Causality and SEO

I had an interesting discussion at work. Around the same time we improved upon our Google News site-map, one department decided to ‘experiment’ by adding a byline to the body-text of our blogs—where bylines were generated by the system elsewhere in the HTML. (The old bylines omitted the word “By [author]“, a fault this department believes costs us dearly.) In the end, traffic attributed to Google markedly increased, but a debate grew between what caused the gains.

But how can one truly prove causality and causation when it comes to SEO? In a perfect world, a researcher would be able to hold constants and use statements like “all things considered equal…”. But when it comes to news, no two stories are truly equal. And since some content lives or dies at the whims of the news cycle, the weight of an article must be judged against all of related news content existing on the internet at the same time.

Even a approach were the same article renders differently for different audience may not stand up to SEO analysis. There are still too many variables at play when you weigh in the content generated by outside sources.

So for our little experiment, if we’d better communicated we could have quantified the gains (or losses) of each method. Instead, we’re stuck with a rift in the office where each side is unable to prove or disprove the findings of the other.