One of the oddest things about all the updates that Google has made over the last few years concerns the lack of impact that some of the changes have had on some businesses I know. Google Panda and other ranking algorithm changes have caused significant drops in traffic for many of these companies. But that’s not the odd thing. What’s strange, at least at first glance, is that a number of these companies have suffered only small drops in sales. This goes against everything we’ve ever thought about in search marketing–I mean, the reason that you work so hard in search is to get traffic to your site, because the more traffic you get, the more you sell–right? Google says wrong–at least partially wrong–and they have a good reason why.
It all goes back to the old question that has always bedeviled lead-based businesses: Why are we getting more leads but not closing more sales? The answer is what in the old days we called qualification. A qualified lead is one that is likely to buy from you. Often when sales leads come in higher and actual sales stay flat, it is because the quality of those leads are poor. Anyone can send you a name, but if that name is more or less picked at random, it doesn’t help you sell anything.
Google, over the years, has been figuring out the same thing. So, it is possible that the ranking algorithm changes that Google has been making are not merely directed at lowering the rankings for content farms and other spam sites, as many have opined, but rather improving search results for searchers across a wide variety of searches. The search leader might be working on a new game.
We’ve known for a while that calling the formula a ranking algorithm–singular–is no longer true. In fact there are many ranking algorithms and Google selects the right one with what it believes are the best mix of factors for each search. Some searches emphasize the freshness of the information, others might include more personalization, yet others might emphasize information over commerce–or vice versa.
Google might be judging some of these factors on clickthrough of search results and on conversions. Google has many ways of tracking conversions–everything from Google Analytics to Google Checkout to the Google toolbar–and we know that they are using machine learning to extrapolate from small sets of data to large ones. So, what that means is that once they have a subset of sites that they can get this information from (and they do), they can then look at every other site on the Web and identify which ones are similar to their master set.
If your site looks very similar to sites that convert highly for a type of search, expect your rankings to improve. If you look like a site that doesn’t convert, expect the ranking to go down–but maybe you don’t care, because that search wasn’t bringing you qualified leads anyway.
Stop focusing on mere rankings as a way to judge your search effectiveness. Besides the fact that rankings are personalized (so you don’t know for how many searchers you rank anywhere anymore), you are missing the point of search–just like those companies gathering more and more sales leads that don’t close. Look at your revenue and focus on your best-converting keywords. That’s where the focus needs to be.
