Summary: Amazon's new e-book reader offers print-level readability and shines for reading fiction, but it has awkward interaction design and poor support for non-linear content.
The new version of Kindle, Amazon.com's dedicated e-book device, recently shipped with an improved display and various other upgrades. It now provides good usability for reading linear fiction (mainly novels), though it's less usable for other reading tasks.
As an experiment, I bought two copies of the same book: a trade paperback and a Kindle download. Alternating for each chapter, I read half the book in print and half on the Kindle screen. My reading speed was exactly the same (less than 0.5% difference), measured in words per minute.
Of course, one person reading one book is not a proper measurement study. So I can't say for sure that Kindle has finally reached the nirvana of equal readability for screens and paper. (Update: a bigger study found that the Kindle 2's reading speed was still slightly slower than printed books, though much better than old studies used to find for computer-based reading.)
When I was carrying Kindle through the house, I felt like a Star Trek character with a datapad. But when I actually sat down to read the novel, I became so engrossed in the story that I forgot I was reading from an electronic device. This fact alone is high praise for the device designers.
(See sidebar for analysis of Kindle's Out-of-Box Experience: unwrapping and "installing" the device.)
Awkward Interaction Design
Kindle shines in one area of interaction design: turning the page is extremely easy and convenient. This one command has two buttons (on either side of the device). Paging backwards is a less common action, but it's also nicely supported with a separate, smaller button.
The device thus offers good support for the task of linear reading — appropriately so, as Kindle's design is centered on this one use case. While reading, your only interaction is to repeatedly press the next-page button.
Anything else is awkward.
Most Kindle interactions are mediated by a small joystick called the 5-way , which lets you move the cursor in 4 directions; pressing down enables the fifth action. Repeatedly flicking the 5-way to move the cursor around the screen is extremely tedious. It doesn't feel like direct manipulation at all. The 5-way owns the cursor, not you, and getting the cursor where you want it requires a lot of work.
Interacting through the Kindle 5-way feels much like many mid-level smartphone user interfaces, though the 5-way is worse than a BlackBerry mini-trackball.
Furthermore, Kindle is slow. Every time you enter a command, it ponders the situation before acting. Even turning the page takes slightly longer than it should, and all other actions are definitely sluggish.
In short: Awkward pointing + slow reaction = a bad user experience that discourages people from exploring and attempting different tasks.
Poor Design for Non-Linear Content
Let's say you want to see The Wall St. Journal 's articles on technology. Where would you click in this screenshot?
Everybody I've asked said they'd click "Technology" to see that section's articles.
False. That click takes you only to the section's first article. To see the section's list of articles, you have to click the number in parentheses that indicates the number of stories. In this case, "(16)."
Intuitive? No way. In fact, after two weeks, I still make the mistake of doing the only natural thing: I click a section's name to see a list of articles.
(If Kindle were a touchscreen device or a mouse-driven UI, I'd further complain about the Fitts' Law implications of using the smallest click target for the most important choice. However, the 5-way controller makes it equally easy to pick a small or a big target; indeed, this is its only redeeming quality. Even so, it's still bad GUI design to make the most important commands tiny and thus reduce their visual emphasis.)
The usability problem with non-linear content is crucial because it indicates a deeper issue: Kindle's user experience is dominated by the book metaphor. The idea that you'd want to start on a section's first page makes sense for a book because most are based on linear exposition. Unfortunately, this is untrue for many other content collections, including newspapers, magazines, and even some non-fiction books such as travel guides, encyclopedias, and cookbooks.
So, the design decisions that make Kindle good for reading novels (and linear non-fiction) make it a bad device for reading non-linear content. Sure, Amazon designers could fix simple UI stupidities, such as the interaction design for a newspaper ToC. But doing so would simply apply a band-aid. To truly optimize the non-linear user experience, they'd have to completely reconceptualize the Kindle design.
Kindle Reader on iPhone
Amazon also provides an iPhone application that can display Kindle-format e-books. In my testing, the iPhone had 6% slower reading speed than Kindle 2.
(As with the comparison of Kindle and print, this measurement is based on too little data to be conclusive. It's a first estimate, pending a bigger study which would be a great topic for a Master's thesis. Hint, hint.)
One reason it takes more time to read a book with the iPhone than with Kindle 2 is the time spent turning pages.
When using my preferred font sizes, reading on the iPhone requires 21% more page turns than reading on Kindle 2. With the iPhone application's default font size — which I find a bit too big — iPhone requires 100% more page turns than Kindle 2.
The iPhone's page turns are also more awkward because it has no dedicated button for this purpose. To turn the page, you swipe a finger across the touch screen; this takes more time than simply pressing the button that your thumb is likely resting on.
Rather than requiring an entire gesture to turn the page, it would have been preferable to have users simply tap the screen. (In fact, the competing Stanza book-reader application uses a tap to forward one screenfull.) Just because you can use gestures on iPhone, doesn't mean you should.
The swiping gesture is the iPhone's generic command for scrolling and forwarding, and supporting it lets users transfer learning from other apps. Still, it's reasonable for a dedicated book-reader application to also support an accelerator shortcut; after all, when reading a book, next page, please is the command that users enter more than 90% of the time. (Accelerators are one of the top-10 software heuristics.)
The Kindle iPhone app is clearly a rush job that violates many application usability guidelines. For example, a slider moves users through a book rapidly, in multiple-page chunks. Very well, but without dynamic feedback, users can't tell where they are in the book while they're moving the slider. Page numbers are probably meaningless, but dynamic feedback could come in the form of chapter and section headings, or at least by showing the first line of text on pages corresponding to the slider's current position.
Help is nothing more than a link to an Amazon.com Web page that's optimized for PC display, thus violating a key guideline for designing good mobile user interfaces. Would it have been too much to hire a writer for a week to create a dedicated set of user assistance docs that display within the application itself?
The Kindle iPhone app displays pages with fully justified text (that is, flush left and right). This reduces legibility slightly, particularly for the small telephone screen's narrow columns. Stanza uses left-justified text only, which is a better choice.
Even though version 1.0 of the Kindle iPhone app has poor usability, it shines in its integration with Kindle and with the Amazon.com website. This is one of the first good examples of supporting multiple computers, and using wireless connectivity to do so.
You can buy a book on the website using your desktop computer and its superior browsing usability. You can then read the book on Kindle 2 at home or on your iPhone in the dentist's waiting room. In fact, you can alternate between reading the same book on the two devices; they'll automatically coordinate so that you pick up the book on one screen in the same position where you left it on the other screen.
Much has been made of a user's ability to buy books directly on the Kindle. However, buying books on a full-featured PC with the full Amazon.com site is much easier; I think they should simply eliminate Kindle's keyboard and use that space to make the screen bigger.
Still, Amazon deserves praise for the smooth integration between buying on the PC and reading your purchase on Kindle. Assuming you're in the U.S. and have wireless coverage all it takes is one click on the PC, and the book shows up on your Kindle without any special installation.
Electronic Books: Good or Bad Idea?
11 years ago, I wrote that electronic books were a bad idea. Has Kindle 2 changed my mind?
The two factors that convinced me were (a) equal-to-print readability and (b) multi-device integration.
I now think there's some benefit to having an information appliance that's specialized for reading fiction and linear non-fiction books that don't depend on illustrations and don't require readers to refer back and forth between sections.
For people who travel or commute using public transportation, e-books can be useful for a few reasons:
- There's less weight to schlep around (carrying 10 books on one device would have been particularly nice during my recent trip around the Serengeti, where we were allowed only 22 pounds of luggage for a two-week safari).
- No dirty fingers from newsprint.
- It's easier to turn pages while riding a crowded train or bus.
Being able to synch with an iPhone extends the usefulness scenarios. If you're waiting in the doctor's office, for example, and don't have your dedicated device, you can pull out your phone and make use of otherwise wasted time. Multi-device reading is one area where Kindle is better than reality: It does something a printed book can't do.
In addition to travelers and commuters, a second big user group will be old people and others with reduced vision. Big fonts are Kindle's other better-than-reality feature. (For more on this, see the sidebar, " Accessibility and Speech Synthesis.")
For people who are just reading a book at home, however, print will do just fine. For linear reading, Kindle offers no advantages and, for nonlinear content, has many disadvantages — so why pay $189 more?
Next Week: Kindle Content
Stay tuned for next week's column about how to write non-book content — such as newspapers and magazines — for Kindle.
Other updates :
- Update May 2009: Sidebar on the new, bigger Kindle DX.
- Update December 2011: Kindle Fire user testing