The Evolution of Corporate Web Sites
Published on April 28, 2004
Hard to believe, but corporate Web sites have been around for over 10 years now. It’s fascinating to see how they have evolved over the years, from the early days of magazine-style brochureware to the most recent trends of two-way Web interfaces.
Corporate Web sites tend to re-invent themselves every 1-2 years and one reason is because Web design is constantly changing and adapting to technology in an evolutionary manner. Web design is a product of its environment, in particular of HTML and Browsers, but also secondary influences like Content Management Systems and the speed of computer processors. Web design is always pushing the boundaries of what current technology allows.
The Early Years: 1993-96
Let’s start at the beginning. Mosaic was released in November of 1993 and it was the first graphical browser to become popular on the World Wide Web. At that time HTML 2.0 had not been released, so Mosaic was initially limited in its functionality. Netscape Navigator 1.0 arrived on the scene in December 1994, with support for basic HTML 2.0 elements only. By the time Internet Explorer 2.0 was released in November 1995, browser technology had begun to support more advanced features such as tables and forms. It was now possible to evolve beyond the single linear page of text that passed for Web design during 1994-95.
At this point the Web was still mostly the domain of spotty scientists, who were more interested in publishing the Periodic Table of the Elements than arranging the visual layout of HTML elements. That all changed in about 1996, when Graphic Designers discovered the new medium. In the eighties and early nineties, Desktop publishing (DTP) on Apple Macintoshes was all the rage with graphic designers. The Web was their next challenge.
Netscape vs. Microsoft
Enter the Marketeers
Around this time Marketing departments started to get interested in the commercial potential of the Web. The Pepsi and Coca-cola Web sites were two interesting examples. Pepsi’s Web site in 1996 was a tour de force of Web technology: Java-animated spinning logo, Shockwave, 3-D VRML, streaming audio, QuickTime. The premier version of the Web site, called “Psychotropic Sundae”, required numerous plug-ins to be downloaded in order for it to work.
Pepsi was one of the first corporate Web sites to fully exploit the Web’s interactivity. It tried to create a virtual world—populated with the “hottest” music, movie previews, digital art, game previews and so on. It even went so far as to nickname its users as “squatters”—people who visited and inhabited Pepsi’s new world.
Pepsi’s goal was to become a mass-market content provider, in order to attract as big an audience to its spot on the Web as possible. It was a broadcasting mentality, but they also recognized that the Web was an interactive medium. Pepsi.com provided a “destination” for consumers to “play” and “hang out”. The Web site was essentially a Pepsi-branded immersive world.
Coca-Cola’s Web site in 1996 was rather tame by comparison—in fact it was styled as a “virtual museum”! But it, too, presented its Web site as a “world” of games and entertainment, a “place to be”.
This sense of place was also an important focus of Amazon.com, which initially billed itself as a virtual bookstore. Amazon had its IPO (initial public offering) in May 1997 and it was one of many signals during that year that e-commerce was on the rise. The Web was beginning to be seen as a channel for sales and marketing, as well as a medium to deliver content.
By 1997 Microsoft had finally caught up with and indeed overtaken Netscape in browser technology. While Netscape was the first to release its 4.0 browser, in June 1997, Microsoft soon trumped it with the release of IE 4.0 in October that year. IE 4.0 was widely seen as being a superior product—for example, IE’s stylesheet support was better than Navigator’s.
1997 was the year the second edition of David Siegel’s book Creating Killer Web Sites came out. It was one of the first Web design books I read. It’s interesting now to look back at this book seven years later. Siegel was famous for encouraging and widely promoting HTML “workarounds” (a nicer term for hacks), in order to obtain the most visually-appealing Web page layouts. The reason he had to do this was because HTML is based on structural rather than presentational principles. So in order to optimize HTML for presentation Siegel had to find ingenious ways to manipulate HTML markup.
One of the main weapons of choice for Siegel and his contemporaries was tables. HTML 3.2 was the first W3C-recommended specification to include tables. It stated that tables “can be used to markup tabular material or for layout purposes.” In a semantic/structural sense tables are most suitable for tabular data such as the Periodic Table of the Elements that the scientists from 1994 wanted to publish. However, using tables for layout of unstructured information soon became the norm. A couple of common examples were adding transparent spacer GIFs to control positioning of content, and inserting chunks of narrative text into tables. It led to messy code with little semantic value—there were tables within tables, then tables within tables within tables, and so on.
Another Web design trend that became de rigueur around 1997 was the notion that Web sites ought to be “sticky”. Web sites should entice people in, like a Venus Flytrap, and keep them entertained long enough to sell them something. This excerpt from Siegel’s book sums it up:
“You must welcome them and make them feel at home in your site. Most third-generation sites have an entry, a center area with a core page for exploration, and a well-defined exit. Third-generation sites pull visitors through by tantalizing them with something exciting on every page.”
This is another example of Web sites aiming to be ‘places.’ Siegel encouraged designers to make their Web sites like a “home.” Or if we take a more cynical view, Web site visitors were treated like teenage boys being enticed into a House of Ill Repute by ladies with long black boots and red fishnet stockings. To extend this dubious analogy, a lot of Web sites had the equivalent of a flashing pink neon sign with the words “Girls Girls Girls” emblazoned on it—the Splash screen!
A Unique Medium
Metaphors were also actively encouraged in this era. In 1997 the Web was still a new phenomenon to most of the population; one of the easiest ways to make the Web seem more familiar, and less alien, was to make the Web look and feel as much like the real world as possible. So we were treated to a variety of shopping mall metaphors, city metaphors, home and room metaphors. It took a few years more for Web designers to realize that the Web wasn’t a mere copy of the real world, that it was a unique medium with its own characteristics.
In the early years of Web design, Web publishing was utilized mainly as a broadcasting medium. This is obvious in the terminology used to describe Web readers at that time—they were most often described as “viewers” or an “audience”, rather than active “users” as we think of them today. This culminated in 1997 with the brief rise and fall of “Push” technology and “multi-casting”, both of which aimed to deliver content and advertising directly to people’s PCs like a television broadcast.
HTML 4.0 and CSS Specs
In July 1997 the draft for HTML 4.0 was released and in April 1998 it became the official W3C version of HTML. The previous versions of HTML had mostly kept pace with the Web browsers, but by the time HTML 4.0 hit the scene, Netscape and Microsoft had forked off into different paths and were incompatible in a lot of ways. This meant that HTML 4.0 inevitably was not supported fully in both major browsers and this muddied the waters terribly for the humble Web designer. The phrase “Optimized for” became commonplace on Web site homepages during this time and designers often had to create two versions of their Web sites—one for Navigator and the other for Explorer.
The browser war also hampered the take-up of CSS2, which was introduced in 1998. It wasn’t until IE 5.5 was released, in mid-2000, that browsers began to support CSS2 functionality and designers began using it regularly to separate presentation from structure.
Transactional Web sites
The e-commerce bubble was expanding like a big balloon during 1998 and 1999 and it led to Web design becoming more and more transactional, rather than simply interactive. Database design became a foundation of most Web sites, using dynamic Internet languages such as ASP and PHP. Transactional Web sites were more complex than their predecessors—requiring real-time computation and personalization of data.
E-commerce portals also became a hot Web design trend in this period. Many a business rushed to make its Web site a portal in the hope of selling not only its own core product or service, but gumboots too. The idea was that once a business had a transactional Web site up and running, they could “leverage” it to sell a variety of products and services.
Unfortunately a lot of e-commerce ventures failed to deliver on their promises and in 2000 the balloon burst. A contributing factor was that the landscape—browser and PC technology in particular—wasn’t mature enough to cope with the complexity of transactional Web sites.
Standards and Standardizing
2000 was the year that saw the beginning of a Web design revolution: Web Standards. XHTML 1.0 became a W3C Recommendation in January 2000 and from that point onward XML was on everyone’s lips.
It is noticeable from about 2000 that corporate Web sites became more and more predictable or standardized. The company I worked for at the time had a motto that summed up this trend: “one face to the world”. It was a multi-national company and during the mid-to-late 90’s it had spawned literally hundreds of different Web sites throughout its worldwide empire – most of them with different designs. But in 2000, it was announced that all Web sites of the company must conform to a single, standard design from head office. And it turned out to be a carbon copy of the design of a major multi-national IT company, whom they’d hired as consultants.
It’s not a coincidence that standardization of corporate Web designs coincided with the rise of Web Standards. Browsers had become a jumble of incompatible features and the chaos it caused for Web designers finally forced people like Jeffrey Zeldman to champion a common approach for HTML and CSS coding called “Web Standards”. There was also a drive for consistency in Web site interfaces and navigation, labelled “usability” by Jakob Nielsen and others.
Content Management systems were also a big part of the corporate Web by 2000 and contributed to the sameness of Web designs. CMS’s rely on design “templates” to produce Web site content, and the use of templates invariably leads to designs being re-used and shared.
One Browser to Rule Them All
Another contributing factor to the standardized Web designs of this era was the increasing dominance of one browser vendor. By 2000 Microsoft had achieved an 86% share of the Web browser market, according to WebSideStory. IE 5.5 was released in July 2000 and it proved to be a solid performer, especially in comparison to the buggy Navigator 6.0 released later in 2000. IE 6.0 was released in August 2001 and it only extended Microsoft’s lead over Netscape. At around the same time Netscape transformed its flailing flagship product into an open source browser called Mozilla, holding out some promise for a competitive browser market in the future.
Extending The Web
During 2002-03 the definition of Web began to be stretched to include not only personal computers but also mobile phones, televisions, and various other household appliances. The catchphrase for this is “Ubiquitous Web” and it relies heavily on XML technologies.
In parallel to this extension of the Web’s boundaries is a trend towards Web sites as a collection of “Web services.” This is known as “Service-Oriented Architecture.” Web sites circa 2004 are no longer thought of as “places” or virtual versions of real-world spaces. Web sites are now a collection of “loosely-coupled” services.
Whenever I visit Amazon now, I no longer see it as a virtual bookstore or even a virtual mart (seeing as it sells everything under the sun now, including electronics and clothing). Rather, I see Amazon as a collection of services tailored specifically for me. It knows my name, it recommends stuff to me, it can show me what’s new since I last visited, it offers me ways to make money by partnering with Amazon, it lets me rate and review items, and much more. Plus I can utilize aspects of Amazon’s Web site via a third party service, such as Allconsuming.net. Or I can build my own service, using Amazon’s APIs (Application Programming Interfaces).
So it’s much more than a place I can go to shop—it’s like an extension of my own person. The Web site is personalized to me, so that it feels less like a place and more like a virtual agent ready to do my bidding. It’s a far cry from the brochureware Web sites of 1994!
The Future: Continuing to Evolve
Over the past decade Web design has gone through many iterations, driven by the ever-changing environment. Web browser vendors have contributed a lot of new features and functionality. The HTML specification has grown from a rigid structurally-based markup language, to an extensible HTML-XML hybrid. And CSS is now widely used to keep structure separate from presentation.
Plus let’s not forget the business drivers for Web sites. Marketing and businesspeople have slowly gained an appreciation of the Web’s unique strengths—for example that it enables a two-way dialogue with customers.
The changing landscape has led corporate Web sites to evolve from textual to multimedia, brochureware to interactive, static to transactional, chaotic to standardized, rigid to extensible, broadcasting to read-write. Web sites are no longer virtual places, they’re more like virtual agents. Today, corporate Web sites exist to serve their users and so their design must be personalized and loosely-coupled.
Web sites will continue to evolve and be products of their environment. Browser and operating system innovation (or lack of) will affect what the Web looks like in another 10 years. XML Web technologies that so far haven’t impinged much on corporate Web sites, like RSS and RDF, will force new ways of designing onto us.
We don’t know what corporate Web sites will look like in 2013, but we do know that Web design will continue to re-invent itself constantly like it did from 1994-2003.