Search used to make sense. Sure, it was ridiculous and barely worked, but at least it was easy to wrap your head around: include a search term, show up on a search for that term. Does your website say “astronauts?” Well, you’ll show up on a search for astronauts. Does your website mention Hamlet? You’d be there whenever a 9th grader had to turn to AltaVista in the heat of a late night paper-writing session.
In other words, primitive search engines looked for simple 1:1 correspondence between search terms and search results – i.e. you’d only get back a page that used your term exactly. That pretty severely limited your ability to locate information unless you had mastered Boolean operators that could return really specific kinds of search results.
It was easy enough.
But search has changed a lot, and Google is a lot less reliant on exact keywords than it used to be – and it hasn’t been terribly reliant on them for a while now. So that’s kinda saying something.
And it’s all thanks to something called semantic search.
What the Heck is Semantic Search?
Semantic search, which Google has been using and refining since the Hummingbird update in 2013, is search that returns meaningful results even when the retrieved items contain none of the actual words being searched for. It focuses on using meaning to determine intent, and then to answer the real question being asked.
Think of it like this. If you google “jfk birthday,” Google doesn’t think you want a website that says “JFK birthday” somewhere on it. Google knows you want to learn when Kennedy was born. And that’s why they’ll return this:
In other words, Google is interested primarily in satisfying user intent and answering their questions more than in returning search results, and they do that by taking a lot into account, from what you’ve searched for already to what the words you’re googling actually mean.
Here’s another practical example: Genghis Khan. If you’re googling “Genghis Khan,” you could mean either the 12th century Mongol warlord, or the hit song by Miike Snow. Just search “Genghis Khan” by itself, and you’ll be directed primarily to the historical figure. But if you’ve been googling around for other Miike Snow songs, suddenly the video for “Genghis Khan” (which is delightful) shoots up higher on your search results page.
What this means is that keyword rankings don’t (and can’t) really exist as objective measurements. You can’t say you rank number 1 for “jelly beans Houston” if that SERP is being generated on the fly for every single person googling it, based on their location, search history, recent website visits, the time of day, and more.
So Do Keywords Still Matter?
Keywords are still super relevant. But that relevance has changed. Instead of being discrete search terms you’re trying to rank for, keywords are indicators of what information people are looking for – and thus major signposts to the kind of content you should be generating. Keyword research gives you a window into figuring out what’s on people’s minds.
Because when you know the question, you can be the answer.
See, here’s the thing. Google’s goal is to be a giant tool by which anybody can answer any question. And that means they need to serve up the best answer they can, as quickly, easily, and directly as possible. And that means no more gaming the system. No more tricks, no more feints, no more focusing on tags and on making sure you have X keyword in Y place at Z density.
It just means being clear, helpful, and informative about the right questions. Which means that your SEO strategy has to revolve around knowing those questions, and on answering them. Nothing is going to get you better organic traffic. Not optimized H1 tags. Not well-crafted titles. Not a fantastic meta description. Just great content that meets users’ needs.