Bruce Tognazzini was Apple's 66th employee, developing the company's first usability guidelines and founding its Human Interface team.
Almost thirty years later, he's a principal at Nielsen Norman Group and still making his feelings known when companies commit design errors.
Here, 'Tog' gives us a variety of thoughts on interface design, freedom, the future of computing, the iPhone's place in world history and why he travels around in a 400 sq ft motorhome while towing a 4x4 and two Segways.
What is the main usability/user experience mistake you see on the web?
The level of open hostility that websites display is breathtaking. For every Bed, Bath & Beyond, with it’s smooth, comfortable user experience, there are a thousand amateurish websites that appear to feel that torturing their customers is a really good idea.
In the main, this has resulted from striving to achieve mediocrity, rather than excellence, but it is as devastating to the user experience as if they had set out to achieve hostility.
The worst single fault is throwing away the user’s work. You see this in travel sites, where the user spends an hour selecting airline tickets for dates five months hence, then tries searching for a hotel for that same period, only to find the site has thrown away the dates and is assuming the user wants a hotel for tomorrow night.
The customer playing “what if” with different airlines and different hotels may have to enter the same group of dates as many as a dozen times during these transactions - often resulting in their making a mistake the last, fateful time, and ending up with worthless airline tickets for the wrong dates.
Then there’s the worst single bit of information that can be discarded: The user’s decision to uncheck the box saying, “Yes! I want you to spam me fourteen times a day for the rest of my life!” that appears embedded in the order page.
Go back to change anything on that page, and they’ll turn the checkbox back on. How do these people imagine customers feel later when the spam they specifically rejected starts rolling in?
You recently blogged about how sites can speed up 'subjective time' for users. Could you expand on that?
- Rid your site of time-dependent media. Specifically, eliminate all Flash and video that is not specifically directed at the product or service being sold or discussed and that is not under the direct and voluntary control of the user.
- Support tabbed browsing.
- Limit the number of pages and interactions necessary for a user to accomplish his or her task.
- Do “boredom testing,” where you observe new and experienced users and see where they fidgit, their mind and eyes wander, or they sit back with arms crossed.
- Work out solutions so that when you must do some work “behind the scenes,” the user is engaged in decision-making and doesn’t miss your presence. Use Firefox’s ability to pre-fetch pages, for example, so when the user is ready to go, you are ready to go, too.
What are your thoughts on the user experience offered by different video ad formats, including pre-rolls and Google/YouTube's animated overlays ?
Users hate them, of course, but that’s not the point, the bottom line is. As long as users will put up with it, it will continue unabated.
“Users” have been putting up with so-called “free” TV, with its barrage of ads, for sixty years, even though they are paying for the ads through increased product costs, just as they are paying for YouTube through increased product costs.
Do any of your early guidelines on human interface design still hold true today?
The following are guidelines, verbatim, from my 1980 edition of the Apple II Guidelines:
- Everything the program expects the user to do should feel intuitively right. The user should feel comfortable within the program, and the program should respond to the user's best guess of the right thing to do at any given moment.
- The program should anticipate as much as possible the needs and questions of the user and be prepared to handle them as they arise.
- User interfaces should always be designed with the end user in mind: The ideal interface for a preschooler is radically different from that for a corporate executive.
- There is a trade-off in program design between making a program easy to learn and making it quick and easy to use later on. A program used once should have considerable design effort spent in making it tutorial in nature; ...a program used every-day-all-day should be terse, allowing the experienced user speed and flexibility.
- Screens should be kept as simple as possible. Unnecessary verbiage and superfluous graphics... should be avoided.
- Inputs should be tolerant and forgiving. [They] should recognise and parse the most natural and widest range of possible responses.
If you were to do another Starfire movie (a film made by a group of Sun engineers led by Bruce about the future of computing) what would it show things to be like in 12 years’ time?
Successful technology-predicting is based on detecting discontinuities and predicting the trends that will flow from them.
In the case of Starfire, we had had two recent major discontinuities — the collapse of the Soviet Union and the emergence of the commercial Internet.
That, along with trends already established, allowed us to predict the invention and successful adoption of the World Wide Web with rather chilling accuracy.
Since then, we’ve had only a single political discontinuity, the rise of world-wide terrorism, and only a single technological discontinuity, the iPhone. (Ironically, the most singular feature of the iPhone, the 'pinch' interface, I invented for the Starfire film 15 years ago.)
World-wide terrorism has given all governments around the world the chance to do something we predicted back in 1994 they’d try to do—'fence in' all the freedom that the web was unleashing.
It didn’t happen from the very start because of the heady intoxication of the fall of the Soviet Union and its fleeting unleashing of peace and freedom.
At the first sign of trouble, however, governments moved to limit their citizens’ freedoms and to extend their own power. Human nature.
This trend will continue, with the counterforce being the naturally open architecture of the web, brought about, ironically, by the United States Department of Defense when architecting the internet’s predecessors: These networks were designed to never go down no matter how hard the enemy tried (even when “the enemy” is one’s own government—something they failed to consider).
The other discontinuity, the iPhone, also launched a new trend.
At first, we’ll see a lot of look-alikes, where the unwary consumer will believe they’re getting a phone with the iPhone’s interface, but are actually getting a phone with an MS-Dos-style interface coated with a thin candy shell of cute icons. These phones are already being announced.
With luck, however, the gestural interface will at last take hold, ushering in a new era in power and simplicity across not only phones, but a wide variety of devices.
What lessons do you think other phone manufacturers will draw from the design of the iPhone?
They are three-fold:
- Make it pretty.
- Increase the functionality, with things like voicemails appearing on a menu, where you can pick and choose without having to listen to them all in sequence.
- Copy the touch interface without getting sued.
This last should be easier than Apple wants people to know. These interfaces have been around for 20 years.
At the same time, not everyone should be building an iPhone clone. The Blackberry community would be well-served to keep making ugly phones with real keyboards.
That’s what professional users will continue to want indefinitely, or at least until haptic (physical) key feedback is added to the iPhone touch screen. Technology for this does exist.
Will all companies have grasped the importance of user experience in another 12 years?
If you look back at my 1980 guidelines, above, and compare them with what is out there in 2007, you will see that the vast majority of companies don’t yet grasp even the rudiments of human-computer interaction (HCI).
We still see, for example, most websites demanding that users enter difficult-to-check data, such as phone and credit card numbers, without spaces, all to save the programmer five minutes and a single line of code.
Such ignorance and laziness ensures full employment for HCI designers for the foreseeable future, and also ensures that the original promise of the web, with its sweeping aside of “bricks and mortar stores”, will continue unfulfilled.
Your motorhome looks rather impressive. What’s your favourite piece of kit?
It has been split between the four-foot dish that unfurls and locks onto the Internet when we park, offering us WiFi in and about the motorhome, and the built-in dishwasher that saves us not only time, but argument.
Lately, however, my wife and I have picked up a pair of Segways. We now travel the highways with the motorhome towing a car towing a pair of Segways on a lift behind.
This has increased our gas mileage to an amazing 34 miles per gallon. (8.5 miles to the gallon for four vehicles traveling together means an average 34 miles per gallon per each.)
Which everyday objects/pieces of technology do you find most annoying to use?
Computers, phones, and faxes. Powerful applications have outstripped the capabilities of all three. Computers are sluggish at best, even on so-called high-speed connections. Horrible interfaces just make them that much worse.
Phones are time-dependent media, with users spending inordinate amounts of time on toll-free calls lost in voice-menu hell. A smashing application for the iPhone would be a touch menu that would replace such voice menus. Obviously, an industry standard would have to be developed.
Fax was a fantastic technology in 1870. Now, it’s just kind of sad. Fax is, however, an example of a principle that futurists must always take into account: old technologies tend not to go away.
They find niches and hang on forever, just like radio, whose death knell was sounded in the 1940s, but which continues to do well today, particularly in the drive-time niches of morning and afternoon.
As another example, when I did Starfire in the early 1990s, I was lambasted for showing a large, full bookcase in the heroine’s office. Books were scheduled to be obsolete by far-off 2004. Actually, they seem to be doing rather well.
Finally, do you have any advice for someone looking to persuade senior management to buy in to usability/user experience?
Pour over your log files and be prepared to point out the places where users are “bailing out,” along with cogent arguments as to why, mentioning things like your lovely, design-award-winning Flash animated-splash screen that takes a minute to load and does nothing toward selling the product.
Then, convert those bail-outs into dollars:
“We are losing 20% of our customers before they ever even enter the site because of our splash screen. Last year, our sales were $140,000,000. If we hadn’t lost all these people, we could have realised an additional $11,420,000. Total lost profit: Around $750,000. A single HCI designer could have prevented that, at a savings of around $700,000 dollars. We also will no longer have to put out release x.01, x.1, and x.2 every time we come out with a new design, because the design will be right. That would save us millions more in engineering resources.”
Companies not only save millions by having HCI talent available to them, they often move back from the brink of extinction.
HCI can be a no-brainer to senior management if the case is made clearly and expressed in terms they understand — money.
Failing buy-in, do it anyway. All that’s needed is a broom closet and a couple of tables.
Forget about video. Just see if people can use your site. A single test with a single user in a broom closet can be such an eye-opener, it can change the course of a whole project. (Of course, even with qualitative testing, 20 or 30 users, over time, are better.)
Even a really bad designer, with sufficient user-testing, will eventually be able to cobble together a decent design — the infinite number of monkeys theory — the worst crime is to not test at all.
Also consider becoming an HCI designer yourself. If you’re concerned enough to petition management, you have the most important prerequisite — you care.
Consider taking a course like the ones I offer, getting people up to a good level of competency in a few intense days.
Usability and User Experience Report 2007