Musical Algorithms

What if there were no Google algorithm to game? Or more precisely, what if there were several different algorithms all being used at the same time - so that no specific tactics or tricks would consistently place your pages 'at the top'? A discussion going on over at the SEO Chat Forums says this is just what might be happening over in the datacenters of the almighty and now almighty-cash-rich Google. Of course, as with all things Google, nobody is sure exactly what is going on, how long it will last, or why. But skipping by that lunacy for a moment, I want to weigh in with the opinion that the 'spinning-wheel-of-algorithms' is a pretty good solution to the SEO problem that Google themselves have created. Certainly a better solution than any other I've thought of or heard discussed. The 'SEO problem' of which I speak (in case you don't obsess about Google for a living), is that ranking highly in Google is worth thousands to hundreds-of-thousands of dollars per search term, and so webmasters and their paid advisors everywhere are constantly, actively, and relentlessly trying to game the system to get their pages up top. The result is that Google's results have been seriously altered (in many cases for the worse) and many web pages are being designed with Google in mind at least as much, if not more, than the humans who will actually use those pages. Millions of words in SEO forums have debated the ramifications of all this, with opinions ranging from 'business is war' to 'just be nice'. Google themselves have mostly stayed silent (excepting the usually cryptic quips of the Google Guy) in an apparent attempt to avoid flaming the fires of an unwinable debate with a predominantly unreasonable crowd. Here's why I like the russian roulette solution so much: # The fact is there is no right algorithm. By mixing them up Google will give searchers a broader range of results and over time I'd suspect that leads to a better online experience. # What gets measured gets done. This is why we have comment spam and cloaking and all kinds of SEO techniques that are currently effective but ultimately silly. If certain types of links start to count less often, there is that much less incentive to be aggressive in getting links. If bold tags count a little less, there is less incentive to use them when a design otherwise wouldn't call for them. # A broader set of measurements further incents genuine goodness. I've never believed the 'black hat' vs 'white hat' SEO arguments - so I'm not referring here to 'kinder and gentler' SEO techniques. I've also never taken much stock in the 'just make a good site for your readers' crowd, because that is really quite nieve. But I do believe the algorithms are constructed to generally reward efforts that make sites better for users. The problem has been that this general effort assumes that nobody is intentionally gaming the system, and that of course is an even more nieve assumption. But if sites are measured on lots of different metrics all mixed into a bunch of different algorithms, then over time perhaps the gaming of the system will be limited simply because effort/reward ratio will increase substantially. And that is ultimately why I have such a positive reaction to the random-algorithm-generator. It ends the game by smashing it into a million pieces. More competition from Yahoo and even MSN and the other search engines could also do this - but those are going to take a lot of time before we see major percentages of the search population shift. By Google itself becoming multi-algorithmic we could see a huge change in the SEO game overnight. Surely there would be challenges and ramifications for both site owners and marketers, and probably plenty for everyone to both like and dislike. But I think it would essentially end the arms race and that alone makes it worth doing. {Via Search Engine Roundtable} Posted by Craig Danuloff at August 15, 2004 6:25 PM