SEO Ambiguity & Pattern Recognition
As search marketers, we constantly have to deal with ambiguity — whether trying to reverse engineer a Google algorithm or simply prioritize and forecast the impact of the SEO recommendations we’re making. In this article, we’re going to explore what drives SEO ambiguity. Spoiler alert: some major survival skills are required. When dealing with SEO […]
As search marketers, we constantly have to deal with ambiguity — whether trying to reverse engineer a Google algorithm or simply prioritize and forecast the impact of the SEO recommendations we’re making. In this article, we’re going to explore what drives SEO ambiguity. Spoiler alert: some major survival skills are required.
When dealing with SEO ambiguity, it’s important first to note that the human brain is predisposed to seeking out meaningful and meaningless patterns. This concept, defined as “patternicity” by Skeptic Magazine founder, Michael Shermer, potentially creates two types of cognitive errors: we believe a pattern is real when it is not (Type 1), or we don’t believe a pattern is real when it is (Type 2).
As a survival mechanism, our brains have steadily evolved to embrace Type 1 errors, assuming that all patterns are real and meaningful. In SEO language, this means that overall, we are prone to creating data correlations where none exist.
In a recent Moz blog, Rand Fishkin pointed out that data correlations do not always give way to causation, and I tend to agree. Imagine you notice a data spike and need to determine whether it reflects a reporting error or is related to a specific change you made. Your pattern-detecting brain is inherently wired to associate the mystery with similar past experiences. This leaves you vulnerable to making a cognitive Type 1 or Type 2 error.
But if you take a moment and opt for a different route, such as using social data correlations (as Brian Massey did) to rule out a reporting error, you may have better results. When faced with ambiguous situations like this, just asking yourself whether you are inching toward a Type 1 or Type 2 error can help avoid false data correlations.
Patternicity also primes us to see the same pattern over and over again. The decision to change title tags is a great example of SEO ambiguity that can easily fall into this category. For example, perhaps you changed title tags in the past and witnessed an impact on rankings of those pages. But will changing the home page title tag impact rankings on other pages at risk as well?
If you assume a home page title tag change will not impact rankings on other pages, the potential risk of this decision may result in a Type 2 error. Making a decision based on the belief that the change will impact rankings avoids risk, but may result in a Type 1 error. Right or wrong, you are hard-wired to potentially make either a Type 1 or Type 2 cognitive error. And in true patternicity form, the more out of control we feel, the more patterns we find. Knowing this and arming yourself with more data will allow you to consciously avoid this type of SEO ambiguity.
Although pattern recognition certainly has its flaws, technology has made tremendous advances toward algorithms that use pattern recognition effectively. In response to the review arms race, or the current trend of buying or selling favorable product reviews, Cornell researchers developed an algorithm in 2011 that detected fake online reviews with a 90 percent success rate. The algorithm’s capacity for pattern recognition surpassed human ability to accurately detect the same patterns.
Once we are aware that our brains default to pattern seeking in response to ambiguous situations, it becomes easier to see how we are pre-wired toward cognitive limitation. When it comes to SEO ambiguity, remember, “this” isn’t always connected to “that.” The best defense is to be aware of this limitation. Think creatively and let go of any assumptions. And above all, test for statistical significance, while leaving room for possibility and results.
Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.