01/28 v0.1 This post is an early draft based on a tweetstorm and will be overhauled soon.
10 years later, iPad is both great and as John Gruber among others say, inscrutable. iPad's app-based springboard isn't a good organizing paradigm for even light complexity work, and it's holding iPad back.
"Work" has a Minimum Complexity
Anything we call 'work' usually involves 2+ things at once. Think of booking a flight. You need to look at the options, scribble some notes, and look at your calendar. You're already at 3 things. Cooking is looking at recipe + calculator. Active reading is book + notes. Etc.
A lot of even rather simple work just has a certain natural complexity to it; think of it like a desk with various papers and books on it. You may need to reference multiple things at once, bring them together, scribble on one, etc. An app-based model is terrible at this. Here's a few reasons why:
Problem #1: Feature Creep
These days 'Apps' are not just single-purpose tools or even multi-purpose products. They're also brands and businesses. Imagine a wrench you bought was also a brand you had to interact with to use, had its own instagram account talking about what bolts it turned today, and required a subscription. That's the state of most apps today.
The direct coupling of app <-> business create the conditions for feature creep. What is ideally a desire to do a single thing is at a bare minimum a task that requires the following:
- Discovering and downloading the app on the App Store, OR
- Finding where you left the app on the springboard
- Navigating boilerplate UI to contain the single unique thing the app does
- Doing or looking at the thing
- Usually then navigating to another app to do a related thing
But the competitive business environment of the App Store creates strong business incentives to:
- Add duplicative capabilities to solve for minimum work complexity, so you don't have to navigate to other apps as much to do the thing
- Create eye-catching app branding so you can more easily find it next time among a sea of other apps on the springboard
- Add ways to bookmark, tag, or otherwise organize data in app
- Add user accounts with required logins to own user relationship
- Offer in-app upsells
- Add permissions prompts for capabilities you don't want
- Announce changes to app in pop-ups
- Send notifications to re-engage and stay top of mind with users
- Create non-standard UI to differentiate app from others, making it harder to navigate
- Update the app frequently so when you go back to it, it's changed
Then costs rise due to above, the app switches to subscription model to keep it afloat.
Most user needs are not worthy of this level of complexity. It's like if only books were publishable and you'd written a short article's worth of ideas - you'd probably add a bunch of fluff to get it to book length, and readers time would be wasted.
If what you need is just a view of some particular data or a 'card' worth of a few simple actions while doing work elsewhere (e.g. translate phrase, convert units, see if there's an update to a discussion you're in, etc), you have to go through a complex multi-step task that at each step tries to get your attention.
Problem #2: Inventory Management
Another reason app-based models suck is app proliferation. People download apps but rarely delete or organize them (organizing apps in particular is a painful experience in iOS). Over time, the experience gets worse and worse. It's like inventory management problems in games; your bag is full and you pick something up, now what?
This problem is especially bad because the web is also a place with apps & content. So users effectively have two entirely different worlds where they work, which creates significant ongoing cognitive load.
Problem #3: Compartmentalizing Work
Another reason still is how we compartmentalize work, and its associated data.
Compartmentalizing data is important because as computing starts mediating every facet of our lives, we're inundated with data about each of these facets. And if that data isn't organized in a way that makes sense, the psychological result is feeling scattered, overwhelmed, and lost.
In the physical world we compartmentalize spatially across a very large multi-location interface, e.g. stuff on our desk, in the living room, etc.
In a document-based world, we compartmentalize with files, which are rough analogs to paper documents.
In an app-based world, we compartmentalize by app (safari, keynote, notes, etc).
The problem with using apps as containers for our work is at least twofold:
- Some apps we use just for one project, while others intermingle data from several.
- Each app has its own way of organizing data, so once you need to do project-level work, you effectively have to find & open a data on that project 2+ times.
Using similar apps for different life purposes is one adaptation that many users intentionally or unintentionally use due to the limitations of the app model. For instance, one writing app for work and another for journaling. But this is a hack.
What this all means is that most of the time when you're doing work, you are fighting the very nature of the system, which can only conceive of work as something that resides in a single app.
"Multitasking" is no Panacea
The app dyads of iPad multitasking improve this a bit (if awkwardly), as do share sheets. However even if multitasking UX were great, the problem would remain that apps themselves are not the way to organize projects. They're tools, not structure.
Now at this point you may be thinking I'm advocating for a document-based paradigm like MacOS. Nope! Too freeform, too much folder-rope for users to hang themselves with.
We also live in a world of data & databases, but those are too intangible as a UX paradigm. What then?
Toward Human-Centric Organization
We've got it backwards! Apps, documents, and data are all computer-centric paradigms. So is 'multitasking', which is really just 'multiple apps' as one task can involve 2+ apps. These all just expose the computer's innards, they're not designed for how humans think about the world and their place in it.
Let's explore real quick how humans think, how that intersects with the organizational paradigm of our computing devices, and discuss why we need to align those across all computing experiences including iPad.
Humans think in concepts/interests, spaces, relationships. When we engage in these it tends to be via a mix of actions we take, content we encounter, things we create, personas and behavior codes we adopt, motivations we fire up, etc. Its highly multi-modal, multi-sensory, multi-faceted. Lets call these Intention Spaces.
Think about when you get home. This space contains:
- Actions (lights)
- People (family, roommates, etc)
- Conversations (what to do about the couch)
- Tasks (pay bills)
- Cues (become sleepy)
- Content (mail)
- Patterns (wake up at 7)
- Identity (be 'dad')
- Needs (sleep)
- and more
The physical and mental space of home helps relate and delineate these. And the same is true for all sorts of other areas of your life:
- Physical Health
- Mental Health & Wellbeing
- Hobbies & Interests
We've evolved to take physical / relational / sensory cues together as the organizing paradigm for our lives. We go to places to be with people and do things. Anything other than this is going against our nature.
Not only are these intention spaces our organizing paradigm, people all have roughly the same ones. We all generally need community, good health, shelter, hobbies, family, work, solitude & introspection, etc. These are our rocks and where we apply our energies in life. When we have problems with these, we become distressed. And the best way to apply these energies is also similar across people. We're happiest when we're focused on the present, and we do our best work when we have the things we need in front of us, and nothing else.
As Devices Improve, We Lose Intentional Space
But right now we're not anywhere near that ideal. To work on something using a few tools involves complex UX navigation through a huge assortment of bright colors, unrelated life data, inconsistent interfaces, ads, predatory internet feeds, etc.
As computing has become more ubiquitous, intentionality has become harder to achieve, because the separation between our life activities has eroded. Our devices, once relatively focused but limited tools among many, has become a sprawling App Vegas that shapes and limits and controls our very approach to life.
Our devices are not built around our intention spaces; they're not structured to reflect our needs. As devices expand to mediate everything in life, things in isolation may become more convenient but being intentional will get harder.
This is essentially 'the medium is the message', the medium is apps, and the message is you can only reason about your intention spaces through the lens of what any individual app can do (minus some high friction app-to-app UX).
Bringing this back to iPad, the symptom is that multitasking sucks and people struggle to get iPads to do more complex work. These are more acutely felt on iPad than iPhone. And the root cause is the app-based organizing paradigm of iOS.
A Proposal: Spaces
I propose a different approach, organized around psychological needs, as represented by intention spaces (Spaces for short). These are collections of tools and data set up around aspects or projects in your life. Some important details:
UI would pre-populate with templates for above mentioned intention spaces, since they're universal human needs. System would understand conceptual differences between major categories like home, work, community, hobby, etc.
Data would be pre-organized by space. In other words, an app like Notes would inherit home, work, health, hobby X, etc folders from spaces. And when Notes was opened from a space, it would just see that space's data.
Navigation between spaces would be the top-level UX (not springboard). This would default you to focusing on one thing. Breaking focus would be the exception, not the norm. Possibly app switcher style UX, but could be zoom UI or something else.
Switch spaces automatically based on certain triggers or cues, including location types from Apple Maps, calendar, time of day, etc. At gym/doctor office = Health; got home = Home; 4pm soccer practice = Family.
Navigation within spaces would show a mix of widgets, files, and apps, bringing data foward as much as possible. Sticky notes, a youtube video, tasks, reminders, simple tools, intelligent agents, etc would be right there where you need them.
Some kind of tray UI scratch space to capture the variety of content you're juggling when working on something. Ideally Pencil-capable. Tray may also offer launcher for apps/tools you need in the space.
ML content analysis helping preserve space organization by looking out for whether work you were doing in one space better matched content of another, then suggesting moving it over. E.g. writing note about hobby while in health space.
Search would scope to the space you're in by default, with option to expand to system. Siri would make some assumptions based on current space.
Ability to create additional spaces for things like projects, hobbies, etc. Offer a space template library for ease of setup & system understanding. Not all need to last long (e.g. planning a trip, running an RPG campaign), and can be archived as a bundle when completed.
Allow one level deep hierarchy of spaces within spaces, like a space for an individual member of the family within Family, or a specific project within School space. These would inherit rules from parent space.
Allow shared spaces, such as among family members, hobby friends, etc. This replicates how we already use shared physical spaces, and solves perennial problems like family photo albums, access rules for different people, large message threads, etc.
Space-based Screen Time limits. Imagine only School space is available to kids during homework hours, or if you want to unplug from work after 8pm. Control spaces by time of day that you want to interact or avoid. Set space timers, etc.
Space-based notifications, reducing distraction until you intentionally shift focus. Social notifications matter, but not equally, so ability to whitelist/blacklist messages from individuals on per-space basis to batch socializing.
Better Siri suggestions scoped to space, e.g. its time to drive to work map widget appears at top of home space, tapping brings that up and also switches you to work space.
Developer Benefits of Spaces
For developers, new abilities to build simpler widgets without trappings of apps. Widgets would do simple things without the bullshit, sort of like watch apps or enhanced Shortcuts. Intelligent agents would be great.
Certain apps may be listed in app store as belonging to certain space(s), e.g. a pregnancy app for health and family spaces. Means it won't get in the way when doing work in other spaces.
Better distinguish media apps from utility apps. Treat media apps like other media, more easily purged, discoverable and usable from Spotify Discovery like UI, etc. This is most especially needed with things like Apple Arcade & subscription bundles.
Scope permissions to space, to reduce apps getting access to more data than they should. Support 'allow location when using space' scoping and other similar ones.
Encourage widgets for the specific things users are interested in, pulling that data to the fore. E.g. A particular Facebook Group for one of your hobbies, a chat thread for community, opentable widget for reserving at your favorite restaurants, etc.
Considering the Path Ahead
When designing software, you're inherently making some interactions easier than others. The friction that different paths cause determine how users wind up using the product. Ideally, desirable things are easy to do, undesirable things hard.
As you can see, a ton of possibilities open up when you think about organizing life by human psychology instead of by what apps you're juggling. It gets easier to set up your life intentionally and manage the noise that devices sometimes bring.
Hopefully Apple and/or Google is working on something like this. Our devices OS affects our human OS, we'd be silly not to acknowledge that nowadays. Wouldn't it be great if we could update our human OS even just a bit? I think this is a way we can.
One must bring a great deal of humility to an exercise like replacing iOS springboard, due to its ubiquity and success. Big changes like this need to be rolled out over a few years. Apple's the best at that kind of long game.
One last bit - this concept isn't limited to smartphones & tablets, but all devices. It will pay dividends on futute platforms like AR where the experience is even more place-based and the UX has very limited input.
Thanks for reading!
If you like this please say hi to me on Twitter (where this was originally posted). For other inquiries you can email me at my firstname lastname at gmail.
Subscribe to Nick Punt
Get the latest posts delivered right to your inbox