Please note that some of these links may have suffered linkrot. They did work as of the date when they were recommended, but the Web is a highly fluid medium.
December 31, 1998
Disney's head
visionary, Bran Ferren talks about the problems in corporate IS
departments: they are the most despised part of the organization because
they lack vision (focusing on fire-fighting instead) and neglect projects
that could improve the company's over-all productivity. Much of what he
says rings true: for example, I am amazed at the lack of resources invested
in the usability of intranets. It's as if many IS people think that it's
enough to make bit-transport available and then people will figure out how
to use the corporate information infrastructure.
December 30, 1998
Jon Katz analyses the difference between the postings in public discussion
groups and the email received from
"lurkers" who don't want to argue (and flame) in a public forum. He
cites statistics showing that 98% of site visitors don't post in
discussion groups and points out the danger in assuming that the 2%
who do post are representative of the rest. A classic example of the
participation
inequalities on the Internet and the problems inherent in many
current so-called "community" interfaces (that are not communities at all).
December 29, 1998
The New York Times has a great weekly column on e-commerce.
This week, the topic is
trends
for e-commerce in 1999: designing for low bandwidth, integrated
advertising, some fraud, $18 billion sales, and better back-end systems.
All good points. In particular, all sites should prepare contingency plans for the
security disaster that will hit the Web soon. "Which disaster?" you
may ask - I don't know, but Internet security is weak enough that something bad
will happen. Be prepared to explain to your users why your system is safe
when the lead story on the evening news is how some other site lost a million credit
card numbers or posted private tax returns on the public Internet (or some
other disaster).
December 24, 1998
The Apple store is "closed for the
holidays" from Dec. 23 to Jan. 5. Amazingly clueless: the Internet is international and not
everybody celebrates Christmas. Websites need to run 24 hours per day, 365
days per year. You can't simply take an ecommerce site down and expect
customers to
come back later. Only redeeming point: a link to a list of local dealers -
but the list is U.S.-only. European and Asian customers might well
take this as a strong indication of Apple's priorities and intentions to offer good
support in their regions in the future. A classic example of how a bad
website can undermine a company and do active damage to its
brand.
December 22, 1998
Jesse Berst hardly needs links to boost his traffic: he
is the most widely read writer on the Web today. His secret? He's a good
writer to begin with and then he uses the recommended style for writing
for the Web. Anyway, his current column is particularly good:
a list of new ways of improving Web search.
December 21, 1998
Chris Crawford's classic book The Art of Computer Game
Design is available in an online version on the Web. The
book is from 1982 so all the examples relate to long-gone games (though
many of the early games were the most fun ones: it's hard to beat
PacMan or Marble Madness). But the book is
really about the principles of designing engaging user
interfaces and not about individual titles. A major point in the
book is that enjoyable and engaging games require a tight interaction
between the player and the game: player participation makes for
great gameplay to a much larger extent than beautiful graphics (a lesson
that is lost on some current game designers and many Web designers).
Avoids the up-front burden of having to program the VCR
before a show is aired. With TiVo, you know that you
always have the latest Star Treks saved (if that's your favorite
show) even on days when you unexpectedly come home from work late.
Easier to use because of the non-command interface; the TiVo UI is a even more of a win because
VCR vendors have refused to employ sufficient usability engineering
resources to make it easy to program their platform (of course, they
probably never thought of the VCR as a "platform" that could get
competition).
A TiVo VP is quoted as saying that "this has the potential to be bigger than
the Internet" - obviously wrong since their system is not
interactive: it's a convenience rather than a revolutionary power shift.
It is not going to change society to find an easier way to watch the same TV
shows, but it's surely worth buying a gadget that allows this.
December 17, 1998
Andrea Williams from the investment banking firm Volpe Brown Whelan
presents the argument for a
customer-dominated Web strategy and emphasizes the need for
loyal visitors instead of eyeball-collection (as I said myself in the
Alertbox
for August 1997). Finally an analyst who understand the reversed power
relationship on the Web instead of seeing it as an extension of the existing
media landscape. Follow her advice and profit.
December 15, 1998
The New York Times cites makers of several high-end products
for not wanting to sell on the Web for fear
of diluting their brand and endangering their channel. How clueless;
how arrogant. The Web is a roaring beast that will devour brands like
Pioneer and Maytag for breakfast. You can't "protect the channel" in the
new customer-dominated future: the only way to business survival is to
blow up the old channel structures. If these companies don't do it
themselves, then new Internet-focused companies will do it for them and
establish new high-end brands that derive their luster from the premium
service they give on their sites. I bet that there is at least one Pioneer
VP who is writing a business plan this very minute for breaking out and
creating new high-end brand to be sold over the Web (that's where most of
us rich gadget freaks hang out, you know).
December 15, 1998
Also in the Times: Disney's new Go
site is based on heavy usability data from Infoseek.
Since Infoseek is one of the best search engines and has a professional
usability group, this is promising news, though there is always the risk
that an old-media company like Disney can mess up any Web strategy.
December 11, 1998
It is now thirty years since Doug Engelbart invented the mouse: an
anniversary
that was much
celebrated this week. We should remember that most of Engelbart's
other inventions in hypertext and collaboration are still missing on the
Web. In fact, his ideal of augmenting the intellect is in grave contrast
to the prevailing design attitudes
these days when many proponents of eyeball-focused Web design attempt to
suppress user initiative.
December 9, 1998
I have long said that frames
are bad for usability, but it has now been discovered that
frames are a security risk:
basically, a third-party site can cause the browser to display a frameset that looks like it is coming from your site, with your URL in the
browser's location box and with your navigation bars and other user interface elements.
The only difference between a legitimate page and an evil one is that the spoofed frameset
includes one frame from a different website. If this frame asks for
the user's credit card number or password then the average user will enter the requested information,
believing that the data goes back to the "good" site, when in fact it goes to the
frame-spoofer's site. One more reason to minimize use of frames.
December 3, 1998
The new Palm Pilot VII comes with built-in wireless
connectivity and is the first step toward realizing my vision of
standard mobile Internet access: has to be standard rather than
an add-on to allow software developers and Web service developers to count
on users' ability to access the Internet. Unfortunately, this first
product is very weak: requires use of a separate email
address (get it: unified messaging is in), only allows access to a
restricted subset of the Web (get it: openness is the
foundation of the Web),
and has a miserable service plan that charges 30 cents per kilobyte
(since a 28.8 modem transmits about 3 kilobytes per second, this is
equivalent of a per-minute charge of $60 for airtime - get it: this is
not a paging service; we want wireless IP and the cost savings associated
with digital). I predict that the Palm VII will still be a success because
of the huge pent-up demand for untethered connectivity, but it will be easy
to knock it out with a true mobile Internet solution.
December 3, 1998
George Gilder is usually too disconnected from reality with
his insistence on abundant bandwidth Real Soon Now. Interesting thought
pieces, but not very realistic. But an essay he wrote earlier this year is
right on in setting directions for the customer-dominated reality of the Web:
"Life Span vs. Life Spam" says that companies used to waste
customers' time in order to conserve their own resources (one of the worst
examples is the telephone message your call is very important to us,
so please continue to hold). In the future, customers won't stand for
this since their time is getting ever more valuable (the ultimate limited
resource), so companies that conserve customers time will prosper. My
added comment: The Web is truly driving this shift - if you have to wait
just a few seconds too much for a page to download, then you are out of that
site and give your e-commerce dollars to a faster site.
December 1, 1998
www.microsoft.com gets a redesign and the General Manager
writes that the design goals were to
"reduce clutter, condense the editorial,
upgrade the navigation and speed up page loading". How can I do
anything but agree since I have been talking about speed, speed, shorter
content, and clear navigation structures for years. It is particularly
interesting to notice that almost every single press release announcing the
redesign of a major website states that faster download was one of the
main design goals. People would save a lot of money if they listened to
the usability
findings before blowing their budget on a "cool" and good-looking
design that nobody can use. Regarding the Microsoft redesign: I still find
it quite hard to navigate the site (try finding the product page for
Barney) and there are several minor usability problems in the use of DHTML
(for example, on the Internet Explorer
download page, the little triangles
sure do look clickable).
November 27, 1998
David Thiel of Microsoft Research explains why it is difficult to design audio
enhancements of user interfaces. His main point: "if you don't keep a
high information to disruption ratio then annoyance will
result" - clearly a lesson not taken to heart by most sound effects heard
on the Web today. Another great insight: "for audio time is like screen real
estate", so keep most sounds short. This short essay should be required
reading for anybody who wants to design interactions that move beyond the
current GUI stagnation.
November 24, 1998
America Online is buying Netscape.
Hard to know what to make of this, but it will probably
mainly be good since AOL has a strong commitment to ease-of-use. Thus,
having them take over Netscape may lead to the first substantial improvements in
browser design since Mosaic. On the other hand, AOL is too strongly
associated with closed proprietary services to be likely to embrace
the innovations that are necessary to advance the Internet to the next level.
In particular, they are preoccupied with selling links to the highest
bidder instead of helping their customers get the best services, content,
and ecommerce deals. As a result, AOL has declined in quality every year for
the last three years, with more and more annoying pop-ups and less and
less useful guidance and good content. The most provocative analysis
of the deal so far is the Online Journalism Review saying who
cares about Microsoft? All they do is make money; they don't try to tell
people what to think. AOL is a much greater danger to a free society.
Another good analysis comes from Doc Searls (maybe
the most insightful marketing expert writing on the Web these days).
November 23, 1998
The Economist surveys the interplay between technology
and entertainment. (To navigate this article, click the section headings
in the left stripe, even though it looks looks more like global
navigation than local navigation.) The article concludes that the "Big Seven" of
Content will strengthen their hold because of new technology. I tend to
think the opposite: that the Internet will lead to more diversity in
content, but it remains to be seen whether this will happen. I have been
getting less optimistic over the last year as the vendors have
refused to improve the navigation abilities of browsers and search
engines. On one point I do agree with the article: broadcast networks will
be losers in the new world.
November 16, 1998
Dan Shafer discusses the benefits of
making page
designs work across a range of monitor resolutions and window sizes,
using the new design of the C|Net services as an example. I am very happy
with the new C|net design: not only is it much more flexible (and should
work better on WebTV and other new devices), it also gets rid of the
"yellow fever" stripe that wasted too many pixels.
November 14, 1998
According to the Sydney Morning Herald,
David
Siegel denounced "cutting-edge Web design" during lectures in
Australia: he has now
decided to design useful sites instead of killers.
Bravo! There is nothing as good as a reformed sinner, so I will definitely
have a lamb roast in Siegel's honor. He is even quoted as preferring a
"conventional and less exciting" design for one of his clients because
it scored better in usability testing. Way to go, Dave!
November 11, 1998
John Rhodes' essay on informality,
jargon, and tone of writing advises use of a vocabulary that matches the
users' own online writing as seen in their email, discussion postings, and related
websites. My take is that Web writing is more informal and relaxed than most
other media, but should not take on too much of an "attitude".
October 29, 1998
David Weinberger: The Web causes the Death of
Documents as people build intranet sites instead of long reports that
nobody reads (unfortunately this
essay is part of a long scrolling page, so the preceding link jumps you to
the middle of the page - not a recommended way of doing things: it is better
to modularize information with each linkable unit as its own page).
October 14, 1998
Case study of an intranet redesign: a "boring" design
focusing on
employee needs greatly
improved the intranet over an earlier design that was driven by the
company's internal structure.
October 1, 1998
Bruce "Tog" Tognazzini writes about the sorry state of Web design, emphasizing
the lack of navigation support and structure visualization in current
browsers. Tog is one of the world's leading designers of Web-based
applications (and for sure the world's best writer on the topic),
so it is gratifying to see that his conclusion is much the same as my
analysis from a hypertext perspective (last year I said that v.4 browsers
are as bad as Mosaic with respect to navigation).
September 27, 1998
BabyWow
is a new software product that helps infants and toddlers 9mon-3yrs
increase their vocabulary. The $40 CD-ROM ships with 8 languages: evidence
that international use was a fundamental part of the design and not an
afterthought. I would expect nothing less from the head of this project:
Tony Fernandes, author of the book
Global Interface Design.
Until Netscape laid off the vast majority of its user interface
folks,
Fernandes was manager of Netscape's User Experience Group: the Web's
loss turned out to be the kids' gain.
September 21, 1998
Robert Seidman's Online Insider
newsletter has a retrospective on his four years of publishing (one more
than me). Interesting analysis of the early years and good predictions
(including: most people won't get fast bandwidth for five years - a close
match with my own analysis derived from a very different method).
September 16, 1998
Trellix has posted an improved hypertext design for the
Starr report.
This version is indeed more readable than the original at the
Library of Congress, but to make a true hypertext, you would have to link
to the original testimony (and other source documents) instead of simply
improving the footnotes. Of course, they don't have access to the 36 boxes
of evidence (and this mass of info would take forever to type in and
convert to hypertext). An even better hypertext would have cross-linked to Clinton's rebuttal of
each point and some independent commentators, though the exact way of doing so becomes an editorial job and
not an objective matter.
The Congress website scores fairness points for linking to Clinton's
rebuttal (even though this is done as a single link to the White House and
not as a step-by-step link from details in Starr's report to the
corresponding rebuttal arguments). The White House gets a minus for Web
usability for using "INITIAL RESPONSE TO REFERRAL OF OFFICE OF INDEPENDENT
COUNSEL" as the link from their home page to the rebuttal. All caps, no
less, and certainly not in accordance with my usability guidelines for
writing Web
microcontent.
September 12, 1998
Article on trends for intranet search claims that only 32% of
intranets are searchable across the company (as opposed to searching a
single server that only holds part of the company's information). I usually
estimate that large companies lose about $10-50 million per year in lost
employee productivity due to poor intranet navigation design, and lack of proper intranet
search probably adds $50 million to that cost.
September 3, 1998
A Carnegie Mellon University study found a positive correlation between Internet use and depression
(warning: rather academic paper). Salon magazine has a pretty
good critique of the
study. I tend to believe the facts of the study as reported in the
paper since the first author is Bob Kraut who is the world's leading
researcher in the field of social use of computers (and my colleague at
Bell Communications Research for four years). My main critique of the paper
is that they have measured "Internet use" as a single variable, without
reporting separately on the use of specific services. I can easily believe
that excessive use of chat rooms does correlate with depression
(I never liked chat much). I have a
harder time believing that it causes depression to read, say, The New
York Times (which broke
the story) on the Web and not on paper.
September 2, 1998
The average salary of user interface professionals was $78,000 in 1997 (though 4% made more than $200,000). Average annual bonus was $9,000.
August 26, 1998
How to prepare your site to handle a disaster or other major event causing a surge in user interest and traffic (page views jumped by 1,100% on one newspaper site during a flood).
August 8, 1998
New York Times article saying that Web users in the U.S. waste 2.4 billion hours per year waiting for Web pages to download (access to the Times requires registration).
August 4, 1998
I have been reading the new book AOL.com about the history of America Online. I cannot recommend the book as a way of getting insight into consumer online services. This is a completely personality-driven book that spends more paragraphs on the restaurants people dined at while negotiating a deal than on analyzing the implications of that deal for how people use networked computers. As an example, the book has multiple portrait photos of Steve Case and other AOL executives but no screen shots of the changing design of the service over the years (or comparative screen shots showing how CompuServe and Prodigy looked at the time - even though such screens would have been compelling evidence for why AOL won). The book is well-written and has plenty of high-tech gossip, so it is perfect for being stuck in O'Hare when United Airlines cancels your flight.
The results of the 9th GVU Web user survey are now out. Currently, only the raw data is available (analyses and other writeups are still in progress). Two highlights: 58% of users have small monitors (800x600 or smaller; percentage calculated only for users who knew - 23% or respondents answered "don't know"). The percentage of users accessing the Web with analog modems was 77% - exactly the same as found in the 8th survey in 1997. In other words, the Web continues to be dominated by low-end users.
July 15, 1998
The New York Times has eliminated the special subscription fee for overseas users. Bravo. In the two and a half year since the Times opened its website, they had gathered 4 million registered users in the U.S. (where it has always been free to register) and only 10,000 paying subscribers from the rest of the world. Pretty clear that the subscriptions were not working. (Also, by the way, they have launched a much-improved homepage redesign. The previous one turned news into olds with its slow download.)
July 3, 1998
Steven Johnson explains why Alexa may defeat the misguided "portal" strategy that is currently all the rage. Johnson doesn't state the critical need for reputation management if the Web is to stay an interactive medium, but this is the first article I have seen (besides my own) to recognize Alexa as the beginning of the next stage of the Web and the benefits from harvesting the collective intelligence of the (soon to be) billions of Web users.
June 28, 1998
Clickviews:
readings on Website publicity, edited by Eric Ward.
June 19, 1998
Bruce "Tog" Tognazzini has launched his AskTog column with advice on Web design.
Tog is now the lead designer for an Internet startup, but for many years he was a Sun Microsystems Distinguished Engineer (that is, among the company's smartest 0.3% technical talent). He also founded Apple's human interface efforts in the early days of personal computing. Tog is the world's most creative and expressive author on user interface topics so I am sure that his column will be worth your bookmark.
Infoseek is sponsoring a debate on measurement methods for Internet readership ratings. My analysis is that these ratings are useless for anything except bragging rights. Ratings should be irrelevant for an Internet advertiser since the only two things that count are (a) page views and (b) click-throughs. Why would anyone care how many other users see other parts of the site?
Esther Dyson's essay on privacy on the Internet: too long to read on the Web (115 KB), so print it out. We need to think more about these issues and commit to protecting users' privacy. Most users are very reluctant to disclose personal info on registration forms because they don't know how it will be used. Tell people, and then keep your promise.
May 3, 1998
Mark Bernstein wrote a surprisingly positive review of the HotWired style guide to Web design. Until now, I had refused to look at this book under the assumption that nothing could be learned from HotWired which used to be famous for an atrocious design with the primary goal of repelling any users over the age of 17. According to Bernstein's review, the book is not actually a guide to user-hostile style but a case study of findings from HotWired's many design iterations. Presumably, the book title was chosen by a marketing department that knows that the Web is desperate for design rules: it is good to discover that the main focus of the book is much more useful for readers than an attempt to emulate earlier designs that didn't work. After reading Bernstein's book review, I revisited HotWired and discovered that it is now possible to read the text on their homepage and to use the page as a navigation aid. I stand corrected: HotWired may have sucked in years past, but this is no reason to reject the insights in the book.
April 24, 1998
The Industry Standard is a new magazine that is being billed as "the newsmagazine of the Internet economy" and as Wired meets The Economist. The first issue has not been published yet (you can sign up to get it for free at their site) so I can't say whether it will be any good, but it has sure generated a substantial buzz this last week. Also, they research their articles well.
April 14, 1998
Bruce "Tog" Tognazzini (author of Tog on Software Design) has recently worked on design issues for Web applications.
He shows seven design iterations of the first page in a healthcare application. Two lessons:
putting something on the Web does not automatically make it easy: most employees would never have been able to sign up for their health benefits without Tog's redesigns
often it takes several iterations to get something right: simply finding a usability problem and "fixing" it in a redesign does not guarantee that the new design will work
April 7, 1998
Internet Computing Magazine names the twenty best Internet consulting firms, partly based on the usability of the Web design they do for their clients.
April 4, 1998
More evidence that advertising and a hard sell are less important on the Web
than simply making it easy for users to
buy: several sites report huge profits from facilitating comparison shopping
and even guiding users to competing sites. (The link points to The New York
Times which unfortunately requires users to register before they can read the
article.)
Susan Feldman and a group of professional searchers (typically librarians) have compared the use of traditional retrieval services like DIALOG with searching the Web.
The traditional services win handsomely: twice as fast (10 hours compared with 21 hours when searching the Web to answer the set of test questions) and 22% higher relevance of the retrieved documents.
Two problems with the study:
The test users were skilled in the use of traditional database search services, so they performed much better with these systems than average users would be capable of.
The traditional services have much higher-quality documents than the Web since they support micropayments for retrieved content (actually, their charges are often much higher than a typical micropayment, but the point is that they compensate content providers, so they get better content). Thus, the more relevant results may be due to the quality of the content as much as it is a result of the differences in search systems. Once the Web gets micropayments, it will probably perform better.
Nevertheless: an interesting study. Also includes the professional searchers' tips for how to get the best results from current Web search engines.
February 26, 1998
The March 1998 issue of BYTE has a good overview of XML.
Recommended Websites