How Humans Search for Things Online: If That, Then This
Search engines help us get closer to things we desire by decoding the intent behind our verbal cues and filling in the blanks between what we want, and the where, when why and how of getting it.
Humans are an “if that, then this” creatures. If I want that, then I do this to get it. Sometimes, we don’t know what it’s called, nor do we know how to get it, but sure as shit, if it exists, Google can find it.
How Does Google Know What You Want? Even When We Don’t?
Google answers questions the only way it knows how “if not that, then this”. Humans, however, are “if that, then this, this and this might happen.”
Humans don’t always know what they want, but they always know what they don’t want.
Humans get the benefit of conversation, context, and follow up questions. type creature. Humans get to ask follow up questions. Robots get to observe follow up searches.
Before the Search
Google bots don’t understand what you’re asking them. Bots don’t know how to answer a question, they just know how to find the answer to a question someone else has already answered. Google doesn’t know what users are thinking, it just connects users who “they think” are “thinking the same thing”
After the Search
After any search, Google determines whether or not the user found what they were looking for. If 99/100 users click the back button after clicking the first search result, Google bots know what you asked for, but they don’t know what you’re looking for. Google makes its money in gold by understanding what happened after the click.
A robot may infer what people want to find, because after they click on a search result, clicking another link on the page indicates that the content display is at least somewhat related to what they were looking for.
Google not only knows what you click on after you search, but it also knows what you click next, but it also knows what you click after that.
Even if you clicked the first search result, and thought “get me the hell out of here”, you’ll probably keep Googling until you find what you’re looking for.
Google evaluates all of these interactions, creates metrics based on how related the search results were (did they click, download, or keep searching), looks at the previous searches, and then infers the most likely meaning based upon those who were analyzed in the same way.
Now imagine this happening on a scale of billions of search queries a day. Every variation of two or more words allows Google an anonymous-looking glass of what users wanted to find…and what they actually find.
Yeah, robots are still a bit awkward and clumsy — they were trained by the best..