useit.com -> Recommended Websites -> Spotlights -> 1999 Q1 Spotlights

Spotlighted Links From January-March 1999

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.

March 31, 1999
Clickthrough rates for Web advertising may be dropping even faster than expected. A new study from Nielsen Media Research found average clickthrough of 0.15% in February. I had predicted that clickthrough rates would be cut in half every year as users turn away from misleading and distracting Web ads, so my expectation was that we would reach 0.25% later this year. Getting down to 0.15% as early as February is almost too good to believe in terms of vindicating my statements since 1997 that advertising doesn't work on the Web. But despite the name, there is no relation between Nielsen Media Research and myself, so they don't have any particular reason to vindicate me. And they are a more respected research firm than most. I do take exception with the comment from one of the researchers that clickthrough is "like measuring success at a trade show by how many T-shirts you give away" - a better analogy is how many of the people who visited your booth send in a response card requesting more product information. Still not as good as measuring actual sales, but a pretty good measure of promising leads.
March 31, 1999
Jeff Bezos recently said that "the No. 1 thing is the customer experience" but Amazon.com's new page design contradicts that goal. Instead of being an optimal book-buying user interface, Amazon now forces users to scroll through more than an inch of irrelevant blurbs for their online auctions that have no relation to the book the user is looking for. The page for a software engineering book featured an auction of a "Miss Spider Plush Puppet" smack in the middle of the book description: polluting the user experience.
March 30, 1999
McAfee's download site seems to be almost totally down - overwhelmed by demand for anti-virus software after Melissa. Yesterday, it was also close to impossible to download virus fixes from Computer Associates, but they seem to have recovered today. The general lesson is that the Web is susceptible to flash crowds and to have contingency plans to handle huge bursts in volume. In the case of McAfee, they could at least pull the heavy graphics from their download page and use a strategy similar to that of MSNBC, CNN, and other news sites when big stories happen: go to a lightweight version of the site to conserve bandwidth. (Historical note: I introduced the term "flash crowds" in the Web sense when writing about the mass of users descending on the site for the 1994 Winter Olympics in Lillehammer, but I didn't coin the term, having taken it from a 1973 science fiction story by Larry Niven.)
March 29, 1999
New York Times story on mainstream computer games like Deer Hunter quotes a gaming executive as saying that games for the larger market "is for people who want to do much the same things on the computer -- hunt, fish, play cards, whatever -- that they do in real life." I disagree. Plenty of games have successfully attracted mainstream users without being about everyday life. Think Pac-Man or Tetris. To me, the key to a mainstream game is that it can be played without an overly steep learning curve (admittedly, a reality-based metaphor allows users to transfer their knowledge about the reference domain to the game domain). "Get a life" implies not spending a week learning a game before it gets really fun. Also, as people get more and more activities (both at work and at home - another part of "getting a life"), they are reluctant to commit 100 hours to a game. Games for late adopter users must be low commitment and provide satisfaction from the beginning. No adventure games where you have to search half a castle for a stupid key to progress to the next room.
March 27, 1999
The ClueTrain Manifesto lists 95 reasons Internet business is different from traditional business. Much overlap with my own comments on Web design over the last five years. Despite the fact that this manifesto and my own writings are online for everybody to read, I predict that most big companies will still not get it because their internal management structures are too well built and capable of resisting customer-centricity until it's too late. 80% of the Fortune-500 companies will be gone in ten years (unfortunately I don't know which will be the 100 companies to change in time).
March 24, 1999
Prix Ars Electronica'99 prize competition for interactive art. This one is too "cool" for my taste, but if we take the middle road between the interactive arts people and the interactive fiction people (see March 22), then we will hopefully get some innovations to break Web content out of its current funk.
March 23, 1999
Comparing Media Metrix's Web traffic estimates for 1996 and 1999 shows that only 3 of the sites on the top-15 list in 1996 are still on the list three years later. In other words, 80% of former top sites have dropped out of the list in this short period. These traffic estimates are good supporting evidence for my argument that Internet stock is over-valued because future users are likely to prefer other sites than the currently popular ones.
March 22, 1999
April 7-9, 1999: conference on the "Next Century For Interactive Fiction". Too "English-lit" for my taste, but still a welcome initiative: we need to find new forms of expression for interactive text.
March 18, 1999
A recent survey of British corporate websites found that 92% of managers expect to increase spending on Web strategy in 1999, with 34% expecting to double Internet spending this year. Of course, this increased spending may well be wasted since most companies don't know how to run a user-focused site. The study also claimed to find that most sites have improved since last year: for example, 35% refresh content on a daily basis, compared to 22% last year. Unfortunately people often follow such superficially correct policies without making the content useful to users or writing it according to the online writing style guide. So the simple statistics presented in these popular reports are not necessarily indicative of any improvements in the Web user experience. On the contrary, when I review actual empirical data, I am depressed that the outcome is the same year after year with very little improvement in Web usability.
March 16, 1999
Dan Shafer discusses the issues in chunking long content into multiple Web pages. Chunking is important since long scrolling pages don't work on the Web (and Shafer notes that most users on his site don't read to the end of long articles). But chunks should be designed to help users, not just to expose more banners. And a ToC is essential for any piece with more than 3 chunks.
March 15, 1999
Bill Gates has written a new book. We shall see if it's any good, but the special book website is not promising: a splash screen (the sure sign of a user-hostile site) that's pretty heavy in graphics followed by a very heavy home page where all the words are graphics. It almost goes without saying for such a low-usability site that the images do not have ALT texts. I wonder if Bill Gates knows that he is being represented in a way that brutally discriminates against people with disabilities (and is unpleasant to use for the rest of us).
March 15, 1999
Most schools don't teach the children how to avoid injuries when using computers; on the contrary, the kids usually don't have ergonomic furniture, keyboards, and other equipment that's appropriate for their size, so they may be even more at risk for RSI than adult users. One of the most important safe computing practices is to take frequent breaks and stand up but children are often too captivated by the computer to take even this simple precaution to prevent damage to their hands and backs. The New York Times visited schools in New York that "remain focused on teaching both staff members and students how to use the programs on their new computers," with desktop publishing being a major curriculum. (Come on: desktop publishing is an obsolete computer skill, why not teach the kids Web design so that they would learn something they could use after they graduate?) Nobody teaches students how to preserve their health. And then we wonder why health care costs keep going up...
March 9, 1999
I am very critical of advertising on websites, but it's fine for sites to advertise outside the Web. I am pleased with OnSale's idea for TV ads that are partly driven by a live feed from the website: the commercial does contain some fluff (necessary to catch viewers' attention), but then it also shows real data taken from the site a few minutes before the broadcast. Good way of making TV viewers aware of the dynamic nature of the Web and attracting them to visit to get something specific that is currently featured on the site. Much better than generic boasting.
March 5, 1999
More than a year ago the Wall St. Journal decided to withdraw from the Web and put their content behind a subscription wall. In the long term (5-10 years), not being a fully integrated member of the Web will seriously erode their brand as people prefer to link to other online sources and as mainstream search engines (Excite's NewsTracker, etc.) feed readers to everybody else. But the WSJ may finally be getting a clue (and their brand is strong enough to survive a few years in the desert). First sign of re-joining the Web is Walt Mossberg's column which this week looks at Web-based calendars and correctly concludes that they are still too clunky.
March 4, 1999
Jon Udell describes a simple way to rank the Web mindshare of a set of websites. His list of computer magazine websites ranks BYTE as number 12 despite the hiatus of that magazine until recently. Using Udell's metric, Useit would be number 24 on the list (scoring below PC Gamer). CNET and ZDNet top the list (no surprise), and LinuxWorld is number 3 (maybe a surprise for some people).
March 1, 1999
InternetWorld's six design critics analyze their favorite sites, ranging from the famous (yes, Amazon.com is on the list) to the obscure.
March 1, 1999
BYTE Magazine has re-launched as an online-only publication. The quality of the new material remains to be seen though the first few articles are promising. One great feature of the site is a complete archive of all articles since 1994 (including, for example, an interview with me from 1997). BYTE was often first to provide substantial coverage of trends in personal computing, so the archive is a true treasure trove. Let's get the 1980s online too.
February 26, 1999
Yesterday in the Microsoft anti-trust case, the judge very reasonably asked a Microsoft VP to explain the benefit to users from integrating Internet access into the operating system. According to the New York Times, the VP couldn't answer the question beyond unconvincing generic comments. This "trial-to-define-the-future-of-computing" turns on usability arguments and yet no user interface experts have been called as witnesses. Maybe the reason is that both parties are afraid of a thorough usability analysis of the case:
February 25, 1999
Update to the search engine referral stats: Yesterday, AltaVista shot to the top of the chart. Why? It's the only major search engine to have indexed my new essay on iCab (a word many people are searching for right now). The others apparently have given up producing daily updates to their index. Other iCab news: iCab now accounts for 0.4% of the traffic on useit, though that's likely to be disproportionally high because of my iCab article (people like to read about their own browser - I have seen that for WebTV as well).
February 24, 1999
Last week's server stats for useit show inflow of traffic from search and navigation engines in the following order: Yahoo, Go (formerly Infoseek), AltaVista, Google, Excite, Snap, Go2net, Mining Co., LookSmart, Lycos, HotBot, AskJeeves, WebCrawler, GoTo, AOL NetFind, RealNames, EuroSeek. Only the emphasized names matter in terms of volume. Traffic to a site from a search engine depends on three factors: the engine's raw traffic (so almost everybody gets a lot of visitors from Yahoo), the interests of people using a certain kind of engine (so useit will get uncommonly many visitors from AltaVista and Google since they tend to be used by the kind of advanced users who care about my topic), and whether the engine emphasizes quality or quantity (so, since I think useit is a high-quality site, it will get more visitors from Go, Google, LookSmart, and Mining Co. because of their emphasis on guiding users to the best sites).
February 23, 1999
Covering the doll's eyes is the killer app of Digital Barney. Why? Because that's Barney's only non-command interaction. Squeezing Barney's toe to sing a song is just as much a command as grep or dir: something you do for the sake of the computer and not as a natural action in your environment. Digital toys will be big in the future, but they need usability just as much as traditional computers do.
February 22, 1999
Too few sites take the long term view on the Web and think about the changes over a multiple-year period. Mark Humphrys does and predicts that many links will evolve to point to stable sites that remain running, retain their focus, and keep their pages at the same URLs over many years. For example, his site has many links to the Irish Times and very few links to the competing Irish Independent because the Times is more stable. Since links is one of the main ways to attract traffic, we can predict higher long-term use of stable sites. Humphrys also thinks that the long-term perspective favors Yahoo! over smaller sites. I hope not, but I agree that many sites are reckless in causing linkrot for any author that tries to give them business by linking to them.
February 19, 1999
At the TED conference Thursday, Douglas Adams (Hitchhiker's Guide) said that we need to make the best of our limitations while we have them. He was mainly talking about other art forms (like film), but the comment rings particularly true for the Web. In a few years, many of the design limitations will be gone and people will shoot for all-out fancy multimedia productions unless we learn now how to produce interactive content that puts the user in control of an information space. It's a necessity now because user-driven interaction is the only thing the Web can do better than other media, and we could well lose the opportunity go beyond traditional one-way video if we don't do so now.
February 18, 1999
In his talk at the Editor & Publisher conference, Bloomberg said that "serious people will always read newspapers". He is wrong. Unfortunately I am not speaking at the conference this year, but two years ago I showed work we had done at that time on flat panel interfaces to the Web (and flat Web devices are coming to market now). Once high-quality Web access becomes a reality (requiring a combination of portable flat screens, high resolution, anti-aliasing like ClearType, fast bandwidth, a decent browser UI like we hopefully will get with IE 6, and content written for the medium), the high-end demographics will desert print first because they will be the ones who can afford the good screens and connections and who will be best served with targeted value-added online content in the beginning.
February 17, 1999
Randy Cassingham explains how he succeeded in narrowly focused Internet publishing with his This is True service. Two important points:
February 16, 1999
The Industry Standard on Web ads that are little programs instead of annoying banners. I think these new ads are also doomed because of the basic "I'm-in-control" attitude of Web users, but anything should be better than banners.
February 15, 1999
The latest statistics show 43 million hosts on the Internet and 4.3 million websites. Slower annualized growth rates than in the past: 66% for the Internet and 123% for the Web. We are nowhere near saturation, but the hyper-growth phase is over. Send in the late adopters.
February 14, 1999
Here we go again: A new survey shows that only 37% of European CEOs use the Internet at least once per week, even though 82% of the CEOs believe that "electronic business will have a strategic impact upon their firms." The 18% who don't believe in ecommerce may lack vision (or they may run a chain of barber shops), but that's better than the ignorance of the 45% who do understand that the world of business will be changing dramatically but avoid a deeper understanding through personal experience. American CEOs use the Internet 50% more than European CEOs: better, but still not enough to prepare them for the future.
February 13, 1999
I try not to say "I told you so" every time I am vindicated, but this one is too good to let go: Slate Magazine removes their subscription block and establishes free browsing. In June 1997 I said that paid subscriptions don't work on the Web (except for outlier cases like the Wall St. Journal). Then, when Slate was just about to introduce their subscription policy, I allocated half of my January 1998 Alertbox to a more detailed argument under the heading Subscription Fees Fence You In. I really did tell them so; several times. (And while we are on the topic of Slate: couldn't you please kill the page numbers?)
February 13, 1999
Robert X. Cringely correctly debunks the Free-PC scam.
February 12, 1999
Review of digital flat-panel displays shows that pretty good ones are now available for less than $1,000. Unfortunately, these reasonably-priced displays have a puny 1,024x768 resolution. Since screen resolution correlates directly with user productivity, I recommend getting a bigger display. The best monitor right now is probably the Silicon Graphics 1,600x1,024 pixel flat-panel display, which unfortunately costs $2,500. In any case, future monitor purchases should be restricted to flat panels: Microsoft's forthcoming ClearType that provides better anti-aliasing and much enhanced user productivity due to faster reading speed will only work with LCDs (not CRTs). Once a new version of Windows comes out with ClearType (2001?) CRTs should only be used by people whose time is worthless.
February 11, 1999
A reader from Europe writes (slightly changed to preserve anonymity): "A very expensive Web advertisement campaign was put in place for one of our sites. This campaign involved permanent banners on major portals. The increase in traffic was literally imperceivable. However, a short spot on [an important] radio show produced an immediate traffic increase of around 900% within 24 hours. A week later the figures started dropping again but one month later the average page hits remained some 100% higher than the pre-radio show figures. Newspaper exposure and other media exposures produced similar results. Is the conclusion to be drawn that the best way to advertise for the net is off the net?"
Jakob's comment: I have always said that advertising doesn't work on the Web. But we need some way of getting people to visit new sites and it seems that off-line promotions do work as shown by this example as well as experience from American TV commercials.
February 10, 1999
David Weinberger's essay Down with Reality says that information is often more "real" than physical objects (in the sense of having more truth or impact). Interesting point as the Web makes more things virtual. But let's make the Web better than reality and not just a replacement.
February 10, 1999
The Fifth Conference on Human Factors & the Web will be in Maryland on June 3, 1999 (too bad that so little research is being done that a single day is enough!). Mark your calendar: this is usually a very interesting event. The deadline to submit papers or panels is March 8. This year the conference will focus on Web applications, but they usually also accept some non-theme presentations.
February 9, 1999
Estée Lauder sues Excite for making another company's banner ads come up when users search for their name. Estée Lauder claims that this is a violation of their trademark. Trademark law protects consumers from being defrauded by one company pretending to be another. The question is whether an average user would expect that the offending banner ad belonged to Estée Lauder and therefore be cheated if it came from another company that would unfairly benefit from clickthroughs when the user intended to go to Estée Lauder's site. Usability theory to the rescue: Most users view Web banner ads as an irrelevant annoyance that is not related to their goal. The vast majority of banners encountered by the average user has a very tenuous relationship with the content contained in the body of the page. Thus, users will expect any given banner to be a random advertisement that is not related to the company they searched for. In fact, it is rare for users to see banners from companies they are looking for.
February 8, 1999
Consumer warning: From now on, we cannot trust any recommendations on Amazon.com since books get included on various recommendation lists based on how much the publisher pays Amazon. A very stupid move on Amazon's part: the fees they collect from publishers are not worth the utter erosion in their credibility. On the Web, trust is very hard to earn and very easy to lose, so even those parts of Amazon's service that remain truthful will now be viewed with skepticism by consumers. In Amazon's defense, promoting books based on kickbacks is standard bookstore practice: when a physical store places piles of a new book in a conspicuous location, it is often part of a co-marketing program with the publisher. But such physical promotions seem less like recommendations and advice and more like sales, so they don't erode the credibility of the store. You expect a shop to move product, but that's very different from whether you can trust their advice or whether the "advice" is noise. Amazon had attempted to integrate editorial content with sales content: a great help to users and a way to sell much more than through sheer sales pressure, but an approach that only works if the editorial remains credible.
Update Feb. 9: Amazon backs off and agrees to clearly mark any paid recommendations. A good step to rebuild our faith in their credibility. These guys are smart: they made a mistake of above-average stupidity, but we can all make mistakes. The difference between failure and success on the Web is mostly defined by the ability to quickly identify mistakes and correct them. Steve Outing's analysis of Amazon's gaffe is appropriately titled "Learning To Be a Publisher, the Hard Way".
February 5, 1999
New York Times reviews Victoria's Secret's Web broadcast: "barely visible images of models in white teddies." Having viewed a few minutes of the show myself (for research purposes, of course), I can only add that not only were the models hard to see; you also couldn't see the products. So maybe it doesn't matter that there was no interaction: no way to "click here to buy this bra" if you saw one you liked (because you couldn't see anything). The camera work and lighting further downgraded whatever slight image quality was provided by the postage-sized video. When will people learn that it is a very different skill to produce video for the Web than to produce video for television? Maybe 1.5 million users were curious enough to connect to the show, but I bet that most of them logged off again after less than five minutes (of the 21-minute video). Adding insult to injury, the VS site pops open another browser window when you follow the link to the fashion show. Keep your hands off my computer, please.
February 5, 1999
"Unique users" and "reach" are irrelevant measure of website success. Even though it is obviously misleading to do so, these measures give equal weight to:
February 4, 1999
AnchorDesk now has two million subscribers to its email notification. 100% growth in a year. To put this number in perspective, the weekday circulation of The New York Times is "only" one million. The two numbers are not easy to compare because the newspaper reaches households whereas the email reaches individuals (also, people can read the website without getting the email notification - and many people are cutting down on their email these days). But Jesse Berst may well be the most widely read columnist in the world now.
February 3, 1999
Bruce "Tog" Tognazzini has a test of user interface knowledge with questions like "name the 5 pixel locations that are the fastest to click on". I know the theory and I still didn't score 100% (gulp). There should have been one more question: "How could the Windows'95/98 taskbar be made more efficient?" (Answer: make the inactive line of pixels across the very bottom of the screen clickable. Read Tog's column to find out why it reduces user productivity to leave this pixel dead.)
February 2, 1999
The Economist says that Web stocks are overvalued and that "the Internet will give consumers more power and make it hard even for firms with a large market share to raise their margins for long."
February 2, 1999
Bill Gates says that he will continue to provide more integrated products despite the anti-trust suit. Good: I agree that integration often helps usability and that Web access should be integrated with the basic system. But let's get real integration this time, please. Windows'98 still has a lot of wrist-pain inducing double-clicks hidden around the UI (and is lacking most of the elements needed for a true Internet desktop).
February 1, 1999
Hal Varian has a great paper on the economics of information goods. I liked the paper better than Varian's celebrated book Information Rules: the paper is focused on problems related to running a website, whereas the book is too focused on issues related to running a high-tech company. The paper and the book include many of the same arguments and the paper is free (Varian would say that he makes more money off his book by pricing the paper at zero: it serves as a preview of an experience good).
February 1, 1999
A program scored 11% on the Turing test (percentage of people who thought that the computer was human while chatting with it). AI has a long way to go. I still believe in usability as a much better way of achieving good design.
February 1, 1999
Jesse Berst says that PalmPilot is making the same mistake Apple made in keeping their platform closed when up against an open alternative (IBM clones in the case of the Mac; a variety of CE devices in the case of the Pilot).
January 31, 1999
A user reports the following experience with Barnes & Noble: The website claimed to have a book in stock and that it would ship in 1-2 days. After placing the order, the user received a confirmation email saying that the book was not in stock and would ship in 1-2 weeks. In fact, it has still not been received after 3 weeks, at which point the user canceled the order. If this story represents more than a single, stand-alone bug, it is a disaster for B&N. On the Web, credibility and trust are everything because your company exists as nothing but glowing pixels on the user's screen. Lie to a user even once and you have lost that customer forever. I had been rooting for B&N recently because of their improved design and because we can't afford to let Amazon.com become a monopoly (as the fulfillment house for affiliate sites, they are Web infrastructure and not a simple shop). Maybe I should root for Borders instead: their site is improving and may have a chance of catching Amazon. (Amazon is still best, so I shop there while hoping for a competitor to "get it.")
January 29, 1999
Yahoo pays $3.9 billion (plus another billion in stock options) for GeoCities which hosts 3.5 million amateur home pages. In other words, each of these small amateur sites is valued at somewhat more than $1,000. This is playing in the press as proof of the value of large sites, but it could just as well be proof of the value of tiny sites. Why would some amateur's site be worth either more or less because of its hosting service? It's not as if GeoCities adds any unique value in terms of navigation or ability to find content. On the contrary, somebody who wants to find what "the people" is saying about a topic would be better off finding amateur sites through Infoseek or another search engine that covers the full Internet. In other words, we can now conclude that anybody who gets free space on, say AOL, and spends an hour or two posting photos of their dog has created $1,000 in value. Or alternatively, we can conclude that Internet stocks are over-valued.
January 28, 1999
AT&T introduces flat-rate calling: from anywhere to anywhere (inside the U.S.) for the same price for wired and mobile phones. The new trend to location-independent telecom pricing will be important for the next phase of untethered Internet use. I predict that we will see prices of less than a cent per megabyte for world-wide communication (from anywhere to anywhere on the planet) in less than ten years (pricing per packet makes more sense than pricing per minute). The Death of Distance, indeed. (death of RBOCs, too, unless they get high penetration for xDSL soon). By the way, the only way I could get from the AT&T home page to the new initiative was to wait for a news ticker to get around to showing the headline I wanted. Oh, misery!
January 27, 1999
A Forrester study of loyalty among Web users found four important factors that drive people to return to a website:
  1. High Quality Content
  2. Ease of Use
  3. Quick to Download
  4. Updated Frequently
All scoring above 50% among home users. Buzzwords that attract much attention in current Internet strategy books received minimal scores (14% or less) in the survey: coupons, brands, "cutting edge technology", customization, and chat.
January 27, 1999
Designing websites for kids, emphasizing the different considerations for each of the four age groups: 5 and younger (graphics and audio only), 6-9 (minimal text), 10-12 (research and creativity), and 13-17 (info about other teens). Additional privacy issues related to young users.
January 26, 1999
MSNBC polled readers about their use of portals. At the time of this writing, 1018 responses had been received. 3/4 of the users decided on which portal to use based on usability: "easy navigation" scored 55% and "fast downloads" scored "21%". 13% said "brand name content" (which can either mean "good content", "reliable content", or "famous content" - we don't know, but the first two are also design considerations). 81% use more than one portal on a regular basis: there just isn't much loyalty on the Internet (one reason high stock valuations are ridiculous). The only result that surprised me was that as many as 22% of the users claimed to click on banner ads (doesn't square well with the low clickthrough rates that have been recorded recently). But I guess that people might select that response category if they have ever clicked an ad.
January 25, 1999
Web advertising rates fell 6% in 1998, following my many predictions (most recently 3 days ago, see below) that advertising doesn't work on the Web.
January 25, 1999
Prodigy Classic closes because of Y2K problems. Data nugget in the story: Classic still has 208,000 subscribers and the new version (launched in 1996) has 433,000 subscribers. In other words, about 1/3 of the users have not upgraded yet. Further proof of my claim of increased user conservatism. We will probably still have at least 1/3 of the Web users on browser versions 1 through 4 in two to three years.
January 22, 1999
Whit Andrews found that only 40% of clicks on banner ads led to something useful. The remaining 60% linked to something "shameful." This is one of the main reasons advertising doesn't work on the Web (as stated here since 1997): Once users have been burned by enough badly done Web advertising, they won't bother clicking on any banners, even the ones that would have taken them to a responsible payoff page. Thus clickthrough is being cut in half every year (I predict 0.1% clickthrough rates by 2001 with correspondingly lower CPM). Web advertising is an example of "the tragedy of the Commons" as bad advertisers ruin it for everybody.
January 21, 1999
It can be computed from Excite's quarterly report that their revenues per page view were 1.1 cents. Much better than Yahoo's 0.5 cents per page view and slightly better than Infoseek's 1.0 cents and Lycos' 0.9 cents per page view. For some reason, none of these companies include this crucial measure of success in their own financial reports, but it's pretty easy to compute.
January 21, 1999
The merger of @Home and Excite worries me. While most media coverage focused on the stock market folly that a press release is valued at $4.4 billion, Scott Rosenberg (always good for insightful commentary) noted the fundamental impact on the nature of the Internet: To keep the Internet free, there needs to be a separation between three classes of services: bit-transport, navigation, and destinations. If a major ISP owns a navigation service (and the "portal" furthermore gives preference to paying destination sites instead of the best ones), then the Internet will turn into a collection of closed proprietary services. Just like the old days of Prodigy, Delphi, CompuServe, AOL, etc. Will you be able to send email to users of other services? Or download a video of more than 10 minutes from an independent website? The anti-trust people had better take a hard look at this merger: vertical integration is a much more severe threat against a freely competitive market than whether Microsoft puts a browser into the operating system (as long as IE treats all sites equally, of course). The best analogy to the @Home/Excite deal is the time when film studios owned large chains of movie theaters. Let's keep ISPs as common carriers.
January 19, 1999
A review of Web-based discussion group software concludes that there are no good systems available. I concur. This seems like a golden opportunity for somebody to make usability the design goal and create a good discussion group package.
January 15, 1999
The Web Consortium releases guidelines for making sites accessible for users with disabilities. Nicely done with priorities for things that are essential vs. things that would be nice. Goes far beyond simply adding ALT texts (though that's obviously one of the recommendations).
January 15, 1999
Why are all the books about the Network Economy bad? I have read many highly promoted books recently, but they are all disappointing: the books focus on the future of the technology industry, so they would be good reading if you have invented the next Java, PointCast, or Palm Pilot. But they have little helpful advice for an online toy store, a medical information service, or a vendor of earth-moving equipment. The perspective of most books is that of thinking up ways to force the "others" out there to use your technology. Never what these folks would use this technology for in their own business. Even though it is not perfect, I think that Kevin Kelly's New Rules for the New Economy: 10 Radical Strategies for a Connected World is an exception. It actually has food for thought for people working on Web strategy in non-tech firms. (I recently bought customers.com - I will post my impressions here once I have read the book.)
January 14, 1999
Despite increased profits, Steve Jobs says that Apple will not revive its R&D spending. He says that "focus and leadership" is more important than research spending. True, if you are aiming at incremental advances. But if you want the next big thing, then you have to research 10 wild ideas, 9 of which never pan out. More evidence that Apple's strategy is to fleece its remaining loyal customers while slowly closing down the company (or at least the Mac).
January 13, 1999
Nice to know that there is at least one place where I beat Bill Gates: The Industry Standard has a who's who of the Internet economy: this week's statistics show that my entry is number 7 and Bill's is number 8 in terms of readership.
January 13, 1999
In November 1998 I calculated that Yahoo had 0.4 cents revenues per page view. Their most recent quarterly results show 0.5 cents per page view, though is is impossible to say whether the improvement is simply due to seasonal variations over the holidays. This type of usage analysis is often left out of current press coverage of Internet stock exuberance.
(January 13)
January 13, 1999
Dave Winer found a great article about a highly embarrassing situation for Paramount: it seems that their Web strategist forgot to read my Alertbox about why Web pages must live forever.
January 12, 1999
Only 0.2% of e-commerce sites have good customer service (The New York Times). Part of the solution is obviously better service, but it is also necessary to improve site design so that users can serve themselves.
January 11, 1999
CNet's Builder.com has a good info package on how to design websites for kids. The presentation is clearly designed for the Web with modularized content and rich linking to their own site and resources from across the Web. Too many other sites never link to the outside under the theory that they don't want users to leave (guess what: the users know where the Back button sits, so they are going to leave anyway if you don't provide the highest possible value - including outbound links).
January 9, 1999
Amazon.com does so many things right that it is an unusual pleasure to report that Barnes & Noble does have one usability advantage: better use of progressive disclosure and content prioritization. Look, for example, at B&N's search for books by a specific author: out-of-print books are (correctly) relegated to a secondary page (progressive disclosure), and the list of in-print books is (correctly) sorted by the probability that the user wants the books. Search listings on Amazon.com confusingly intermix in-print and used books; even though some prioritization is used to show best-sellers first, the full list contains duplicates of the short-listed books (making it harder to scan and understand). Another mistake at Amazon: the search results do not have a unique URL, so I can't link to it (you have to do the search yourself). I still like Amazon.com best: lower prices and better book-description pages, but B&N wins on search. Congrats, B&N: this is why we need multiple websites in each category.
January 9, 1999
It is now possible to make Google the default search engine for both IE and Netscape. Why not go to a useful search engine rather than the one that paid the browser vendor the most money.
January 5, 1999
Microsoft launches a special MacTopia subsite with targeted information for their Macintosh customers. Great idea to extend the trend toward vertical portals and embed a hub on the site itself. But unfortunate "out there" design: lose the pulsating logo, please. And it's a disaster to deviate this much from the general site design and navigation. Microsoft's webteam just completed a commendably user-focused redesign and then they throw away all the gains in the new, completely inconsistent, subsite. To compound the problem, the IE for Mac page has a third design. Fragmenting website spinning out of control faster than you can say "department infighting".
January 5, 1999
Appropriately skeptic review of the electronic books: "the type on both machines is one generation removed from dot matrix"; with comments on better electronic text to come in the future. (Seattle Times)
January 1, 1999
No news will be posted until January 5: I am on vacation in Big Sur without my laptop. As the Internet becomes ubiquitous, true luxury will involve being offline and out of touch. But check Tomalak's Realm for daily links to the best content on the Web that day.

Other Months in 1999

Other Years