Google doesn’t need to read minds, they rely on behavioral science. There are distinctive patterns of behavior that appear in the data that indicate relative usefulness. You are on the right track, they do use CTR, bounce rates, TOS, session paths, keyword term revisions, and a host of other metrics in their data analysis. Take what you think they do, and multiply it by 20 and you will be much closer to what they actually do.
You’re right, they cannot “read mental patterns”. however, they can and do read behavioral patterns.
In fact, Google holds Patents in the heuristic analysis. One of the most common heuristics methods used in data search algorithms is known as fast a frugal heuristics. Anytime you have a large amount of data you can accurately predict the relative usefulness of a snippet of text. This isn’t science fiction, it is math. For more information about this topic, pick up a book on simple heuristics.
Perhaps you are right? Maybe Google lets those hundreds of scientists and mathematicians they employee sit around doing nothing. Or, maybe they do actually use science and math to solve difficult equations.
There are brilliant methods, worked out by scientists like Markov, Gigerenzer, and Goldstein, just sitting there waiting for someone to apply them to an algorithm. Perhaps they haven’t overlooked those methods and actually do use them?
Maybe we are both underestimating?
Let’s just agree to disagree on this point.
Let me share a brief story with you…
About ten years ago I built a little website for my daughter’s new profession. Not being an expert on the topic I randomly selected a bunch of stuff that I thought might be interesting.
To my surprise, a page deep within the website, a page that I thought would be overlooked filler information, started receiving massive amounts of traffic. A quick check revealed that it was listed no.4 in SERP for a keyword that got a lot of traffic. So I went in and expanded the content on that page, and made it easier to sort and navigate. Within a couple of months, the page hit and retained the number one position for many years. I never built any links to that page.
After much study of this situation, and the type and nature of the content, I realized that Google had a method of analyzing content for usefulness. I looked at thousands of other websites to see if I could find the same pattern of high ranking content without backlinks and found a definite pattern of ranking content based on content usefulness.
This led me to an adventure where I have been documenting content structure to find a number of patterns that automatically rank higher based on nothing other than the predicted use of that content.
Now before anyone asks, I will not reveal exactly what I have learned because I don’t want to see it exploited anymore that it already has been. I’m not the first person to discover this, and I’m sure I will not be the last. I am perfectly comfortable with most people thinking it doesn’t exist.
If you are smart enough to figure it out for yourself then good on you.
