In case you’ve never heard of it, Search Engine Optimization (SEO) is the practice of trying to raise your site’s placement in Internet Search Engines. Everyone wants to be on the first page of Google results, and preferably in the first five listings on that page. The increase in traffic that comes from better search rankings can be enormous and lucrative.
I might be called an “SEO skeptic.” That’s not because I think SEO has no benefit — In fact I think understanding SEO is essential to managing a Web site well. So how can I be a skeptic?
I’m an SEO skeptic because too often SEO is treated as a specialized practice and I believe focusing on it in isolation can:
- De-emphasize Web content as content created primarily for people, and
- Over-emphasize a small subset of Best Practices for Web Development
Unfortunately, no matter how well-intentioned SEO specialists may be, their work can’t be performed in the absence of information architecture, standards-aware development, and user-centric content practices. While targeted SEO work might lead to short-term gain, it carries genuine risk of creating long-term problems. This is why I advocate treating SEO as a benefit of having a well-managed Web site rather than a type of work to be done for its own sake. Spending money on an SEO specialist should be the last thing you do after you make sure your Web team is adequately trained and that training is reflected in their performance.
Nearly all SEO benefits can be had as byproducts of trained workers doing long-term quality content and development, but SEO work done without also doing long-term quality content and development can actually harm the health of a Web site.
The Blue Collar Rocket Science Breakdown:
The best SEO results come from managing it as 95% skilled blue collar work, and 5% rocket science.
Quick Detour: SEO in The Semantic Web
While SEO as we know it today will be a consideration for years to come, the way the Web is understood by search engines will be changing as a new set of ideas and standards often collectively called “the Semantic Web” takes hold. Initially, semantic information will only impact what information is shown in search results, but there may come a time when “Semantic Web Optimization” is more important than SEO.
I will take a deeper look at the state of the Semantic Web in my next blog post.
SEO 101
There are a number of components to SEO. Here are summaries of the main ones:
- Write your content so that a robot can easily extract keywords from the content.
- Make your content something that is worth linking to from other sites, encourage people to link to it from other sites, and actively find legitimate ways to link to it from other sites.
- Use HTML that has semantic meaning, instead of HTML that is only concerned with how things look.
- Use URLs that reinforce the keywords in the content.
Number two is far and away the biggest one. If you get linking right, you’ll rise in the rankings. Get it wrong, say by spamming blog comments with links to your site, and you’ll sink like a stone. Don’t be greedy with your links and try to only have inbound traffic: The Web is an ecosystem. You link out, others link in, and Google notices.
SEO and Me: A Wary Relationship
In late 1998 I was working on Disney’s GO.com portal initiative. I once participated in a conversation with one of the search engineers who worked on the Infoseek search engine that was at the portal’s core. We were asking him to share tips to get GO.com pages higher than competitors’ pages but even though we all worked for Disney he was, rightfully, not enthusiastic about sullying the purity of the search. All he was willing to divulge was the things people did that hurt their rankings, which at that time mostly had to do with abuse of <meta> elements (or “metatags” in the imprecise terminology of most SEO conversations).
But some things seemed to be helping. Even in those days of HTML that followed no standard, the use of semantic heading elements appeared to have some weight. That was my first moment of starting to understand something very important:
HTML is a language for describing scientific research papers.
Tim Berners-Lee didn’t come up with HTML because it would solve the problem of how to sell shoes over the Internet. He came up with it because there was no common language for describing documents or referencing their sources. He was sick of needing a bunch of different programs for reading papers, and he was sick of trying to track down the sources for the papers. HTML provided a standard way to mark up papers, and HTTP provided a way to link them to each other.
The problem was that once it was released, HTML spun off into a wild, uncontrolled thing with no standard definition. Those of us who were making Web sites in the mid-90′s hacked it into a layout language, ignoring semantic meanings, and generally messing everything up.
In 1998 search engines hadn’t yet become as important to the Web as they are, so SEO wasn’t a primary concern. As a Producer I was focused on content creation and user experience. I came up with best practices like this:
- Write your content so that a reader can easily extract keywords from the content.
- Make your content something that is worth linking to from other sites, encourage people to link to it from other sites, and actively find legitimate ways to link to it from other sites.
Yep, those are the same first two items as in the SEO components list, except in number one I changed “a robot” to “a reader.” This highlights something that I think many people, including many SEO consultants and their clients, overlook:
In the current pre-Semantic state of the Web, Search engines are robots that try to experience the Web the way human beings experience the Web.
If people can glance at your content and quickly understand what it’s about, then robots will too. If people think sites linking to your site are good, then robots will too. If people think sites you link to are good, then robots will too. If people see you have a lot of links into and out of your site, they will see you as an active, valuable part of the Web, and robots will too.
If you focus on creating content that is just for robots, you may succeed in raising your rankings. But you might alienate your human readers.
That conversation with the Infoseek search engineer, as well as a few less-than-fantastic experiences with Content Management Systems got me interested in how data is structured. I went into XML, which primed me for learning XSLT a couple years later during a contract sojourn at Microsoft, which led to something of a sliding left turn into Web development. But long before that, something much more important happened, although I didn’t know it until years later…
In 1997, the HTML 4 spec was published. By moving formatting out of the markup and into the CSS language it provided a clear solution to the mess that had been made of the Web. It also made it much, much easier to build, update, redesign, and maintain sites. But until browsers caught up, in about 2001, most sites still had to do things the wrong way. But when the early adopters of clean HTML 4 began trouncing the competition in Google search results the SEO consulting industry really took off. What their clients perhaps didn’t understand was that SEO wasn’t the magic, standards-compliant HTML 4 was.
Fortunately for SEO consultants, but unfortunately for their clients, it’s possible to get pretty good rankings by hacking a few semantic elements into a site built with old-style HTML. So it was easy to make recommendations that would deliver desired results, but ignored the deeper underlying problems. This could, and often did, turn what might have been a single, unified, push to modernize Web sites into a multi-year agony of diminishing returns and frustration.
It was with the turn of the century that I started actually caring about how HTML was structured. But still I come at it from a background in managing content creation and user experience. I’d internalized that semantic HTML makes Web design, development, and maintenance much easier and cheaper. You don’t have to be a developer to appreciate that. So when I took on management of the HTML production at REI, I had one primary Best Practice that I drilled into my reports:
- Use HTML that has semantic meaning, instead of HTML that is only concerned with how things look.
Where have I seen that before? Oh yeah, it’s the same, word-for-word, as point number three in my SEO components list. So that leaves us just number four from the SEO list:
- Use URLs that reinforce the keywords in the content.
Really that’s actually a part of user experience that’s as old as the Internet’s Domain Name System itself. No one wants to make their users go to the Web site address 207.46.197.32 when they could go to Microsoft.com instead. Some User Experience people had been arguing addresses built out of query strings like http://rei.com/search?query=Tents&button.x=0&button.y=0 instead of http://rei.com/search/Tents were bad experiences to be avoided whenever possible, but they didn’t get far with the backend developers until SEO gained momentum. Of course, the reason this is important to SEO is if it’s bad for people, it’s bad for robots.
When Bad Things Happen to Good Web Sites
SEO is an easy sell to management because it can return significant gains rather quickly, assuming there are gains to be made. (Remember, there are only ten spots on the first page of Google results and your competition has an SEO consultant too. Maybe even the same one you have.) But if you have issues with SEO, it’s pretty much a given that they stem from deeper issues you need to address.
SEO by the Gallon
Last summer, a local site development company with a national base of clients was looking for a manager of their SEO team. I was just coming off my work at REI, where I had achieved major SEO improvements in as an integral part of managing the implementation of the site redesign. I figured I was plenty qualified for the posted position, so I applied.
I got as far as an interview with the person who would have been my Director. About an hour after the interview ended, the recruiter I’d been working with called me back. In a tone of voice that clearly indicated she felt I’d been wasting their time, she said the Director told her I wasn’t a real candidate because I was “a developer.”
The turning point in the interview had come when I told the manager that I’d looked at the HTML from a few sites that they managed and could see SEO improvements to be made. He didn’t want to hear that. He wanted a special SEO sauce that came in a jug and could be poured over a site without getting on anything inside it. Because I showed I think of SEO as part of a holistic approach to content and development, I had disqualified myself.
If I thought it was one director at one company that thought that way, I wouldn’t mention it. But it’s a misperception I’ve run into a number of times.
Audit Yourself
If you are tempted to seek SEO expert advice, first make sure that the fundamentals of your site are in the best possible shape that your available resources can make them:
- Ask your Web Developers if their HTML will validate against the spec. If it won’t validate, there may be a perfectly good reason why it doesn’t, but if your developers can’t tell you the reason they may need some additional training.
- View the site with CSS turned off (an easy way to do this is with the Web Developer Extension for Firefox.) Ideally, you will see nothing but text, images, and tables, all lined up neatly in a column. Again, if you see something else your Web Developers may be able to give you a perfectly good reason why, but if they can’t some additional training may be in order.
- Review your content. Can you tell what a page is about at a glance? Does the title description in the frame of the browser describe the page? Do important keywords get used in multiple locations, without being so obtrusive as to interrupt the user experience? Do links to other pages or other use meaningful text, or do they say something generic like “read more…”? If you find problems and you’ve never had a dedicated Information Architect working on the site, you might need one now. IA’s provide the skeleton plans for sites that designers, content creators, and developers flesh out. It’s their job to think all this through before anything gets set in stone.
- What about the URLs? Do you see the anything in there that describes what the page is about? If not, you may want to work with your server administrators to change that. But beware, just changing one URL for another could cause your search ranking to plummet. You need “301 Redirects” in place first. If your server adminstrators don’t know what that means, they need additional training.
Last Thoughts
Those are the thoughts of an SEO skeptic. It’s important to understand what SEO is and what it isn’t. One thing it is, is a zero sum game. At some point, you and all your competitors are going to have put as much energy into getting on that first page of Google as you can. Maybe you’re already there. Only ten links can do it, and there’s no guarantee that you are going to own any of them.
What is your strategy for succeeding without a focus on SEO? Is the rest of your site up to snuff? Is your Web team trained well enough to create competitive product? If you focus on that and let SEO happen, you’ll be spending your money wisely.
Post a Comment