Mobile Community Design Research and design information for mobile community developers.

Hi everyone, in the last few years my professional work has taken me a bit farther away from mobile devices. While I still find the subject fascinating I am not coming up with as much material to post on a regular basis. Consequently it makes more sense to start up a new blog on more general tech and user experience design topics.

Please have a look at my new blog: User Design at http://userdesign.com

I recently attend the first occurrence of Ignite San Diego and gave a short talk on tablet computer design.

Description:
Currently mobile computing is dominated by laptops and phones, but there is a usability and utility gap between these devices which is currently unmet. I will demonstrate what market segments would use the product, what functionality should be prioritized, and address issues such as form factor and input/ouput methods that would be suitable. I will also address what Apple is likely to be doing with their forthcoming tablet computer, and why most industry reviewers don't have a clue how people will actually use it.

Video:


Some of you may note that this is very similar to a blog post I wrote back in mid-2007.

I was recently in Germany and had the opportunity to tour the brand-new Porsche museum in Stuttgart. I certainly recommend going if you're in the area. While it is a different kind of mobility than I normally discuss on this blog, it is important to draw from different design disciplines when designing new products. In this case, Porsche has clearly gone far beyond making high-performance machines capable of doing airplane speeds on the ground, and long ago entered the realm of fine art and designing for exotic experiences. So perhaps those of us designing mobile phones and wearable computers of the future could learn a bit from how Porsche designers think about their products, and how they have managed to give consumers a whole new definition of how they can be mobile.

One of the long-standing problems in the design process has been the lack of digital tools to support early GUI design work. Surprisingly, one of the most useful tools to come about recently to aid this process was the digital camera. The point-and-shoot camera (or the camera phone for impromptu work) meant that any design that was scrawled on the back of a napkin or the nearest whiteboard, could be captured and distributed to others for remote collaboration. This works great for initial concepts, but it can become tedious as the design becomes more solid and where re-drawing complex parts of the design repeatedly becomes inefficient.

Up until now there have been very few digital tools that supported low-fidelity (think napkin-like resolution and speed) prototyping. People have often told me that Visio, Visual-basic, Axure, Photoshop, Dreamweaver, etc, can fulfill this role - but they can't. The problem is that they are either too high-resolution or too code-focused. What is needed is a tool that is very very fast to use. It needs to be informal. The user shouldn't have to worry about exact alignment of objects, and they shouldn't have to edit complex tables of variables for components. They should just be able to *think* about an object and have it there on the page in the 2 seconds it would take to draw it.
So, I've been waiting for this type of tool to come into existence, but the problem is that most programmers don't think like designers. Thus, the tools that they create don't have 'laser-like-focus' on the requirements of the intended users. Fortunately it appears that Peldi, the developer of Balsamiq does understand what is needed.

Here is a little fictional iPhone app concept that took me about 2 mins to "draw" in Balsamiq. This tool is very interesting in that it has a very minimalistic UI and can be installed from the same executable on either a Mac or a PC.


The software uses a nice hybrid of direct object manipulation and programming. The programming side of it is extremely minimal and most users won't even realize what it is doing behind the scenes. Instead of having pop-up boxes with fields to edit to modify objects, there is a simple scripting language to modify objects. In this case, it is a list of table labels separated by commas. This has the advantage of being extremely fast to modify. This is critical since prototyping essentially happens at the speed of thought and needs to keep pace with collaborative conversations.

Another awesome feature that should be included in more applications is a hybrid of search and object manipulation. In this case the user wants to *very* rapidly get a new object on the page. Since they have used it before, they know the name of it. By typing 'note' or other search terms, they can have that object automatically dropped onto the active page.


The real brilliance of the app is in what Peldi has *not* chosen to do with it. The objects aren't designed to look perfect - they look like sketches and set viewer expectations accordingly. There aren't a huge number of options for each GUI component. There is also nothing standing in the way of rapidly getting the appropriate objects laid out on the page - which is the point of the application.

It's a nice simple tool, and there's something to be said for not loading it down with features. However, things that might be considered:

• The GUI component library is awkward, and horizontal scrolling is bad. Possibly smaller icons and a right-hand vertical display of objects might work better.
• More GUI objects are needed. For iPhone development, many of the standard components look different and some of them act differently. Being able to import additional widget libraries, or to allow expert users to code their own components and share them would be great.

This is a great tool and I'm planning to start using it as a possible replacement for whiteboarding, or possibly as the immediate stage after whiteboarding. Whiteboards still support extremely rapid generation of ideas without the need to select objects types prior to creation of objects. If I were to think about further optimization around the use-case of 'rapidly get desired objects on the page' I would look at how to start expressing ideas without prior selection of object types - which is not an easy problem to solve.

You can download a trial copy of Balsamique or try a web-based version here.

One of the interesting things about doing research (as opposed to industry development) is that you're supposed to be ahead of the leading edge products currently being produced. The idea is to think up things that would be useful, but which currently seem impossible for one reason or another.

Example: Backpackers had the problem of being able to bring enough music with them on the road - there was never enough storage, and if they lost the device, then they lost their music library. A good solution was to "stream" their home music library to them while they're on the road, via a high-bandwidth wireless connection. At the time I developed the concept, there was nothing available. There was satellite radio, but it was cost-prohibitive, and the units were large and could only be installed in cars. There were no mobile devices capable of streaming music. Now the iPhone has several apps doing things similar to this: Pandora and Orb. It still may be cost prohibitive while travelling internationally due to international roaming rates, but that won't last long. In 2006 personal music streaming seemed like science fiction.

The challenging thing about doing mobile research is the pace at which the technology is developing. If you complain about there not being enough storage space in an MP3 player, sure enough in another year it's not really an issue anymore. So if you're researching mobility, you identify some really interesting concept and run with it. But it's likely that before you finish the Ph.D., somebody's gone and built the thing you dreamed up - it's both exhilarating and highly irritating at the same time.

Now that I am getting into iPhone development, I am suddenly rather interested in how many of my previous mobile product concepts have already been created; or to put it another way, what was the rate of mobile tech change in the last 3 years? The following isn't statistically significant, but it provides an interesting sample.

Method:
In my thesis, I included an appendix of 46 product concepts. I recently reviewed all of these concepts and rated them High, Medium, Low/none for the variables of 1) Market Interest, 2) Development Interest, 3) Already in Market.

Results:

• 23 of the 46 concepts (50%) rated a High/Medium for 'Already in Market'.
• There was a high correlation between being rated 'Already in Market' and High on Market Interest - so suppliers are responding to demand and creating things there is an obvious need for. These included things like maps, media storage, discussion forums, find resources in the environment and telling people what you're doing (Twitter).
  
• There were still a significant number of mobile device concepts that really haven't been completed yet, and which still have high Market Interest. I'm not going to give away the specifics, but they include things like:

1. resources that people want, but which aren't digitized.
2. methods of creating groups and pairing people.
3. ways to digitize money and other important paper items.
4. things that require larger amounts of typing while mobile.
5. specific data from local businesses.
6. awareness of social properties of geographical regions.
7. timely awareness of critical information of personal interest
   
8. niche market mobile shopping and delivery
9. problem solving assistance

So from my small sample, 50% of cutting edge ideas get put on the market in a 3-4 year time span. Those ideas are probably under development for at least a year or two before they become publicly available. So that means: when you come up with a cutting edge idea - you're probably not the only one having it. However, there's the other 50% of the ideas that either nobody had, or nobody could execute on in 3-4 years. Typically those problems, such as the nine items listed above, are a little more difficult to conceptualize, and more difficult to build. If you're looking for a topic for a start-up, and don't want a lot of competition out of the starting gate, maybe this is a good space to look in.

Increasingly I am of the opinion that use-cases (user tasks) are not sufficiently considered during design processes, particularly in mobile devices. Sometimes the use cases are simply forgotten (the user can't do something they should be able to do), but more often they haven't been weighted appropriately (things that should be easy are hard and vice-versa).

I take a lot of notes, and it's interesting to see how well cutting-edge technologies support this basic and fundamental activity. The following video demonstrates a light-weight comparative task analysis of five technologies across three use-cases.

Technologies:

1. Sticky notes
2. Notepad / artists drawing pad
3. Sidekick 3
4. iPhone 3G
5. Fujitsu U810 Tablet computer

Use-cases:

1. Write down an idea quickly
2. Write extended text entry
3. Find an idea from a previously entered note
   

Comparative Task Analysis of Note-taking on Five Technologies
14 min.



Conducting a comparative task analysis is a useful way of ensuring that your new advanced technology actually performs better for typical tasks than the older and low-tech technologies which it is hypothetically supposed to replace. Note how poorly the cutting-edge tablet computer performs for this basic task - it's no wonder they are not selling well. Despite incredible potential of the device and form factor, tablet computer manufacturers are focusing on RAM and processor speeds instead of creating a product that is optimized for taking notes (hint: this is a market opportunity).

Labels: comparative task analysis, design, HCI, human computer interaction, interaction design, iXd, mobile, note-taking, notepad, technology, usability, use cases, user experience, ux

Now that I've had a chance to break in my iPhone 3G, I think it is time to review what Apple should fix in upcoming OS and hardware versions. Looking back on my previous review of the iPhone 2G (before it had been released), most of those comments were reasonably accurate. What is even more surprising is that they are mostly still valid for the 3G version. So apparently Jonathan Ive and John Geleynse aren't reading my blog. =)

For the record, some of what I point out below is fairly minor details that won't matter to the average consumer. However, for UX professionals and students, this critique should raise the bar a bit on what mobile devices could be - even when we're admittedly reviewing a device that is best-in-class in many ways.

Top 3 Greatest Things about iPhone 3G

1. The app store
   This phone would be far less useful without 3rd party apps. Having a dictionary, travel tools, voice recorders, etc makes this a great pocket knife for a wide range of users.
   
   
2. Seamless Gmail contacts integration
   Past phones have always made me manually enter contacts, and then they would get out of sync with my mail programs. Now they're all in one place, and a change in one goes to the other.
   
   
3. Free Exchange integration
   This was marketed by ATT as part of the business plan, but it appears that it doesn't cost anything extra for customers on a normal plan (crossing fingers)
   

Everything else that is less than perfect

(Ranked least to most important)

17. Mobile Me - why? - Mobile Me sounded great. I signed up for the trial; then I realized that I already had my email on Gmail, and I already had my contacts synced with it, and I already had my work calendar via exchange, and I didn't need a web-based hard drive. So what's the point for $100/yr?

16. No help system - One of the main usability heuristics that should never be violated is the existence of a help system. Even if your interface is usability tested and clean and simple, you still need it. There will still be advanced features people need to find. There will still be troubleshooting necessary. The iPhone needs a searchable help system, or at the very least, a desktop icon to a web-based iPhone documentation formatted for the iPhone screen.

15. Tedious customization of home screen - I had to do a web search to find out how to customize the main screen. I was scared to death that it simply didn't have the function. Hidden (completely hidden) functionality isn't a great idea. I still think it's awkward, even after finding out how to activate it. Particularly difficult is the action of moving an icon between 'panes/pages'. The user gets used to the physics model used elsewhere in the OS and expects to be able to "flick" an icon to the next screen, but alas this doesn't work.

14.Buried settings screens - The settings has an information architecture which is far too deep.
Settings > General > Network > VPN > Add Configuration
Trying to remember where wi-fi vs bluetooth vs VPN screens are is enough to give one a headache. Some of these things are actually frequently used if you want to extend battery life, so this is clearly not the right format for these features. (this can be solved by making the tree broader).

13. No external access to notes - The sidekick 3 still manages to get some things right that Apple hasn't learned from. For example, allowing desktop/web access to note content. I can't access my iPhone notepad content in iTunes, or in MobileMe. Some of us track multiple to-do lists and notes. The lack of a way to import or export this, or view it on a desktop or via the web is a clear unsupported use case.

12. GPS turn-by-turn? - There's a GPS. Why doesn't it act more like my Garmin? Am I really going to have to carry both my Garmin and my iPhone when I take that business trip to Europe?

11. Confusing standard widgets - I watched the Apple podcast by the Apple UX group on UI design (User Interface Design for iPhone Applications). It's actually pretty good, and clearly focused at introductory concepts for an engineering/programmer primary audience. That said, some of the basic widgets they are recommending developers use have inherent usability problems. One of the biggest ones is the 'two part' row items without any visual warning of their action or of their split personalities. Another big one is the 'long alpha list' with a miniscule search button at the top which is extremely close to the add contact button. It's nearly impossible to get to the search field when I'm sitting still, much less walking and trying to dial a contact.

10. Audio plug on the wrong side - Every time I put this phone in my pocket while listening to music, it comes back out upside down. This is because the device points head-down in the pocket and the headphones point back up toward my ears; but when I pull it back it's natural to keep the headphones at the bottom, which is of course the wrong side. Also, when I have it plugged into the charge cable (because it didn't come with a dock), the headphones point toward me, the charge cable points away, and the screen is upside down. Ignoring the undoubted hardware constraints, the headphone jack should be on the other end of the device. The current design would be OK for a generic dock, but users will primarily be using this on the move.

9. Modal mtg alert before end of call - Some of the Exchange and calendar alerts have gotten a bit too pushy (too modal to be precise). Several times I have gotten off of longer conversations to find a dialog demanding that I decide to 'Accept, Reject, or Maybe' a meeting request. I am apparently supposed to make this decision in a split second, before being able to hang-up my phone call / switch calls / do anything else. This device needs to be designed to have _everything_ be able to be interrupted or set aside for later.

8. App updates don't retain location - The auto-update features are great, but it would be nice to be able to set them to auto-update for certain applications. Furthermore, they need to remember their location on the desktop. If you move them to screen 4 of 6 in the bottom right corner, they need to go back there after you update them. Currently they drop in on the main screen like a new application.

7. Hard physical top edge of hardware - I think somebody got a little too carried away with this device being an Internet tablet. It doesn't actually feel comfortable when used for voice calls. Call me old-fashioned, but I still don't like using bluetooth headsets and talking loudly into thin air for all of my calls on noisy streets. Some more rounded corners on this device would make it more comfortable to have against the face. And I might further add that a little more thickness would be have been justifiable if it had allowed a larger battery to be used. 8 hours really isn't enough for a business phone.

6. Non-ergonomic headphones - I just don't get it. The standard apple headphones don't stick in my ears. They fall out. Every time. I can't be the only one. Considering this is a premium music device, is it really that difficult to ship with earbuds? I won't even mention the lack of otherwise-standard A2DP support to allow stereo bluetooth headphones, which would have been the truly elegant (i.e. wireless) solution.

5. Back button - where is it? - This is another thing the Sidekick got right. The iPhone is very modal. Only one app running at a time. Moving to a new app generally leaves you stranded - it has no memory of where you've been. For example, if you are in Contacts, and you click the address of a contact, it predictably brings you into Google Maps, showing that location (which is great). But what do you want to do next? Well, go back to editing or reviewing the contact of course. To do that you need to do this:
1. Home button > 2. contacts > 3. click to search > 4. type in name > 5. select name. Now you're back to your contact. Wouldn't a back button (probably hardware) have been easier. Currently the iphone is very IA focused instead of task focused. One more small hardware button on the top left really wouldn't ruin the aesthetics that much. Another option would be a universal gesture to move back to the previous screen.

4. Undo purchase option - where is it? - This is dangerous. People have been accidentally buying $1000 iPhone apps. Laugh if you will, but the next time you're using your iPhone drunk at a bar, make sure you click carefully. Good design should prevent errors. The Amazon Kindle has a great feature for its online ordering system. You can buy a book (with no confirmation screen), and if you don't like it within 10 minutes or so, you can undo your order. Undo is much more elegant than warnings and it actually solves the problem instead of just making it less likely to occur. There could be a standard undo gesture, that is universally supported.


3. Relying on iTunes - It's oftentimes unclear what gets updated over the air, and what gets updated via iTunes, and further it's not always consistent. Some apps are "too big" to download over the air and require iTunes for installation. I've experienced some technical problems where some apps wouldn't download, and it tells me to connect to my desktop computer. However, my gmail contacts sync wirelessly. My podcasts don't update until I connect to the computer. I know Steve likes the Digital Hub concept, and it was good for a while, but the end is nigh. We travel for business. We use multiple laptops. We plug in at work. We go on vacation. Do you really think we want to wait until we get home to get our new podcasts? Do we really want to wait for application updates? Maybe Mobile me was intended to solve some of this, but it is clear that Apple doesn't get the fact that we no longer want our PC to be the hub; instead we need wireless networks to keep us securely connected (even if we don't know it yet). Amazon has thoroughly and elegantly figured this out with their always-on and free Whispernet service.

2. No universal horizontal rotate support - Please provide an operating system that natively supports different handling preferences. The keyboard still has unacceptably high error rates (see below) which are greatly improved in the very few applications that support the wide-keyboard mode. Most applications (with the exception of Safari and many games) don't support the landscape screen rotate mode. It should be mandatory that all applications support both modes. Content can re-size, or be previously designed for both modes. Some types of content are simply much easier to use in landscape mode. If I could rely on a horizontal keyboard, I would probably rarely use it in the portrait mode. Why this wasn't fixed in the 3G version I don't' know, given that it was a glaring problem in the 2G version.

1. Keyboard still sucks - Sorry, but I've heard the "but after it adapts to your spelling and you learn where to press the screen" arguments and they just don't ring true. At the end of the day it's an awkward keyboard. Anything with a dictionary that suggests words is a bad idea on a small screen device because it requires a large amount of attention from the user (increased cognitive load) to monitor it and make decisions. And we all know that mobile users have even less of that available than busy desktop users. It's also even more likely that the dictionary will auto-complete the wrong word for you, resulting in confusing or embarrassing messages - so it's actually creating additional problems. I've been using this keyboard for 2 weeks and I still regularly mis-type on it. I used to be able to write short emails on my Sidekick (same for the Blackberry), but on the iPhone, I have added the "Sent from an iPhone, excuse the brief message" signature file. It just takes too long. I'll wait till I get back to the desktop. The easy design solution is to _always_ allow the landscape mode, which does result in less typing errors due to the larger letter key areas. The hard solution is to offer a slim sliding hardware keyboard. Apple, if you really want this to be the mobile web device of the future, it absolutely has to support extended text entry - this is a chance for innovation.

Update: 8/10
I forgot three things that should have been on the list:
A) No copy/paste. It is a frequent use case that on the web or in various forms, that you want to copy a bit of a page and use it elsewhere. This is a general purpose productivity-enhancing feature. The sidekick supports it through an hidden shortcut.

B) Poor design of mp3 player interface. Many of the widgets to control location in the track and see the entire track title are very small and hidden. Trying to back up to a previous location in a track (common for audiobooks) is very error prone. This interface has focused too much on showing glitzy album art, and hidden commonly used features as a consequence.

c) No support to 'background applications'. While it is commendable for Apple to not force memory management tasks on the user, the optimal solution is probably not 'only one app at a time'. Clearly some apps will only be useful if they can always be on. For example what if your phone application could only receive calls when you were in the phone application? There needs to be an option to run applications in the background in a controlled and safe manner.

Labels: 2g, Apple, interface, iphone 3G, mobile, review, ui, usability, user experience, wearable