• Register
• Log in

SEO Theory - SEO Theory and Analysis Blog

Algorithm analysis, Web community relationship analysis, SEO practices and techniques, industry news, etc.

• Home
• About SEO Theory
• SEO Glossary

How search ruins the Web experience

Posted by Michael Martinez on May 14, 2008 in SEO Theory

Google has been proselytizing its “rel=’nofollow’” strategy for about three years now. Although the road has been rough Googlers probably feel like they have achieved some success. There are now quite a few Web application vendors and services that automatically embed “rel=’nofollow’” in user-generated links. Despite occasional rumors or allegations that Google has strong-armed people into changing their systems to comply with “rel=’nofollow’” I get the impression that they have been delivering the message more conventionally (such as Matt Cutts’ keynote address at Web 2.0 Expo 2008).

You draw more flies with honey than with vinegar, and all that.

Google’s message is fairly consistent: they want to provide a good search experience for their users, they want to drive traffic to “worthy” sites (I feel Matt means sites that offer something of value more than sites that Googlers may personally want to champion), and they want to help Web site operators de-incentivize the search spam community.

However, the Google problem is not so easily solved. These kinds of strategic initiatives have stripped large portions of the Web of their natural connectivity. I don’t mean that “rel=’nofollow’” blocks connectivity because obvious people can still click on the links. Rather, people have become more reluctant to link to each other because of Google’s complaints, warnings, and admonitions. It seems like today you’ll more often than not see Googlers telling people to go ahead and link to sites they trust without fear rather than telling them not to link out.

That is, Google isn’t just peeing in the search spammers’ well, it’s peeing in its own well. Not only are Googlers now having to spin their spin, they are slowly fabricating an artificially linked Web that looks nothing like the PageRank-mapped Web they want to document.

Google’s dedication to PageRank has unquestionably forced it to spend millions of dollars on “solutions” to find and filter out link spam. PageRank did not make Google the success that it is. Google’s complex relevance algorithm is what has ensured that searchers would find something worth clicking on. PageRank is not about relevance, it’s about value and unfortunately the Web was not created for value-based relationships. Hence, a search technology based on value works poorly with the Web.

But Google is not alone in its struggle to make PageRank work as promised. Many other people have enlisted in the cause either without fulling understanding the challenges Google faces or in spite of those challenges. Google has done some amazing things through the years and a lot of people have placed immense faith in Google’s technology. That communal respect should not be dismissed or undervalued. There are many SEOs who believe in PageRank, who want it to work, and who want Google to succeed in all its search initiatives.

Nonetheless, the SEO industry needlessly pursued PageRank for years, driving Google to refine its filtering technologies to the point where Google had to implement Web apartheid, a patently unfair system of segregating Web sites into those sites that are given favorable promotion in search results and those sites that are virtually ignored in queries to which they are most relevant. Google and the SEO industry had an opportunity to create a symbiotic relationship that could have benefitted everyone, and I think over the past couple of years some steps have been taken to build that relationship on both sides.

But Google has, in fact, been forced to throttle itself in its neverending battle to throttle search spam. That is, had Google not had to fight all the needless link spam the SEO community created through the years, Google COULD have devoted more resources to actually improving its search results, reaching out to the searchers, and providing better, more relevant listings. Not that Google ever stopped seeking to improve its search results, but it would be a great day for Google, search, the Web, and everyone in general if Matt Cutts had to find a new career path.

It’s good for any Web site to be at the top of the search results but most Web sites are not at the top of the search results. If Google returns on average 200 listings for every query, and if searchers run 2 billion queries per month, an average of 398 billion listings do not make it to the top of the results every month. There is immense pressure being placed on Google to drive all sites to the top.

Search engines document and index the data that we package and present on the Web. In some cases, they even repackage our packaged information and present it to searchers, de-incentivizing searchers to actually look at the source sites (think of all the people who may use Google to look up the meaning of ’seo’. Why do they need to look at the source sites? (Most people would not know that the majority of the definitions Google cites are wrong, incomplete, poorly written, or irrelevant — although “Seo” is indeed a Korean personal name but did Googlers intend for us to use the define query operator to look up personal names?)

Google is transforming itself into an encyclopedia, one I admit I use on a daily basis because — like millions of my fellow searchers — I just don’t have time to read all those Web sites. I hate seeing Wikipedia show up in so many queries because I know the information cannot be trusted. But the truth is that my bones don’t yet realize that Google’s information cannot be trusted. I still use the define query operator a great deal, and I occasionally force myself to click on some of the source listings to see if Google really got the right thing.

Most searchers won’t do that very often, no one will do it all the time. Hence, we’re using Google as the Supreme Web Interface — we filter all our information through Google and, frankly, Google is just not good enough at condensing human expression into a single format to be doing this kind of work.

I’m not sure Googlers are really wanting us to use their search engine that way, but Google puts AJAX and Web 2.0 to shame because you don’t have to wait for Javascript to go through all its nonsense when you search for content on Google.

Microsoft and Yahoo! could learn a thing or two from the Google search experience. It’s not about the listings. Google has demonstrated that relevance doesn’t have to win out. It’s about the presentation. Any idiot can get 1 million page views by creating a site that is interesting or informative. It doesn’t have to be accurate or correct. Google doesn’t have to be any more accurate or correct than Wikipedia.

Which goes back to the Wikipedia Principle: “search results tend to favor minimally acceptable content because of the prohibitive cost of identifying highly acceptable content.” In other words, “it’s too expensive to be right in an environment where being satisfying is sufficient.”

If the content search engines serve is only acceptable, it does not have to be comprehensive, completely correct, or all-inclusive. And that means that as search engines improve their ability to deliver acceptable information to their users, the incentive for people to click through on search results declines.

Not that we need to hang up our Web sites just yet. Rather, we need (as search optimizers) to refine our skills at being compelling sales people. We have to hone our search listing pitches so that people continue to click through to find the deeper content. But we also have to be cognizant of the fact that search results pages themselves are now acceptable content in many queries, and some people will never look past those search results pages to find what they are looking for because they will feel they found something good enough.

Search is becoming the Web and the Web is becoming a database.