- December 30, 2001
- Godiva Chocolates wins my award for best recipient user experience for e-commerce this holiday season. I have received quite a lot of gifts that were sent through e-commerce sites, and the package from Godiva was the best in several ways:
- Clear indication that it's a present: big "GIFT MESSAGE ENCLOSED" printed on a small pouch on the outside of the box (very important in these times of some uncertainty about unexpected mail deliveries)
- Easy to find out who gave me the present (at least if the sender included their name in the gift message, which people do tend to so)
- Very nice and Christmassy wrapping and presentation
In contrast, the worst shipment I received didn't say anywhere who the gift was from - I only discovered after quizzing various friends and family. I have also received many gifts where one had to search through the shipping peanuts and other elements of the packaging to discover the gift message - or cases where the gift message was well hidden among SKU listings and other codes on the packing slip. I even scared a friend myself this Christmas by having him sent a gift that arrived in an anonymous box without the company's return address. He almost dressed up in a biohazard suit before opening the package.
Lesson: e-commerce sites need to work on the recipient experience - not just out-of-the-box but even before you open the box.
- December 24, 2001
- NORAD tracks Santa Claus' travel path. Nicely done multi-lingual kids' site from the North American Aerospace Defense Command.
- December 16, 2001
- In praise of permanent collections:
Over the weekend I visited the San Francisco Museum of Modern Art which is currently hosting a special exhibition of Ansel Adams photographs. The special exhibit galleries were so crowded that you had to wait in line to see every single print. The experience of visiting the exhibition was like buying sausages in the old Soviet Union: one long line slowly snaking through several rooms. In contrast, on the floor below the special exhibition, the permanent collection of photography had approximately one visitor per gallery, even though it contained equally great pictures (including one by Ansel Adams). Lessons for web design? Not many, since art museums and websites are as close to opposite ends of the design spectrum as you can get, but there is certainly one commonality: Waiting is no fun; show me what I came to see.
- November 28, 2001
- Macys.com has increased its conversion rate by 150% after a redesign that mainly focused on improving the usability of the site's search feature. The magnitude of the sales increase confirms the potential for improvement on current websites: most sites continue to lose most of their prospects because they are too difficult to use.
The fact that improved search accounted for most of the increase in sales does not surprise me. We know from our research last year that website users only succeed in half of their initial queries.
Macy's search improvements are partly due to a software solution called IntuiFind, which costs $250,000 and addresses the verbal disagreement phenomenon (the fact that many users will use different query terms than the vocabulary used by the webpage author). Smaller sites can get the same benefits from a manual analysis of their query logs and by following more of the 29 usability guidelines for search design: the average site only complies with a third of these guidelines, so it is no wonder that users have trouble searching on these sites.
- November 27, 2001
- The New York Times has impeccable ethics. When they make a mistake, they fess up. The newspaper ran a big article today disclosing that its own website helped promote a stock-touting scheme that will probably result in big losses for many elderly readers. Subscribers received an email reading "From NYTimes.com, A Special Investment Report from StockTopics.com [...] Aggressive Growth! Strong Buy Recommendation!"
It would be quite understandable if many readers assumed that the dubious stock recommendations in the email were editorial recommendations from the Times. Savvy Internet users might appreciate the subtle distinction between the two "from" elements: the email was from NYTimes.com and the investment advice was from StockTopics.com. But we know from many usability studies that most users, and especially many elderly users, have a less thorough conceptual model of the Internet. I know somebody who for a long time thought that Jesse Berst was writing him a personal email every evening, even though Jesse's newsletter was circulated to millions.
The touters also had their email sent out through CBS MarketWatch, and to CBS's credit, that website has a headline at the bottom of the homepage today, linking to the Times article about the calamity.
The stock-touting email from CBS MarketWatch contained the disclaimer that the site "expressly disclaims all warranties in connection with the material in the e-mail and shall not be liable for damages arising from any statements contained in it." I guarantee that a lot of users would have blanked out this beauty of legalese. People are so used to useless disclaimers on investment websites that they don't read them and don't understand them. That's why investor protections would increase if regulators would lighten up on their demands for worthless disclaimers of the type "past performance doesn't guarantee future results."
- November 26, 2001
- WAP continues to offer a substandard user experience that makes it useless for all common applications. When I released my report on WAP usability based on field studies in London a year ago, WAP supporters complained that it was unfair because the technology would continue to improve and would get more users. Today, it is clear that almost nobody uses the WAP features on their cellphones. A small supplementary study this month found WAP as hard to use as it was last year.
Even worse, WAP services continue to include design mistakes that we documented in our report last year. For example, I tried to check the stock market on a WAP phone. The service forced me through the following awkward set of steps:
- First, I had to wait through three splash screens:
- Microsoft Mobile Explorer (software provider) logo
- Vodaphone (cellular carrier) logo
- Vizzavi (WAP portal) logo
Considering how slow WAP is and that users are paying by the minute, there is no reason to force them to download all these logos every time.
- Select WAP access (other options included email, though that was also an option on the top-level phone menu, so why add an additional step here?)
- Select financial info
- Select financial indices
- Get a screen that lists several financial indices:
- Dow Jones Ind. Index
- FTSE 100 Index
- DAX Index
- Select one of these indices to get the first screen with real content: the current value of the index
Several of these steps could have been eliminated, but worst of all, why didn't they follow our advice to provide the content (index quotes) on the screen that named all the indices. The word "Index" could have been eliminated from each line, and the space used to provide users with immediate answers instead of forcing extra navigation. Since I just selected "financial indices", there is no need to explicitly state that each of the options is an index.
- November 21, 2001
- Tidbit from Europe: The portal Scandinavia OnLine (sol.dk|no|se) is being sold for $49M. The portal holds $37M in cash and the buyer will take over tax deductions worth $15M. Thus, the portal business itself, with 8 million "unique visitors", has negative value (-$3M). I have never thought that generic portals were the best use of the Web (specialized, targeted sites are better), but the negative valuation is interesting.
- November 8, 2001
- My new book, Homepage Usability: 50 Websites Deconstructed is now in bookstores in the United States and can be ordered from Amazon and others. It will reach other countries by the end of the month.
- 113 usability guidelines for homepage design which you can use to score your own homepage. Get less than 80% and you need a redesign project. Get less than 50% and you better make it a top priority to improve the usability of your homepage before you lose any more business.
- We are also including statistics showing how frequently major websites use different design elements on their homepages. This chapter may be dry, but it's very useful in settling design arguments.
- The most fun part of the book is certainly the detailed usability critiques of 50 homepages, all shown in big, beautiful color pictures with plenty of bubbles indicating their design mistakes.
This is the first coffee-table book about usability and will be a great gift item this December if you need a present for somebody who is interested in web design. Buy it yourself to increase the business value of your company's website. Buy it for others to give them the joy of looking at the many pictures and the biting commentary.
- October 26, 2001
-
My new book, Homepage Usability rolling off the assembly line
(in bookstores November 5).
- October 26, 2001
- I can't believe this: XP has a huge productivity benefit that Microsoft neglects to trumpet in its advertising and promotions. We are finally getting anti-aliased text in the operating system. Back in the mid-1980s IBM research found that anti-aliased text increases reading performance by 10-20%.
The annual productivity gain for a business professional will be $2,000, or about the price of buying a nice PC with XP on it. (I discovered this major selling argument for XP, not from Microsoft's mighty marketing machine, but through a small link on Dave Winer's site.)
- October 25, 2001
- Somebody hire Bill Gates a tech writer: The Windows XP "upgrade advisor" will check your existing PC to see if it can be upgraded to XP. As stated on the web page for the advisor, "some of the language in the tool may imply that you are actually installing Windows XP" because they are reusing the same software that scans your system in the initial phase of an actual installation. Come on, what would it cost Bill to have a writer edit the messages for this tool? He is imposing a cost to the world economy of $3 million in lost productivity by causing confusion among the millions of people who will use the tool.
- September 30, 2001
- Case Studies Wanted: Measurable Impact of Design Changes
I am collecting examples of metrics that quantify the impact of design changes. If you have data you are willing to share, please email Shuli Gilutz at shuli@nngroup.com
Any type of user interface is of interest: websites, intranets, mobile devices, traditional software, consumer electronics, etc. Also, the design changes may refer to a complete redesign of the total UI or it could just be a single feature, screen, or design element that was changed.
Measures of interest include anything that measures how the use changed. Could be an improvement, but data that documents how a design change had negative results are just as interesting.
Example metrics:
- conversion rate
- size of average "sales basket" for e-commerce
- training time needed to learn a feature or system
- sign-up rate for newsletters or other desired action
- time on task and other productivity measures
- subjective satisfaction scores
- average page views or duration of visit per user
- number of times a feature was used
- percentage of users who bail out at a certain step
- number of calls to help desks or tech support
- learning or comprehension scores
- other usage or impact metrics, including new ones you have invented yourself
For whatever metrics you have collected, please send both the "before" and the "after" number as well as a short description of the kind of system you were measuring and the type of change made to the design. If you can send us before/after screenshots, so much the better (whether or not you will allow us to publish the screens), but we are also interested in data where you cannot send screens.
We can keep your contribution totally anonymous and purely report your numbers as statistics that don't include any indication of what company or design they refer to. Of course, we would prefer to give you full credit if you are willing to be named. Also, if you send us screenshots, please indicate whether we should treat the screens as confidential info or whether you will allow us to publish the images. If I get enough good case studies to warrant a full report, then all contributors will obviously get a free copy.
If you can help in this project to document the range of measurable effects that can result from a user interface redesign, please contact Shuli Gilutz (shuli@nngroup.com) who is the member of my team responsible for collecting the data. She can also answer any questions you may have about this project.
- September 11, 2001
- When a major emergency happens or a big story breaks, websites often experience a flash crowd as many, many times the normal traffic hits their servers. At such times, it is best to change over to a very lightweight homepage without any graphics or fancy formatting. At the time of this writing (10:13 AM Pacific time), MSNBC.com has scaled back its homepage somewhat and CNN.com is serving a truly simplified homepage. In contrast, BBC.co.uk is saying that service may be "slow due to the weight of traffic" but they still have too many images on their homepage. Once things move fast, it is hard to redesign a homepage (and interior article pages) on the fly, which is why I have often recommended that sites have contingency plans in place and have a pre-designed lightweight page template ready to be put up as required. Some of the things that can be done: eliminate graphics, eliminate scripts, rollovers, etc., pull ads, replace fancy logos with a simple one (as done by The New York Times), eliminate stylesheets (too bad, but any extra hit on the server must be avoided during a flash crowd), and place staging servers and other extra machines into production. (I thank Paul Nagai for several of these suggestions.)
Historical note:
I first used the term "flash crowd" as relating to the Web to refer to the overloading of the webserver for the 1994 Winter Olympics in Lillehammer. I borrowed the phrase from a science fiction story by Larry Niven that referred to the crowding that would happen if the world had Star Trek-type transporters and curious people could instantaneously beam into hot spots; the Web is experiencing this phenomenon virtually.
Compare screenshots:
- September 8, 2001
- Doc Searls correctly explains why Microsoft is less evil than AOL. Micro-summary: Microsoft will win because it views users as customers. When a Linux guy says that MS is less-than-fully-evil, you gotta pay attention.
- September 1, 2001
- President Bush launched a new version of the whitehouse.gov website. In some ways it is much improved over the previous site: substantially more content and real edited decks (story summaries below the main headlines). Good use of breadcrumbs to support navigation inside the site. But kill the rooms metaphor for the site structure: the site is not about the building, it's about the institution and nobody outside the Beltway knows what the Blue Room is.
The new homepage complies with 68% of the usability guidelines for homepage design. This is a fairly high compliance rate compared to many other websites but still less than one would expect from such a high-profile and well-funded project. (To be fair to the White House team, the homepage usability guidelines have not been published yet; the first public presentation of the full list will be October 22.)
The design includes several good elements to enhance accessibility, but employs a close-to-unreadable cursive script on a busy background for the main navigation buttons and a very faded photo of the White House itself. Both will hurt low-vision users. The ALT texts are over-done with extraneous word count that will clog screen readers. For example, the text for the faded photo of the building included the irrelevant info that the graphic includes the words "Welcome to the White House" (even though the blind user will have heard the welcome read out at the beginning of the page).
The biggest news may be the attempt at making the site multi-lingual. Often well-done, with direct links between corresponding content in different languages. But many weaknesses as well: Try to go to the White House homepage and ask for Spanish. Most likely you will not end up on the Spanish homepage because there are too many other links to Spanish content. The most prominent link is one that goes to an overview of political initiatives. Worst of all, when you link to the Spanish subsite, the navigation remains in English, and the homepage link takes you to the English homepage and not the Spanish homepage.
- August 29, 2001
- As of August 29, The Industry Standard website is still operational, including the full archive of the printed magazine as well as a scaled-back news service with recent articles. Good. It's interesting to note how much cleaner, simpler, and pleasant their homepage looks now that the site has scaled back. Especially if you run a utility like WebWasher that eliminates blinking ads.
- August 19, 2001
- Commenting on the demise of The Industry Standard, Dave Winer says that it would be "a final act of kindness to the Web to keep the URLs live even after the magazine has ceased to publish." I have commented many times on the need to prevent linkrot and the value of archives of old articles. The question is whether there is a business model for saving the archive of a dead magazine now that the site cannot sell subscriptions. Even without having to pay reporters and editors to create new content, the pure cost of bandwith, server maintenance, and even a single ad salesperson may be more than the advertising revenues. If nothing else, I hope that the owners put all the pages in safe storage and bring them back once we get micropayments.
- August 19, 2001
- Interesting dilemma: due to an editing error, the email announcement of this week's Alertbox column was sent out without the URL. Usually, users can click from the email to the column, but not this time. What to do? Send another email? No, because the mailing list signup specifically promises to protect subscribers' mailboxes and send them no more than one message every two weeks. Instead, I decided to rely on the ability of almost all subscribers to know how to go to useit.com and click from the homepage to the current column. Not ideal, but better than overloading peoples' mailboxes. Of course, the real lesson is to proofread email newsletters, not once, but twice before sending.
- August 17, 2001
Back from a few weeks' vacation in Africa. One of the interesting sights was this billboard which was all over Zambia.
- August 15, 2001
- Good article lambasting Macintosh system software for displaying ads even though the user paid Apple for the Mac (and for the OS upgrade). We need to reassert respect for customers and their rights to own what they have bought. This also goes for the telecoms industry which often tries to keep paying customers within a walled garden of mobile content instead of giving them the best service for their money. The Mac article is particularly notable for expressing the frustration users feel when they lose control over their own purchases and it becomes clear that they are marks and not valued customers.
- August 14, 2001
- Is Bluetooth doomed?
I don't know, but the author sure has some compelling arguments. Also, being up against a version of Ethernet is a tough battle that I would not wish on anybody. All I really want is seamless integration between devices and the death of cables. Who cares whether the solution is called 802.11, except that it's more nerdy than the old Viking king's name.
- July 16, 2001
- New book: My design guidelines for e-commerce user experience are now available in full-color printing and hardcover binding.
All guidelines are based on findings from detailed usability studies of twenty e-commerce sites with users in the United States and Europe.
- July 12, 2001
- Yahoo's revenues for the second quarter of 2001 amount to 0.2 cents per page view. This is the same value I computed for Q1 but only half of the 0.4 cents per page view they made from 1998 to 2000 (some quarters they even made 0.5 cents). Luckily, daily page views are up substantially, so even at half the value per page, I think Yahoo can make it back and become profitable again.
- July 9, 2001
- The online game for the movie "A.I." is a great example of innovative use of the networked nature of the Web to create a new media form. Unfortunately, this specific game suffers from being overly difficult. This makes the game exciting for a small number (7,000) of hard-cores, but is a turn-off for the rest of us. However, it will be eminently possible to create other forms of networked scavenger hunts that allow for multiple levels of difficulty and involvement.
- July 5, 2001
- Reading user reviews of Ultima Online it is striking how many of the negative experiences were related to difficulties getting started with the game. New users were summarily killed off by experienced players or they had to spend lots of time in boring activities before getting any excitement. A newer release of the game, Ultima Online Renaissance, has supposedly addressed some of these problems. In any case, it will be a challenge for more and more systems in the future how to provide a multi-user environment that accommodates both experienced and novice users without allowing the experts to walk all over the novices. eBay, anyone? Computer games in particular are notorious for appealing to a very narrow segment of the population and being either too difficult or too boring for users who are not fanatical gamers. My best idea right now is a "handicap" system like that used in golf: novices get extra points to put them on a more equal footing.
- June 26, 2001
- Scott McCloud discusses micropayments and Napster/music sharing in his latest online column. The format of the column is interesting as one of the very few examples of content that is designed for the Web. The modified comic strip format crystallizes the content in a way that pulls in the reader and presents a relatively large amount of information without having to read too many words. The content of the column is interesting as well: I completely agree that the new economy is on hold while we are waiting for micropayments and the associated ability for websites to connect to their customers in a two-way flow of services and money being exchanged simultaneously.
- June 14, 2001
- The telephone industry has released the specs for a next generation of mobile phone Internet access called M-Services. They provide a Flash demo of one of the new phones (good use of Flash to show a changing object; a better use would have allowed the user to experiment with the phone). It is very hard to assess usability based on a simulated demo, but the new design does look promising relative to current WAP phones for experienced users. Novice users might suffer worse problems - a usability test could tell. The design borrows the GUI paradigm without GUI input devices: still using painfully indirect manipulation through pushbuttons. My preliminary judgment: progress, but will probably ultimately fail because the telephone is the wrong device for mobile interactivity. You can put prettier lipstick on it, but it remains a pig. The "Danger" device sounds much more promising.
- June 10, 2001
- Dan Gillmor wrote an excellent column on how network computing can turn nasty if providers care more about what they can do to their users' eyeballs than what they can do for their customers. This is why it is so important for users to pay for services. In fact, maybe it should be a buying criterion that the service provider makes all of its money from paying customers and nothing from other sources that would cause them to treat the customers poorly. It's probably well worth paying the extra money in order to get better service as network computing moves to the center of our work experience and even parts of our leisure time.
- June 8, 2001
- The debate about the "smart tags" in Internet Explorer overlooks the history of this feature (which automatically adds links to certain words when displayed in the browser). In hypertext theory, this is known as implicit links and is quite a useful feature. Long before the Web, the InterLex system provided links from any word to a dictionary and the Video Linguist allowed language students to link to translations of words in foreign-video subtitles. More recently, several browser-add-ons have provided implicit links on the Web, such as the ability to link from a company name to a stock quote pulled from a different site or to perform a query for the term on Google.
There is no doubt that "smart tags" are useful. There are just two problems with the implementation:
- Use of squiggly underlines as link anchors: in rich hypertext you don't want visible link anchors since they will clutter up the screen and reduce reading speed. When most things are linked, you don't need marks. Use a technique similar to HyperCards and display the anchors when the user holds down a modifier key.
- If the link destinations are all Microsoft properties, then the feature obviously becomes an abuse of power. It is good for the browser to include hooks for implicit links, but users should subscribe to a variety of services for the destinations.
- June 7, 2001
- I recently made two purchases through ads on Google. I have always said that there are two exceptions to the general rule that advertising doesn't work on the web: classified ads and search engine ads. Even so, it amazed me that I spent money twice in a week based on Google ads. Why do they work so well? Several reasons:
- The ads are text-only meaning that they will actually be seen by users.
- They are precisely targeted at people looking to buy specific items. In my case, the two query terms that led to purchases were:
- The name of my former long-distance provider which went out of business recently. Buying placement on searches for defunct providers is a great way to attract new customers since these people are actively looking to move.
- "Conference calls."
- Both queries related to telephony service, which is easy to advertise in a format that is easy to scan: cents per minute. Thus, the offers were incredibly clear in the ads, making me more likely to click on them. Admittedly, many other B2B services are much harder to describe (but that's why writers ought to make the big bucks in the Internet industry).
- June 6, 2001
- WIRED reports that e-books are doing poorly. Not a big surprise for me, since that's what I predicted three years ago. In contrast, e-reports are doing well. As an example, I am selling about a million dollars a year in downloads of my usability guidelines for things like designing the PR section of a corporate website, designing for international users, and improving checkout and registration. Why the difference? Two major reasons:
- Content: Practical guidelines for use in projects are suited for distribution over the Internet, since people want instant gratification the day they are having a design discussion. In contrast, novels and other traditional book content work well through traditional distribution channels.
- Format: I use standard PDF that anybody can print out without further ado. E-book publishers use proprietary formats that require special readers that users don't have (sometimes even proprietary hardware as well, which people don't want to buy). E-books also employ annoying copy protection schemes that make it hard for people to read the content they have paid for. As always on the Web, simplicity wins.
- June 4, 2001
- FirstGov: please rethink.
I have recently visited some federal government websites which have started to feature a FirstGov logo. See, for example, the left-hand navigation bar on the White House or the Air Force sites. (Note that these two versions of the logo are inconsistent: the White House includes the tag line "Your First Click to the U.S. Government" in a tiny, unreadable, font.) This placement is likely to be ignored by many users since they have been trained to assume that such graphics are irrelevant ads. As a further problem, the name alone is insufficient to communicate the services provided by the site, so simply putting the FirstGov logo on, say, the Air Force home page is also not going to help users discover the service, even if they do consider the graphic. As it turns out, FirstGov is a portal site that seems quite useful. I have not conducted a full usability review of the FirstGov site, so I don't know if it is in fact any good. However, it seems useful enough that it should be better integrated with these other government sites and not treated as an add-on.
- June 1, 2001
-
Weird new service from Amazon: if you have JavaScript enabled, you will likely see a box here with five book links from Amazon.
The five books are selected from books that have been purchased frequently by readers of Useit.com (more precisely: bought by people who follow links from Useit to Amazon). The contents of the box come directly from Amazon's server and they keep track of all the sales statistics. Thus, if you reload this page, the box should update and show a fresh set of books that Amazon has been selling recently.
I have mixed feelings about this idea. Most important, I view it as the responsibility of the website editor (i.e., me) to recommend books. I don't like "recommending" things I have not read (in principle, it's Amazon doing the recommending, not the website, but most users will not understand the difference).
On the other hand, the auto-updated recommendations leverage the collective intelligence of the readers. People who are smart enough to read Useit should on average be attracted to good books.
- June 1, 2001
- Salon has an unreasonably pessimistic analysis of the future of online music, claiming that everything is over and that traditional record companies are basically all that's left after Napster lost. Well, Napster deserved to lose since it was based on ripping off other people's work. That does not mean that we cannot see new models for distributing music online that has been developed for the Net to begin with. Sure, the traditional recording companies own the currently famous music stars, and they deserve to own the stars since they paid for them. New stars will emerge, and they don't have to sign away 99% of their sales to a recording company. New talent can establish itself by distributing new recordings at a very low cost - or even for free. Then, as they get more famous, they can increase the price per download and still be much cheaper than the current cost of a compact disk. We might also see completely new business models, such as the ability of a fan to micro-invest in an up-and-coming musician: maybe pay $100 to a new rock group you have discovered in return for the rights to get all their future recordings for free. Finally, some really big stars are out of copyright: how about Mozart on the Internet? Anybody could set up a service to experiment with ways of optimizing personalized mixes of classical music and creating single-listener radio stations.
- May 27, 2001
- My review of the documentary film Startup.com. Three-word summary: go see it.
- May 25, 2001
- I have come across one of the rare examples of a Shockwave animation that works: the skateboard tricks demo at Sports Illustrated For Kids. (Ignore the annoying introduction and music.) Animation and 3D are appropriate in this design because:
- the animation is actual content and the best way of showing how the skateboarder moves
- the multiple 3D views enhance the user's understanding of the trick
- limited controls make it easy to change the 3D view and avoid the "lost in 3D" feeling that comes from full freedom of movement
- it's short and thus not boring (except for the intro which takes too long and yet is too hard to understand)
- May 24, 2001
- The White House has updated its website to fix a usability problem I pointed out two days ago in an interview with WIRED: earlier in the week, the White House for Kids page consisted of a series of attractive icons, all of which were inactive. We have known since 1994 that "under construction" doesn't work, so I was appalled to see this design on a high-profile site seven years after I published the finding. But good work should be acknowledged: the whitehouse.gov staff fixed the usability problem in two days and replaced the offending icons with a set of simpler text links that all lead to appropriate kids-oriented content. Bravo.
- May 22, 2001
- Reviews of RadioShack.com and Dell.com by Kara Pernice Coyne, senior user experience specialist in Nielsen Norman Group.
- May 21, 2001
- The New York Times is highly critical of the usability of the new Infiniti Q45 (access to the Times requires free registration). The review of the new car Sunday asked the rhetorical question, "If you find Windows 98 frustrating at the office, do you want to operate a scaled-down version at 70 m.p.h. on a rain-slicked highway?"
Because of the Times, I now know what car not to buy. Two lessons:
- As products become more and more complex, the press must focus on usability in reviews of categories far beyond computers. Consumers need this protection because it is hard to judge usability before you have bought and tried the product: I always say that one cannot assess usability by asking people to look at a design; they have to use it.
- It's amazing that the designers of a $59,000 product like the Q45 could not afford a usability engineer to simplify the user interface. But like many manufacturing-oriented companies Infiniti probably didn't consider software usability to be a real profession, relying instead on the judgment of their automobile engineers. (Or, if they did have a usability professional on staff, the engineers clearly got to override his or her recommendations.)
- May 6, 2001
- Two years ago, I predicted that poor intranet usability would cost the world economy $100 Billion in 2001. I just computed a current estimate of the cost of poor intranet usability for Business 2.0 magazine. Surprise -- the result is exactly as predicted: $100 Billion lost per year. Think about this next time you cannot find something on your corporate intranet - they were warned years ago, but most companies chose to spend all their money on public websites. Of course, it's great to improve usability for customers. Still, it's good business to invest in employee productivity as well.
- May 5, 2001
- David Strom reports on the difficulties in finding out whether an airline will offer in-seat laptop power on a specific flight. Classic example of the usability problems caused by weasel language like the site telling users that "your seat may be equipped with a DC power outlet" (my emphasis). Strom only briefly touches on a second problem: each airline requires you to buy a different power adapter, and there is no simple way of buying them.
- May 4, 2001
- One of the best article I have read about Internet marketing. Aaron Goldberg (VP at Ziff-Davis Media) explains in three screenfulls what took me a 432-page book: the Web is not for building awareness; its place is "further into the marketing process. That is, specifically, preference- and purchase-building activities."
- April 26, 2001
- Happy to see that "la bibliothèque idéale du développeur web" (the ideal library for Web developers) at Amazon.fr has ergonomics as the first category following basic principles. That's how it should be, but not how it used to be.
- April 21, 2001
- The UPA'2001 conference program is now online. The Usability Professionals' Association conference is the main annual technical conference for people who are serious, full-time, hard-core usability experts. Highly recommended. (I am presenting the closing plenary on the topic of Usability--Past, Present, and Future, taking a twenty-year perspective on the discipline.)
- April 12, 2001
- Interviews with participants in the final stop of the User Experience World Tour: Is the Slowing Economy Helping or Hurting Usability?
- April 11, 2001
- Yahoo's quarterly results for Q1 of 2001 show that they made 0.2 cents per page view. This corresponds to a CPM of $2, though some of the revenues derived from non-advertising sources, so the average CPM must have been lower than $2. Of course, we all know that nobody pays list price for Web advertising, but a more surprising fact comes from comparing this latest quarter with earlier results. For as long as I have tracked this statistic, Yahoo has been making 0.4 cents per page view. This was even true for the last quarter of 2000, so having the value of a page view drop to half of its historical value is big news. (The one exception to the 0.4 cents statistic was Q4 of 1999 when Yahoo made 0.5 cents per page view - in retrospect, the 1999 holiday shopping season was the high of advertising folly.) Well, I have said since 1997 that advertising doesn't work on the Web. I used to say, "with the exception of Yahoo and a handful of big sites." Too bad that even this exception doesn't hold any more, since I am a great admirer of Yahoo's design. I have faith in the company's future exactly because the site has great usability: if anybody can convert users into paying customers for value-added services, it's Yahoo.
- April 7, 2001
- My latest analysis of search referral logs shows that Google now dominates Internet search, accounting for 46% of the search referrals directly as well as many of the visits from Yahoo. Google went from number six in 1999 to a clear winner in 2001, showing that the Web is nowhere close to being locked down: there is still plenty of opportunity for new sites. MSN also exhibited strong growth, whereas AltaVista, LookSmart, and Search.com may be in trouble. A similar analysis last year provided a strong predictor for the death of Go (formerly much beloved when it was called Infoseek).
- March 29, 2001
- Esther Dyson's PC Forum was the usual schmoozefest for the high-tech intelligentsia (Steve Hayden started his talk by saying, "hello, smart people"). My old boss, Eric Schmidt was there as usual and everybody, including me, congratulated him on having joined Google.
The mood at the conference was more somber than last year because of the stock market. Marie Tahir from my group was also at PC Forum and said that it reminded her of visiting Poland after the Solidarity movement had been suppressed by martial law. An initial euphoric feeling of freedom where everything seemed possible had been crushed, leaving people more depressed than before, saying "will it ever be good again?"
Poland became a democracy a few years later, so let's hope that the network economy will revive as well. For sure, PC Forum demonstrated that innovation is still alive, with several new technologies being shown. A standing-room-only crowd attended my usability session.
Both of the co-inventors of the spreadsheet were at PC Forum:
- Bob Frankston insisted on using the term "first mile" for the link between a house and the public network, thus taking a more user-centric view than the telephone companies' preferred expression, "last mile."
- Dan Bricklin showed me his newest essay where he argues that is it OK for some things to be difficult to learn. General-purpose devices like cars and computers provide sufficient subsequent leverage to pay us back for the initial investment in learning. I agree, and even while I think that general-purpose tools should be as easy as possible, I also think that there are benefits from powerful tools that can be applied widely. The argument for power only applies to a small number of general-purpose tools: the majority of interfaces should be at the receiving end of these tools and must be very simple so that the tools can be applied. In particular, Web browsers and other client software need more powerful features; websites need simplicity.
- March 20, 2001
- Report from panel at the World Economic Forum in Davos:
Can Technology Alleviate Poverty?
- March 15, 2001
- The sales statistics for the WAP Usability Report show three things:
- Usability has become a world-wide concern: we have sales to countries from Kuwait to Slovenia
- Most interest in usability remains focused in the United States, the U.K., the Scandinavian countries, and Australia. We still need to evangelize more in the rest of the world.
- An American e-commerce site really can have 60% of its sales outside the U.S.
Sales for December 2000 to February 2001:
| USA | 40% |
| Canada | 4% |
| Latin America | 2% |
| U.K. | 16% |
| Scandinavia | 9% |
| Germany | 3% |
| Spain | 2% |
| Netherlands | 2% |
| France | 2% |
| Other Western Europe | 6% |
| Eastern Europe | 1% |
| Israel | 2% |
| Asia | 6% |
| Australia/NZ | 4% |
| Africa | 1% |
As part of the expanded evangelism program, we will license translation rights to the WAP report, our e-commerce design guidelines, the guidelines for designing PR sites, and the Alertbox. Please contact my project manager for international, Luice Hwang, at hwang@nngroup.com if you are interested. (Note: rights to the Alertbox already sold for Japanese, Korean, Spanish, Portuguese, and German. Sorry.)
- March 9, 2001
- Dave Winer presents two good ideas for how Yahoo can make money:
- charge for their mailing list service (I agree: the reason I use SparkLIST rather than Yahoo for the Alertbox mailing list is that SparkLIST allows me to be a paying customer and keep control)
- an easy way for list owners to collect subscription fees with a cut of the take being kept by Yahoo
The last idea (credited to Scott Loftesness) is particularly brilliant because it leverages the networked nature of the Internet. You can have a business that consists of nothing but allowing other people to have a business and collect money that would have been too difficult to collect in the past. So far, eBay has been the best example of this approach, but maybe Yahoo's mailing lists will be the next.
- March 7, 2001
- A review of the new VisorPhone confirms the potential for the tablet form-factor to deliver significantly improved usability for mobile telephony. Traditional phones must die. Tablet communicators will be especially important for delivering mobile Internet services; something that still seems to be lacking on the VisorPhone because it does not come with integrated Internet access. When customers have to do something special to add-on a service, many won't do so, and the service will not become ubiquitous. If connectivity cannot be taken for granted, then software developer will not design for it and service providers will be less likely to deliver services for the new platform. One of the main lessons of the Macintosh was the benefit of shipping a mouse as standard with every Mac ever sold: this guaranteed the success of the graphical user interface in an age where all other software was keyboard-driven.
- March 1, 2001
Marie Tahir and Jakob Nielsen enjoy summer in Sydney as the User Experience World Tour visits Asia/Pacific. Tahir reports on usability issues in each city:
- Tokyo: Obstacles to Iteration in a Gadget-Driven Culture
- Hong Kong: Design and Usability Challenges in a Cultural Portal
- Sydney: Strategic Survivors Flourish Down Under
- February 18, 2001
- Kicking off the Asia-Pacific part of the User Experience World Tour: Tokyo tutorials today; Main Event tomorrow. Hong Kong later in the week. More detailed reports later, but I have to write right away about the incredible screen quality on the new i-mode phones. Japanese usability professionals at the conference do have complaints about some aspects of i-mode, but I found the 503i phone easy enough to use that I navigated my way in a few seconds to the Alertbox I had posted from my hotel room earlier in the day. Even with a high-res color screen, I still don't think that a full-length column is the best content for mobile Internet, but the improved readability offered by the 503i at least makes it reasonably nice to browse and read medium-length information. The 503i also has a cute (and functional) joystick and a UI that starts to approximate a real GUI as opposed to the odd designs found on European and American cellphones. I particularly liked the way they use the (still small) screen space to the best advantage by dynamically expanding the currently selected area as you use the joystick to move around a 2-D menu. (And I haven't even talked about the next-generation devices showcased in the DoCoMo store next to the conference - candyland :-)
- February 9, 2001
- Having a link to the Amazon payment service on my home page for three days resulted in payments of $314. Considering that the home page had 22,095 page views during this period and that Amazon keeps $62 as their fee, I conclude that the value per page view was 1.1 cent. That's not so bad, considering that Yahoo's revenues are 0.4 cents per page view (computed based on their report from the fourth quarter of 2000 - however, Yahoo's revenues have been 0.4 cents per page view for two years, so I am starting to view this as a constant baseline for the earnings potential of an advertising-based site - note that 0.4 cents per page is a CPM of $4).
- February 6, 2001
- Disney's search engine Go.com is dead. In April 2000, I wrote that Go was stagnating as indicated by the referral traffic I was seeing come into useit.com. Your logfiles may not be the best source of investment advice, but take a look anyway. They do reflect true user behavior and can be a leading indicator (as my logs clearly were in this case).
- February 5, 2001
- New experimental micropayment system:
Amazon has launched a new payment system that relies on users voluntarily paying for content. I have my doubts whether this idea will work. I believe more in micropayment systems that automatically deduct pageview charges from the user's account. Anyway, here's how Amazon's system works:
There is a graphic here, that will have your name embedded if you have an Amazon cookie set on your computer:
Click the graphic to go to a page where you can pay what you think the content is worth (in my case, what you think the Alertbox columns are worth).
Kind of spooky to have your name displayed on a website where you are not registered. The reason it works is that the image file is being served from the amazon.com domain and thus has access to Amazon's cookie on your harddisk. The useit.com website has no access to this info, so as far as my server goes, all users are still anonymous.
I expect many users to get very confused about this mingling of features from different sites. On the other hand, I have predicted for some time that stand-alone websites would disintegrate as the basic unit of Web use. This payment system is another nail in the coffin for website structural integrity.
- February 1, 2001
- My report from the World Economic Forum in Davos, Switzerland.
- January 19, 2001
- David Coursey's argument why he might buy an Apple laptop essentially goes as follows: these days, the Mac is an MS Office machine. He may well be right, even though I also think there are hardware considerations in buying laptops, and it's hard to beat Sony on size (i.e., lack of same). Anyway, this supports my comment in April 2000 that the real platform these days is Office and that Microsoft's monopoly at this new level is of more concern than what they do with Windows. Maybe George W. Bush's new Justice Department leadership will take a fresh look at the anti-trust case and bring it into the 21st century instead of continuing to litigate a 1990s problem (I don't think so, but one can always hope).
- January 11, 2001
- Samsung announces a cellular telephone without a numeric keypad; just what I recommended in my last Alertbox.
- January 4, 2001
- Just before Christmas I wrote an Alertbox explaining why Web services would have to start charging their customers in 2001. As of January 10, Yahoo Auctions will start doing exactly that. These guys are smart. I continue to believe that Yahoo is the company with the best understanding of Web usability. (Disclosure: I caved in and bought YHOO stock a few days ago when it traded at 27. Full disclosure: I have a poor history of investing in Yahoo - I sold my last position when it traded at less than half of the all-time high; I thought Yahoo was over-valued at 100, after which it promptly climbed to 250.)
- January 2, 2001
- In an interview with Dr. Linda M. Bartoshuk, the New York Times explains why subjective rating scales often are misleading. I can't tell you how many times I have watched users struggle with a clearly useless and impossible design only to give it a "6" on a 1-7 scale.