenging and fun at the same time. In this article, I describe the ways people play with their auto-schedulers and the impact this will have on the future of personal productivity.
Although the idea has been around for decades, recent advances in cloud computing and Artificial Intelligence (AI) have made auto-scheduling a practical reality. I started using one about 18 months ago and cannot imagine ever going back: it’s become a part of my new, daily routine.
In the first two articles in this series (here and here) on the gamification of task management, I described a number of games which users of task management apps are trying to play. I outlined six levels of technique/app usage which users climb up, as shown in the diagram below.
They don’t make this journey for idle reasons. Those who climb quickly are the ones who are trying to manage increasing numbers of time demands. In other words, in an effort to eke out greater personal capacity, they travel from Levels 1 to 6.
Furthermore, as I pointed out at the start of this series, this is more than a grim endeavor. They are hoping to achieve a game-like, engaging experience as they make their journey. Unfortunately, they haven’t found much luck. Few designers at any level have paid serious attention to the user experience or made any efforts to seriously transform it. At most, they make cosmetic differences.
As a result, while every functioning adult on the planet manages their tasks at one of these six levels, only a few million use electronic task apps. Do the math: out of three billion people with access to the internet, the biggest task management app only accounts for 10–20 million regular users.
That represents a huge gap. Plus an opportunity.
At Level 6, where the technology is new, the idea of auto-scheduling isn’t widespread. The most recent feature at this level is one offered by Todoist Smart Scheduling. It’s their first foray into Level 6 features, but they aren’t the first.
SkedPal, the developers of which I have done work for, is the most mature product and the one I currently use. Timeful, which was acquired by Google, Sheldonize, TimeTo and Focuster round out the collection of known players in the niche.
A comparison of apps at this level versus those found at Levels 3 and 4 show a big difference. They function as static databases, repositories of tasks. If the user adds in due dates or start-dates for tasks, the program reports which ones are overdue at any point in time, but that’s all.
Level 6 applications promise an experience that’s quite different. They are responsive — like the difference between using a monitor that offers touchscreen capabilities versus one that doesn’t.
In these programs, when you make a change to your tasks or calendar, they reply with a fresh new schedule. Multiple changes can be introduced one at a time, or grouped altogether. Sometimes the results it produces are surprising.
SkedPal users I interact with are regularly delighted (or shocked) by the calendar the program creates, the task which it suggests you work on next. Without the tool, you simply would not have remembered.
This makes sense. These programs are meant to do much more than your mind can do, in just a split-second. Now, there’s no danger in creating as many tasks as you like. As you enter hundreds of new ones or change a whole bunch at a time, the program responds the same way: with a fresh,optimized schedule.
This interactive give-and-take makes the UX game-like. You change your inputs and receive a different reaction each time. With continued use, as I show below, you become more productive in response to its unique feedback.
It’s obviously not a tool everyone would appreciate or use. Many have no need for its power. But there are hard-working people who find these capabilities useful, according to my research and experience.
Who Level 6 Users Are
Most users who migrate to these apps generally do so after making an effort to master Levels 4 and 5.
At Level 4, they used a Complex To-do list program that helped them store and view their tasks in different ways. At Level 5, they have substituted these apps with a digital calendar, which they use to juggle their tasks manually.
As I mentioned in the two prior articles, these two techniques have their limitations. While they are perfect for users who have fewer time demands, they both fail to meet the needs of users who manage a large number of tasks. Once they pass a particular threshold, Level 4 and 5 approaches get in the way.
A few get unstuck by using Level 6 auto-schedulers. To understand why they make the switch, let’s analyze their behavior using the same approach used in my prior two articles in the series: The Job to Be Done framework, and the mechanics of gamification.
The Job Level 6 Users Are Trying to Get Done
If you are a Level 6 user, you hire the technologies and techniques offered by auto-schedulers to accomplish three specific purposes.
Reason #1 — You are looking for a way to manipulate a large number of tasks, and want the app’s robotic features to save you time. With it, you should be able to create a schedule and refresh it in seconds rather than hours. You are no longer a victim of an unplanned interruption or the passage of several days.
Reason #2 — You hope to expand the feeling of being on top of things. Perhaps it has been lost for some time and now you want it to be restored. Or maybe it’s become fleeting and you want it to be continuous.
Reason #3 — You expect to experience an increased sense of mastery, a growth in your capacity. You want to be able to do more and therefore achieve new goals. You hire Level 6 apps in order to become a measurably more productive person who never allows tasks to fall through the cracks.
These are sound reasons. Each of them indicates the kind of expectation a Level 4 or 5 user would have if they were up-leveling, having decided to hire a Level 6 app.
However, my 18-month experience points to a far greater possibility. Beyond the three reasons, there are other reasons to hire an auto-scheduler.
New users of auto-schedulers quickly learn that the app needs to be run almost every workday, preferably early in the morning. The logic is simple: the schedule for yesterday is now stale. Anything that was left incomplete must be rescheduled by the app.
Also, things change from day to day, as you complete tasks and generate new ones as a result. Some become obsolete. The project you thought was important is deprecated, forcing you to change gears.
Auto-schedulers can do more than respond to these changes — one day they will capture unique data and build a database which reflects your behavior. When this happens, it will provide you with two further reasons to deepen your use of an auto-scheduler.
Further Reason #1 —Most people become “Total Task Schedulers” when they adopt Level 5 behaviors: they begin to manage the majority of their tasks in their calendar. Now, at Level 6 they can use the app’s data to improve this skill. While it’s possible to gather your own data at Level 5, a good auto-scheduler should provide you some unique insights into your performance in this area. For the first time, you can examine historical data, looking for strengths and weaknesses.
For example, if you are failing to visit the app every workday, the system’s data should reflect your inability to develop this habit, and perhaps suggest the creation of a mitigation plan.
Further Reason #2 — Not only can you focus on improving your skills, you can also use the data to improve the quality of your schedules. Each day when you reschedule your calendar, a good auto-scheduler should indicate the likely success of your schedule. For example, a calendar that is too tightly packed should throw up a red flag.
These two reasons don’t represent trivial attempts. Instead, they present an opportunity for app designers to educate and engage users, converting them from “users of an app” into “players of a game” whose expectations change over time.
In today’s world, that’s a given in well-designed video-games. Once a player moves past the entry level, they provide abundant opportunities to make measured improvements. Most of these new levels can’t be perceived at the start: they only become revealed with higher accomplishment.
The only difference here is that the game being played is “real.” In other words, it comes from a real-life situation, rather than the heavy fiction/fantasy used to construct a game like Angry Birds. Although the initial context is different, the fact is that all games are based on a common pool of mechanics as I mentioned in the prior articles. Fictional or non-fictional: they can both be made to be engaging.
While SkedPal does not have these levels built in, my experience has evolved, just as it would in a video-game. Here are some of the games I have found myself playing, even though the app doesn’t recognize them.
The Games Level 6 Users Try to Play
As I mentioned, all well-designed video-games consist of distinct levels which require higher levels of skill. Here are three “phases” I have discovered which can be used to craft these levels.
- Games at the Learning / Onboarding Phase
As I mentioned, new users spend the bulk of their time, in the beginning, trying to understand how an auto-scheduler works. For example, SkedPal makes a clear distinction between fixed and flexible tasks, keeping the former unchanged, while optimizing the placement of the latter. This distinction is new to most users.
Also, the user is asked to provide a set of weekly time maps. These user-defined graphs map out preferred zones such as “early weekday mornings between 6am and 9am” or “book reading times between 6pm and 8pm on weekends.”
For most users, these two distinctions are brand new. They must learn how SkedPal uses these and other components to produce an optimized schedule. An expert designer like the creator of the Octalysis framework for game design instructs: “During the onboarding phase, you train the users to become familiar with the rules of the game, the options, the mechanics, and the win-states.”
One of the first games to master in this phase is one I alluded to before: a “Daily Practice Game.” It’s simply the habit of entering the app on a daily basis in order to optimize one’s calendar. Mastering this practice is a small but essential win-state, similar to the one achieved by players of Farmville. They learn early in the game that there is a penalty for ignoring for several days: it results in the death of their crops.
Another game users could play is related to one of the first things they want to know when they log in — “Which tasks have become stale?” As I mentioned before, they need to take action to bring everything current. This daily action could be part of a “Task Recovery Game” which takes a certain amount of time and activity which could, in the future, be measured.
Occasionally, these changes are not sufficient and the player must also adjust a part of their task infrastructure, such as a time map. This could be part of a Framework Adjustment Game” in which the customizable components of the app are used to refine one’s schedule.
These three games are all built on the recurrent actions needed to keep an auto-scheduler running smoothly. As of today, I don’t know of any app which overtly promotes these games, but as I mentioned before, the potential exists to weave them into the onboarding experience.
My limited research shows that in the absence of explicit support, each of these games is played quite poorly at first. Over time, things improve as users teach themselves the required underlying behaviors. However, if they were offered as part of games at an introductory level, there might be more who make it to the next phase.
Some do make the transition and after a while, the behaviors become habits. When that happens they up-level to the next phase.
2. Games at the Features Phase
In the next phase, the user learns to exploit the program’s more sophisticated features. For example, version 2.0 of SkedPal indicates when there is a problem with your newly updated schedule.
It offers a Hot List which reflects the number of tasks which cannot be scheduled due to logical problems. For example, if you scheduled a task with a hard due-date for yesterday but didn’t complete it, the task would not be rescheduled. Instead, it would show up on the Hot List as an issue to be addressed.
A player could adopt a “Hot List Game” in which the user tries to minimize the total number of errors produced. SkedPal also indicates when a task is scheduled too tightly by adding a highlight to it in the form of a small icon. Avoiding these highlights could be part of a “Too-Tight Game.”
Another game is related to a feature I mentioned before: the way SkedPal distinguishes between your fixed and flexible tasks. The program synchronizes your fixed calendar in Outlook/Google with your internal calendar of flexible tasks. Unknown to some, it also allows the creation of additional fixed calendars.
I happen to use two fixed calendars. One acts as an appointment calendar with other people. The other is set up for personal appointments (with myself) and also to insert buffer times. Appointments with myself are solo tasks which occur at fixed times, such as workouts at the gym. Buffer time-slots are used to make sure I don’t over-schedule myself: each day, I have 1–2 hours of unscheduled time.
They are both used to ensure I maintain a balanced calendar which isn’t unrealistic. It’s a “Calendar Balance Game.”
Games at this phase and the one below it represent an important start, but I think they only scratch the surface. As I play them, I have generated lots of questions related to my usage.
3. Questions at the Self-Knowledge Phase
While games at this phase don’t actually exist, I continually ask myself questions which indicate that they are likely to be played in the future. Here are some of the questions I have about the schedule I produce each day:
– Is my current schedule too tight and therefore unrealistic? What is a lead metric of its quality?
– Are there items which are being postponed several times? Which ones are they?
– Are my time estimates unbiased?
– How often does a suggested next task actually get done? Why? Why not?
– What is the likelihood of a scheduled meeting actually taking place as scheduled?
There are also questions I have related to my skills as a “Total Task Scheduler.” I’d like to know:
– What is my track record for following my schedule and doing what I planned to do?
– What practices should I change to improve my skills?
– Are my high priority tasks being completed before those of lower priority?
– How often do I abandon my schedule to do something entirely different?
– Are there trends in my scheduling I need to be aware of?
– How many tasks in my system are actually dead and should be deleted?
I see each of these questions as the seeds of future games. They require their own metrics which could be collected by auto-schedulers. Put together here in their raw form they represent a start. I’m interested to hear from you in the Comments, especially if you have other questions that we should be asking at this phase.
What Designers Can Do
In the world of video-game design, there is an important role defined as a “Level Designer.” This person’s job is to create the missions, locales and stages of each level in a particular game. They possess a distinct skillset which is critical: they keep players engaged long after the Onboarding Stage, turning a player’s curiosity into a near-obsession.
Their skills could be applied to the phases I have shared above, to carve out levels of self-knowledge which I believe players of task management crave. As you may imagine, this job requires equal doses of psychology and computer science.
Unfortunately, I only know one or two developers who are thinking along these lines. Most are so close to their product and its functionality that it’s hard to step away to ask and answer these broader questions. They take time, plus no small measure of introspection.
But this is to be expected, according to Clay Christensen and other innovation experts. There’s an in-depth interplay between users and apps that must be studied over time. Answers and insights, they warn, don’t come from surveys or focus groups.
Instead, designers and developers must stand far out in the future, ahead of their users. Way out. It’s the only way to divine their needs, the jobs they are trying to get done and the games they want to be engaged by.
It’s a worthwhile effort.
Why should the most engaging software available to use be limited to trivial pursuits such as shooting angry birds or taking pictures of virtual characters?
Why can’t we put more effort into developing apps which help us learn skills which are important? Why can’t users become immersed in activities that help them get better, even as they complete important tasks? Wouldn’t we all be better human beings if we helped each other bring the fun of game-play to the stuff we really care about?
If there’s an epic quest someplace for task management app designers, and even game developers, perhaps it lies in answers to these questions.
This is the final article in this three-part series, published on Medium and on my website — http://2time-sys.com. If you are a developer/designer and would like to be notified when future articles and series become available, visit this page to sign up for updates — http://www.2time-sys.com/application-designers/.
If you aren’t a developer, you can also receive an immediate update by registering to download my article — “8 Edgy Ideas from Time Management 2.0” — at http://www.2time-sys.com/special-report-8/.
I hope you have found this series to be useful. Please leave me a comment on Twitter — http://twitter.com/fwade or on Medium’s Comments. I look forward to hearing your thoughts.
Thanks to my editors for their invaluable input: Marcia Oxley, Suzy Wilkoff, Jo-Ann Richards, and Tammy Emam.