Manage updates with the Download App
E-mail to a friend
Send us feedback
Print this page
Is search ruining the Web?
It's easy to overinflate the importance of online search. Sometimes I can't help but wonder how it's even remotely possible that Google's stock is trading at more than $225 (at press time). But then I think of every new small business trying to make it on a shoestring marketing budget--actually every Web-based business, big and small, including CNET--and I realize that they're absolutely dead in the water unless they can somehow show up nice and high in search results. If Google tweaks its algorithms just a little bit, thousands of Web sites either have a very good or a very bad day. Search is the big dog; and it, more than standards, usability, or even aesthetics, drives the evolution of Web site design.
The cottage industry that's sprung up around improving a site's search results is called search engine optimization. At its best, SEO is a discipline that influences Web builders and designers to maximize their search engine results with some simple and uncontroversial changes. At its worst, though, the term includes a collection of questionable business practices, shady companies that promise clicks for cash and only sometimes deliver, and a tool that allows the proliferation of advertising-filled Web sites (free registration required) that do nothing but show up in search results and provide no information in exchange for ad impressions. It's also creating quite a heated debate about standards-based design and usability vs. search methodology. And as much as I'm a fan of standards and efficiency, I think the standards and usability are going to suffer the most.
The easy stuff
In its most basic form, optimizing your site for search engines involves having a lot of words on your Web site that a search engine can pick up. In the old days, site designers could artificially increase their results by embedding hidden text keywords in their HTML. You'd search for something common, like "Mickey Mouse" or "pizza" and, most often, end up inadvertently clicking through to a porn site that swamped you with pop-ups, crashed your browser, and got you fired from your job. Modern search sites changed all that--and in fact, Google will actually penalize sites that use hidden text by lowering their PageRank. So, hidden text is a no-no at most aboveboard sites, but they're still looking for ways to improve their results.
Often, in that technique, you just list every single possible word you can on your home page. The old usability standards, such as keeping text above the fold, are out the window in that case, and sites risk becoming busier and busier as designers cram every possible search-result term into every available cranny. You'll see every possible tool listed at the major portal sites, and CNET's certainly no exception. The result is a lot of helpful navigation but a potentially overwhelming page.
But that's the simple stuff. There's a lot more to search engine optimization. For example, some search engines give weight to the text in the browser's title bar. This is the text that goes in the field when you're building a Web page. If you want to follow the guidelines of good usability, this field should display an ontological view of your progress through the Web page. So, on a review of the Siemens S66 at CNET.com, located in the Cell Phones category of CNET Reviews, the title bar should read: CNET.com | CNET Reviews | Cell Phones | Siemens S66 review. But if we do it that way, we risk having our review slip down on the results page, because the phone's name is only the fourth item in our string. So, our title bar says: Siemens S66 review - Cell Phones - CNET Reviews, and our review is the second result on Google. And the guidelines take a little bit of a hit.
This is only the most simplistic example of the brewing fight. Things get a lot more complicated when you start to talk about wider-reaching standards such as CSS.
The harder stuff
In a nutshell, search engines don't really read standardized markup such as CSS. And this is a startling discovery. It's easy to assume, from the outside, that Google would have a natural, abiding, and intense interest in CSS and would adjust its algorithms accordingly. After all, some SEO experts suggest that one way to trick search sites is to put whatever keywords or content you want in an H1, and then use CSS to tell the browser not to display that text, or to display it in a different size. So, if Google doesn't read the CSS, doesn't it risk inaccurate results? At the moment, at least, it's considered safer (if not completely safe) to use CSS to cheat search engines than to use simple HTML.
Meanwhile, though, the CSS camp wants search engines to start reading CSS so that standards-based design will be more search-engine friendly, and it won't become the hidden text of the new search order. And that's a tough argument, since standards have an unfortunate history of bending to meet the big dogs of the Web world (think about how coding for Internet Explorer often means broken pages in Firefox or Netscape). Indeed, advocates in the CSS community complain that the mentality around search engine optimization is, "it works in Google." Basically, there's a natural friction between the way people think Web sites should be designed, and the way they almost have to be designed in order to make it in the race for page rankings. And in such a battle, design for search engines are destined to win, because it's just bad business to ignore a dog that size.
By Molly Wood, senior editor, CNET.com
Thursday, May 5, 2005 It's easy to overinflate the importance of online search. Sometimes I can't help but wonder how it's even remotely possible that Google's stock is trading at more than $225 (at press time). But then I think of every new small business trying to make it on a shoestring marketing budget--actually every Web-based business, big and small, including CNET--and I realize that they're absolutely dead in the water unless they can somehow show up nice and high in search results. If Google tweaks its algorithms just a little bit, thousands of Web sites either have a very good or a very bad day. Search is the big dog; and it, more than standards, usability, or even aesthetics, drives the evolution of Web site design.
What do you think? Would you rather have standards or search results?
Post your comment hereThe cottage industry that's sprung up around improving a site's search results is called search engine optimization. At its best, SEO is a discipline that influences Web builders and designers to maximize their search engine results with some simple and uncontroversial changes. At its worst, though, the term includes a collection of questionable business practices, shady companies that promise clicks for cash and only sometimes deliver, and a tool that allows the proliferation of advertising-filled Web sites (free registration required) that do nothing but show up in search results and provide no information in exchange for ad impressions. It's also creating quite a heated debate about standards-based design and usability vs. search methodology. And as much as I'm a fan of standards and efficiency, I think the standards and usability are going to suffer the most.
The easy stuff
In its most basic form, optimizing your site for search engines involves having a lot of words on your Web site that a search engine can pick up. In the old days, site designers could artificially increase their results by embedding hidden text keywords in their HTML. You'd search for something common, like "Mickey Mouse" or "pizza" and, most often, end up inadvertently clicking through to a porn site that swamped you with pop-ups, crashed your browser, and got you fired from your job. Modern search sites changed all that--and in fact, Google will actually penalize sites that use hidden text by lowering their PageRank. So, hidden text is a no-no at most aboveboard sites, but they're still looking for ways to improve their results.
Often, in that technique, you just list every single possible word you can on your home page. The old usability standards, such as keeping text above the fold, are out the window in that case, and sites risk becoming busier and busier as designers cram every possible search-result term into every available cranny. You'll see every possible tool listed at the major portal sites, and CNET's certainly no exception. The result is a lot of helpful navigation but a potentially overwhelming page.
But that's the simple stuff. There's a lot more to search engine optimization. For example, some search engines give weight to the text in the browser's title bar. This is the text that goes in the field when you're building a Web page. If you want to follow the guidelines of good usability, this field should display an ontological view of your progress through the Web page. So, on a review of the Siemens S66 at CNET.com, located in the Cell Phones category of CNET Reviews, the title bar should read: CNET.com | CNET Reviews | Cell Phones | Siemens S66 review. But if we do it that way, we risk having our review slip down on the results page, because the phone's name is only the fourth item in our string. So, our title bar says: Siemens S66 review - Cell Phones - CNET Reviews, and our review is the second result on Google. And the guidelines take a little bit of a hit.
This is only the most simplistic example of the brewing fight. Things get a lot more complicated when you start to talk about wider-reaching standards such as CSS.
The harder stuff
In a nutshell, search engines don't really read standardized markup such as CSS. And this is a startling discovery. It's easy to assume, from the outside, that Google would have a natural, abiding, and intense interest in CSS and would adjust its algorithms accordingly. After all, some SEO experts suggest that one way to trick search sites is to put whatever keywords or content you want in an H1, and then use CSS to tell the browser not to display that text, or to display it in a different size. So, if Google doesn't read the CSS, doesn't it risk inaccurate results? At the moment, at least, it's considered safer (if not completely safe) to use CSS to cheat search engines than to use simple HTML.
Meanwhile, though, the CSS camp wants search engines to start reading CSS so that standards-based design will be more search-engine friendly, and it won't become the hidden text of the new search order. And that's a tough argument, since standards have an unfortunate history of bending to meet the big dogs of the Web world (think about how coding for Internet Explorer often means broken pages in Firefox or Netscape). Indeed, advocates in the CSS community complain that the mentality around search engine optimization is, "it works in Google." Basically, there's a natural friction between the way people think Web sites should be designed, and the way they almost have to be designed in order to make it in the race for page rankings. And in such a battle, design for search engines are destined to win, because it's just bad business to ignore a dog that size.
