useit.com ![]() ![]() |
Jakob Nielsen's Alertbox for March 8, 1998:
Similarly, it is an utter waste to look at jerky and pixelating postage size stamp Internet video that purportedly shows some smart person delivering an important keynote. Not only is the video useless; because it is so bad, it actually distracts from the audio message. Spend the available bandwidth on high-fidelity audio supplemented with slowly changing high-resolution photos of the speaker, the audience, and (most important) the visuals used in the talk. Even better: deliver something that can't be done in the physical world by indexing the talk and allowing the user to jump directly to segments of interest while reading short abstracts of the other parts. Then integrate the speaker's original content with hypertext links to background information and annotations and comments by other authorities in the field.
Instead of impoverished facsimiles of reality, design from a basis of strength and go beyond reality to things that are impossible in the physical world. It is painful to use the Web, so we need to reward users: give them something new and better that they didn't get before.
Internet malls make no sense, but in the early years of the Web, several companies built sites that approximated real-world shopping malls. In the physical world, malls help by offering one-stop parking and allowing visits to several shops without having to walk outdoors through rain, snow, or glaring sun. On the Web, any shop in the world is a single click away, so there is not much benefit from aggregating different shops on a single site.
Instead of malls, Web shopping is enriched by linking arrangements that allow the smallest website to have the world's largest inventory and allows cross-selling of related products at other sites. A good example is the "associates programs" pioneered by Amazon.com where any site with a targeted topic area suddenly turns into a full-service bookstore selling books related to the site's focus without the need to store any books. A butcher can sell cookbooks and a florist can sell books on flower arrangements. The key insight is that things do not need to be on the same website to be integrated: the hypertext link allows new forms of customer service that are impossible when people have to move in meatspace.
Amazon.com has a robot that interviews authors: in reality, this means that authors get asked a standard set of questions and end up editing the interviews themselves. The downside is that the questions are not specific to the individual author's book and interests. The huge upside is that it becomes feasible to interview thousands of authors. In contrast, B&N has a very small number of author interviews. Doing interviews in a way that wouldn't work in the physical world has allowed Amazon to feature a much richer service.
Furthermore, Amazon links its page for each book to its interview with the author, meaning that users don't have to specifically search for author interviews to find them. In contrast, B&N keeps the interviews in a separate part of their site, so if you find a book you will never discover whether there is an author interview available. This difference in linking shows the benefits of thinking beyond reality. In a physical bookstore, it would be too much work to add a sticker to all the books with a reference to an interview that was kept at the sales counter. But on a computer, the link is simply one more field in the database: as soon as the database is updated, all future users will see a link to the author interview whenever they are interested in a book.
Most companies understand Web business very poorly.
Everybody involved in Web design or Web strategy should do as much of their
personal business on the Web as possible to get direct experience with the
"Web lifestyle." Following this philosophy, I recently bought a present for
my nephew from the website of a famous toy store. After completing the
order, I received a confirmation email that listed the name of the toy and
its price. So far, so good. The email then went on to say, "if we have
the toy in stock, you will receive it in about one week; if it is not in
stock, you will receive it in three to four weeks." Have these
guys never heard of integrating order processing with inventory
management? Don't send an IF-THEN-ELSE
statement to
your customers: that's what computers are good for.
In the physical world, sales clerks may have to tell customers that they don't know when an order will ship. But the Web should be better than reality: check inventory and shipping schedules in the background and hold the email confirmation until you know the ship date. If you can't find out within a few hours, then send a preliminary confirmation followed by more precise information the minute you know. Again we can go beyond the real world: sending an extra message when a condition is triggered is virtually free on the Internet. Also, since the message doesn't go by snail mail, it will reach the customer in time to do some good.