Search ranking factor studies have become very popular, but there’s a real danger if they lead to unwarranted conclusions.
In this episode of our popular Here’s Why digital marketing video series, Stone Temple’s Eric Enge explains the cautions we should apply to ranking studies, and some ways in which they are useful.
- See all of our Here’s Why Videos | Subscribe to our YouTube Channel
- Damaging Our Industry with Ranking Factor Studies
- Do Links Still Matter for SEO in 2017?
Mark: Eric, sometimes our industry seems obsessed with ranking factor studies and claims. Recently, our friend Jenny Halasz wrote a blog post rant about that which has drawn a lot of attention and debate in our industry. Provocatively titled “Damaging our Industry with Ranking Factors Studies,” Jenny alleged that because they are too often misinterpreted many such studies may do more harm than good. What do you think, Eric?
Eric: First I think we should be careful about saying such studies damage the SEO industry. Even Jenny is careful to say in her post that the real damage isn’t the studies themselves, but the interpretation given to their findings, whether by the study authors themselves or by others who read and share them.
Mark: Most of these studies are based on correlations. Explain what that is.
Eric: Put very simply, a correlation is an observation that two or more situations tend to occur together with varying frequencies. So correlations are rated by their strength, that is, the more frequently the situations occur together or the more strength in one seems to be when the other is present, the stronger the correlation.
Mark: But as is often said, correlation doesn’t prove causation.
Eric: Right. It’s possible for two things to occur together but have little or nothing to do with each other. My favorite example is the fact that ice cream sales and drowning deaths are highly correlated. So someone might conclude that increased ice cream sales causes more drownings or even more sillyly, vice versa. But we know the real reason the two things correlate so well.
Mark: They both go up in the summer when the weather’s hotter.
Eric: Exactly. Anyone leaping to a conclusion that sales of ice cream causes more drownings didn’t show that a third factor not being measured may have affected both. So because we don’t know for certain most of the individual ranking factors search engines use, some SEOs or other companies supporting their work conduct correlation studies looking for things that may or may not be present to varying degrees with pages that rank high in search.
Mark: So these studies can turn up some interesting things though sometimes, right?
Eric: For sure. But as we said, you can’t jump from the observation that something correlates well with high ranking pages to the conclusion that it’s a direct ranking factor. In fact, sometimes it might be the opposite.
For example, one study found that pages that rank well tend to have higher direct traffic and concluded that the amount of direct traffic the site gets is a ranking factor. But it’s just as possible–and I’d say much more likely–that it’s simply the case that sites that are able to rank well also tend to be sites that have become so well known and trusted that they get a lot of direct traffic. In any case, there’s no warrant to conclude that direct traffic is a ranking factor.
Mark: And then on top of that there can be a problem with trying to tease out individual ranking factors.
Eric: Yes, and in reality the reasons one page ranks over another are usually quite complex. Search engines never apply just one factor in evaluating where a page should rank. It’s always a complicated mix of factors with the weighting of each varying by the query and other circumstances.
Mark: So it’s nearly impossible to isolate any one thing as a causal factor.
Eric: Right. Go back to our ice cream and drowning example for a moment. The correlation only seems significant if you ignore that there could be other factors, such as time of year. And in the case of search, the problem is compounded by the fact that we don’t even know what all those factors could be.
Mark: So do correlation studies have any value?
Eric: Yes, they can have value if approached with the right cautions. A particularly high correlation might launch further investigation or spur deeper thinking that could be useful.
Take that direct traffic correlation for example. It’s much more useful to think of it in terms of what are the overall things that sites are doing that have both good rankings and high direct traffic? What does their brand marketing look like? What are they doing to build a great reputation in their industry?
Mark: And occasionally, they can be useful when we do know something is a major ranking factor in evaluating how important it is. A good example of that is your series of studies evaluating just how important links are in ranking.
Eric: Yes, because in that case Google clearly states that links are one of the most important ranking factors. So the only question is, how powerful are they? But even there, we’re careful to state that links are never the only factor involved. For example, if you have really terrible content all the links in the world aren’t gonna help you rank.
Don’t miss a single episode of Here’s Why with Mark & Eric. Click the subscribe button below to be notified via email each time a new video is published.