Nate Tsang

You both agree that SEOs sometimes infer cause and effect when it can be impossible, detrimental, and unnecessary. In other words, inferring a direct causal relationship without sufficient data can lead to mistakes. I really liked that part of your article and I think it was very well put, and a good argument for sticking to solid SEO fundamentals.

"Maybe our only differing point of view is on whether correlation is still interesting when it doesn't point to an underlying causal factor."

However, I think this is the subtle (but very, very important ;) ) distinction between your viewpoints. I see exactly what Carter is getting at, and its not just a semantic distinction.

You're attributing an indirect cause and effect here, whether or not its done consciously. If you think factor N is "interesting" because successfully ranking sites are doing it, you're forming a hypothesis that N might have a cause and effect relationship. Of course, you're filtering that hypothesis through your own logical understanding of how Google works and the available data, but you are doing it whether its done consciously or not.

Example: You notice that 78% of high ranking sites use blue and gray in their color scheme. 

Would you find this correlational data "interesting" because so many successfully ranking sites are doing it?

I'm guessing no - because in your subconscious mind, you tested the hypothesis "does blue/gray cause higher rankings?" against everything you know about SEO, and your brain immediately dismissed any possible causal factor. However, if your brain didn't dismiss it and you find it "interesting", you're no longer just looking at correlational data, you've made a cause and effect hypothesis, whether you're aware of it or not.

That is the subtle but important (imho) distinction between what Carter is saying, and you're saying. Its not just semantics, because its much easier to get cause and effect backwards (or infer cause and effect where it doesn't exist), when you're not even aware that you're making that connection.

The article also says "What you need to do is drive your social media fans through a funnel. First, collect their emails, and then offer your products or services through email."

That seems pretty consistent with "Don't be self-promotional. Don't sell directly to your audience." in the context of social media use.