The Heretech

Stock photos are the Muzak of the Web

Stock photos on corporate web sites are lot like Muzak in elevators, or the distracting TVs left on in restaurants and bars. They're portions of the user experience that normally attracts none of your attention. If you do stop and think about them, they're just plain weird.

Why does anyone think I'd like to hear a mash-up of Bruce Springsteen and Mantovani? Why is the TV turned on to a soccer game, or for that matter, turned on at all? And why are all these people in the stock photos shown here so damn happy about using their computers?

Everywhere I've ever worked, people stare at the computer screen with the grim determination that befits the job they're doing. This is your job, not a vacation in Electric Lotusland. We can guess the reasons for the goofy grin on the face of the guy on his couch: he's probably watching YouTube videos of people driving their cars into ditches, or poking alligators with a stick, or dissecting The Phantom Menace as the worst movie ever made. But what about the woman on her couch? She must be happy that she'll be doing something else soon, because there's no way she can work an eight hour day on her stomach.

On a more serious note, I just wonder about the value of the investment in these photos. They feel more like the lint that inevitably collects around web pages, and less like some devious graphic cue that reaches into my reptile brain to tickle my happy centers. Me on IBM web site, me see happy people, so me am happy...

There may be no good reason. Once, when the music in a department store was cranked so loud that it was giving me a piercing headache, I asked a clerk why the store inflicted that pain on the customers. He shrugged and said, "Some woman from Corporate comes out here every once in a while and yells at us if the music is too soft." Which is the same as saying, "It's loud because someone thinks it should be loud," which is no explanation at all.

[Cross-posted at The Forrester product management blog.]

Unscheduled web design rant

Recently, the online magazine Salon changed its design. Now, it exemplifies many common  mistakes in organizing and presenting web content.

For starters, try making sense of the front page without getting a headache. A bajillion little boxes, all seemingly placed at random, each of which is supposed to encapsulate some category of content. The problem is, it's not always clear what those categories are. For example, the meaning of TMI and Open Salon are not self-evident.  (Hint: One of them leads to blogs. So why not call the link Blogs?)

Second, repetition rules, adding to the confusion. The same story appears in multiple boxes. The tabs at the top of the page repeat some of the categories of little boxes in the middle of the page. (But not all.)

Third, the design forces me to click more often than I should have to. Every time I start reading an article, I hit a couple of paragraphs down labeled Continue reading. But I'm already reading the article, dammit, so why make me click one more time to see the rest of it? (Perhaps because the page isn't visually confusing enough, so the web designers had to cram more article snippets into that space.)

Fourth, there's the confusion of ads. It's not that ads are a vice per se. However, when they're not clearly differentiated from the rest of the content--often because their positioning changes with each page--it takes your brain an extra cycle to separate the ad content from all the rest. Only a few ergs of mental energy necessary, perhaps, but in conjunction with the other design problems, it adds up.

Fifth, juvenile aesthetics. For a magazine that's supposedly pitched at literate news junkies, the graphic message is often, "You're all a bunch of teenage boys at heart." I don't know how many times I've clicked on a page to see someone's bare ass, or snarky photos of ugly people, or titillating headlines like "Mom's First Nude Shoot." None of these items individually is necessarily objectionable, but again, when you add up...

And finally, the layout keeps changing. If the editors are trying to fix the problems with the site, kudos for their good intentions. However, the more they tinker with design, the more confusing it becomes.

Many newspapers and online magazines have embraced many of the same design principles. By which I mean, the same mistakes, such as the random, ugly jumble of "contentlets." Just because The Washington Post jumps off a bridge doesn't mean you have to.

And when Mr. Customer gets angry, PEOPLE DIE!

Not every unwelcome change is necessarily a fiasco. It's how you handle the transition that determines, to a large extent, whether people perceive it to be a fiasco.

Case in point: Oracle's transition from its creaky but familiar Metalink support site to the new support portal. I'm sure that anyone who has ever been involved in the transition from one web application to its replacement has learned just how many unexpected glitches can creep into the process. Certainly, Oracle has faced similar scenarios, such as when it switched the main corporate web site from a custom-built content management system to a new infrastructure based on Oracle Portal.

Unfortunately, the corporate web site, where people go to find high-level marketing information, is not the same as the support web site, where people go to get problems with mission-critical applications resolved.

Some of the new support site's problems, such as slow performance, are to be expected when you're putting a web application through its real-world paces for the first time. Others, such as the Flash UI that won't work in organizations that block Flash, are regrettable design decisions that should have been avoided. The team behind the new site can certainly fix these issues.

However, as Oscar Wilde once noted, "In matters of great importance, style, not substance, is the vital thing." Considering the style with which Oracle has handled the blowback from the My Oracle Support launch, they might take a few moments to ponder Wilde's maxim. For example, this snarky post from an Oracle support manager is getting linked pretty widely. In fact, the phrase, "Get in front of this one…seriously," might appear on a few T-shirts and bumper stickers, once the dust settles.

To their credit, people at Oracle are trying to both be responsive, and appear to be responsive. This detailed post on how Oracle handles usability issues in the My Oracle Support portal may include way more information than the average customer wants to read. But heck, when you're explaining the categorization and prioritization of usability issues, you're definitely being transparent to your customer.

Nonetheless, some Oracle customers are still alienated by the It's coming, deal with it message of the initial launch. And certainly, the industry is littered with examples of (allegedly) improved user experiences that either didn't improve anything, or that weren't marketed effectively to the people affected by the change. And, of course, there's the unfamiliarity factor of any new UI, which offends people who have built their daily work around the old, familiar UI. Why Oracle didn't heed these cautionary tales is anyone's guess.

Whenever anyone designs a user experience, they're well advised to remember another Oscar Wilde maxim, "Every portrait that is painted with feeling is a portrait of the artist, not of the sitter." The natural tendency is to build a UI that makes sense to you. Whether or not it makes sense to the person using that UI is another matter entirely.

[Cross-posted at The Forrester product management blog.]

Everyone's a super-user!

It's bad enough that some upgrades from Windows XP to Windows 7 can take almost a full day. What's worse is how Microsoft tries to fudge the problem, claiming that only "super-users" will face this new education in the virtue of patience.

A "super-user," in their definition, has 650 GB of data and 40 applications. Now, take a look at your system and ask yourself, Are you a super-user? If you like music, you probably have 60-80 GB of music already stored on your hard drive. Install Microsoft Office 2007? You've probably consumed another 60 GB minimum of Microsoft's own applications. Take many pictures? If so, you're eating up a lot of additional hard disk space every month.

In other words, it's easy for the average consumer to approach 650 GB rather quickly. Computer games are notorious space hogs: World of Warcraft, for example, requires several gigs of space, just by itself.

And 40 applications? If you add up productivity tools, browsers, photo editors, the printer applications that companies like HP want you to install when you buy their equipment, video editors, desktop security applications, archiving tools, and other utilities, you're going to approach the super-user threshold for installed applications, too.

In the drive to get Windows 7 out, and restore its reputation in the process, Microsoft may not have prioritized installation and upgrade. Unfortunately, as user experience pros will tell you, those two processes define the first experience users have with a new operating system.

Last week, in the podcast, I gave a favorable review to Luke Hohmann's book, Beyond Software Architecture, since it explained why considerations like installation and upgrade are not simply the unfortunate necessity you fit into the end of the development cycle. I wonder if Luke might add the Windows 7 upgrade to a revised edition of the book.

Slides are not self-explanatory

I'm not a fan of the recent fad for PowerPoint decks full of pictures, and few or no words. While the text-packed presentations were guilty of a different sin, the answer is not to remove so many words that someone looking at the slides post oratio can't figure out what the presenter was saying.

In other words, your slides are a UI. The user experience must fit the use cases, including people reading your slides, instead of listening to you talk. You could jettison that use case altogether--if someone didn't hear the presentation, too bad for them. However, I'd bet that most speakers don't want that much economy of communication.

If you're teaching a creative writing class, and you're coaching a chronic blabbermouth, you might want that person to learn how to write haikus. That approach might help them write shorter novels, but they're still going to write novels, not haikus.

Playtesting is more than usability testing

In an earlier post, I noted how some best practices for designing boardgames apply to the technology design process. Today, I'll throw in another analogy that might explain why the user experience for high-tech products often falls short of the mark.

Games people play
People who think up boardgames face difficult design challenges. Not surprisingly, they don't get it right on the first try, or even on the fifteenth. Since the criteria for making a top-notch boardgame are difficult to meet, designers "playtest" their games repeatedly. Objective opinions are important, so this QA process involves as many "playtesters" as possible.

Each criterion is important. Obviously, games need to be fun, not just tedious die-rolling or card-playing exercises. What makes a game subjectively fun, however, depends on several elements. To delight your audience, the game has to pose an interesting challenge. People often deride Monopoly, for example, because it doesn't pose any interesting choices. Roll, move, buy property or pay money, repeat for several hours.

The game also needs to be balanced, obviously, so that one player doesn't easily crush the others. Balance has subtle aspects, such as keeping the game competitive even when someone pulls out ahead (the "runaway winner" problem, or its flip side, the "beat on the leader" problem).

Another challenge is replayability. Some games don't provide different possible outcomes, or give room to explore different strategies, to merit more than one play. While chess may not appeal to everyone, it's undoubtedly one of the most replayable games, allowing for different strategies in the opening moves, middle game, and endgame.

Are you experienced?
As important as these individual criteria are (including a few more I haven't mentioned), game design isn't simply addressing each of these items in a checklist. Many bland games meet all of these criteria, but just don't provide that great of an overall experience. (The old warhorse Settlers of Catan is my favorite example.) Other games might not be perfect on every score, but they're still a hoot to play. (On my top 10 list, Cosmic Encounter fits that description.)

The best games provide an experience. For example, party games like Pictionary or Apples To Apples are all about creating an experience (often unintentionally hilarious) that's more important than who wins. If Pictionary didn't give every team a chance to win, however, the experience would suffer.

And in many cases, the experience is actually a byproduct of the rules. Boggle, for example, makes a task that no one would do on their own, arranging random letters into words, into something that can be pretty fun.

In thinking about the experience you want to create, it's important to keep in mind the person for whom you're creating it. For someone who occasionally watches a WWII documentary on The History Channel, Axis & Allies might create the right kind of experience. For a more die-hard history buff, a more detailed simulation of World War II, such as Europe Engulfed, would be a better choice.

I put a spell on you
By now, you probably see where I'm going with this. Technology products need to meet particular criteria, too. Don't force the user to make too many clicks to perform a simple task. Don't overwhelm the user with too many choices. Make it clear which part of a multi-step task the user has reached, and how to move backwards and forwards in that sequence.

Meeting these criteria might make a product tolerable, but you won't necessarily have created a good experience for the user. You may also have overlooked the sort of experience that fits the user. For example, most workflow tools have highly complex UIs that appeal to a particular kind of technical user who wants a very big palette of states, actions, decisions, and outcomes. Business users, on the other hand, want a more paint-by-numbers approach.

Creating the right experience can pay off, as Apple has shown repeatedly. Just as attractive components helped the "Eurogame" phenomenon, the aesthetic appeal of Apple's hardware generated enthusiasm among people who otherwise saw desktop computers and cell phones as an unfortunate necessity.

But even Apple's experience isn't for everyone. Until the Mac OS included a dual-boot capability, Apple's coolness factor didn't convince software engineers to buy their hardware. Now, Apple has created a particular user experience for that technical audience which shares only a few elements with the consumer market. (And Apple still might have blown this opportunity through bad design decisions.)

Already, we've seen signs that technology products don't get the testing they deserve. In some technology companies, usability testing is, to be frank, is either cursory or non-existent. Even when engineering teams ask users for their feedback on particular UI components, they're probably missing the much bigger picture. If you want to capture and keep users, you're going to need to do more rigorous "playtesting," with the goal of creating an overall experience.