Search Engine Optimization July 29th, 2013
On one hand, SEO is not an exact science, but this is only the case because we don’t know the exact parameters of the Google algorithm. If we knew each and every parameter and the weight it carries in determining page position, then ranking higher would not only be science, it would be as simple as connecting the dots. It is important to remember that even if we knew with utmost certainty the way Page Rank is determined, getting our web property to the top of the page might still be impossible due to the high competition for a particular keyword. In other words, we can know that having a TV commercial play during the Super Bowl would help crush the competition, but we might not have the millions of dollars it will cost to pay for the spot.
If SEO is almost a science, we should be able to run experiments to explore and discover how the Google algorithm works. In science there are a few ways to increase your understanding of something. You can do experimental manipulations and record the results of the experiment. You can also record and describe lots of details and features across a large population and then correlate those features to results you see. In some situations, this kind of correlative study is all that is possible. This exact type of study was published in June, measuring 30 different website parameters of top ranking sites. While the results of the study don’t imply causation, they do give some startling details about things that were important for SEO a couple of years ago that now have little to no impact on ranking.
In the past year there has been a resurgence major brands on the search engine results page (SERP). This is an attempt to represent reality on the SERP. This leaves a big question as to how Google recognizes a website as a “brand”. There is certainly no master list of brands that get special ranking instructions. The “brand signals” must arise from the typical signals surrounding all websites. However, Google has identified a particular profile that identifies a website as a brand.
Perhaps the most indications from this correlative study of website factors and page position, was the rising important of rel=nofollow links. Traditional SEO work has avoided nofollow links since they don’t appear to pass link juice. However, this most recent study ranked nofollow links as the second only in ranking impact to total backlink count, Facebook Likes and Google+ links.
Comments, tweets, links, shares, and most social mentions are nofollow links. Seeing a large number of nofollow links in a link profile is a good sign that there is a lot of social media conversation about the website, which is difficult to fake without significant effort. The thing to keep in mind is this data is correlative, not causative. It doesn’t show that a website added a bunch of nofollow backlinks and jumped in rank. It shows that websites near the top often have a lot of nofollow links in their profile. This should adjust your attitude about only looking for “follow” links. A healthy mixture of nofollow could be the the signals that you are a popular company in the real world and well-deserving of page one ranking.