Overloaded with content? Julie Joyce explains why it’s essential for SEOs and link builders to learn how to wade through a sea of potential misinformation.
I like to joke that whenever Moz publishes an article about links, half my clients immediately email me about it. But that’s really not too far from the truth! People with big audiences have a lot of power.
But people can make mistakes — even people with strong expertise in a subject — so you do need to be careful trusting information without backup sources. In my opinion, the best thing about Moz is that in the comments, people will call you out, ask questions and offer alternative points of view.
People will also call you out on social media, but I’ve noticed that it doesn’t always happen with smaller sites that have smaller audiences. If an individual is writing on his or her own blog and just getting started, especially if that blog doesn’t allow comments, the writer can say lots of untrue things and no one will even notice — other than maybe your poor client.
When the data isn’t what it should be
People do make proclamations based on faulty data. I recently was involved with a study that turned out to be based on a lot of incorrect information for various reasons.
When I pointed this out to the person conducting the study, he never spoke to me or interacted with me again. Did he let anyone know what I’d said? No. It would have changed his conclusions. He didn’t want to admit that he’d been wrong or overlooked something.
This cannot be an isolated experience.
Many sites publish studies, so let’s say a client forwards you a study that draws its conclusions from a very small sample. Do you really want to trust your marketing to a test that only looked at three sites?
If I have built exactly three links for two clients in the same industry (let’s say gaming), and both of those sites improved five spots in rankings for their most important keyword, I could theoretically say that all gaming clients needed only three links to see an improvement in their rankings. But the fact is, such a small sample is unlikely to produce statistically significant results.