We could give you a sassy intro on Google and what an algorithm is, but we’ve done that (see intros on both links, literally). And to be quite honest, we prefer to only repeat ourselves accidentally. Besides, if you’re here, you probably already know enough about Google or you’re algorithm-inclined to the point that if we tell you any more basic information, you’ll explode in a shower of 1s and 0s. The good news is we have plenty to say about Rankbrain and how it functions in 2018.
Google’s Rankbrain in 2018
If you don’t know or maybe have forgotten, super quick refresher: RankBrain is one of Google’s predominant algorithms. Sure, but why does that matter? Well, Google isn’t super forthcoming about what their algorithms look for in websites for obvious reasons. If those reasons aren’t obvious, it’s because black hat seo is, as always, a thing. A dumb thing, but a thing.
If you’ve forgotten, black hat SEO is the uncomfortable pair of underwear in the SEO world. Should you wear them when you know you own happy, resilient underwear? No, of course not, but they’ll get the job done. Hey black hat SEO: maybe wash your underwear more often and you won’t have uncomfortable and unnecessary decisions. Anyway. If Google were open about their algorithms and what they look for, black hat SEO proponents would latch on and exploit these measures to a point that could drastically compromise the integrity of search results. That’s why we have to dissect what we know about their algorithms instead of reading an easy pdf from Google on what they look for.
To hop back on the RankBrain train specifically, let’s cleanse our palette with some quick facts:
RankBrain Quick Facts:
- RankBrain ranks alongside other preeminent algorithms such as Panda and Penguin.
- RankBrain is considered a part of algorithm Hummingbird.
- RankBrain first launched in 2015.
- Google considers RankBrain to be the third most elevated/important ranking factor for its indexed version of the web.
- RankBrain’s most speculated role is to identify relevancy of content/topics as well as implications of what was inferred by a search query (more on this below).
- From what we understand, RankBrain can rule out poor UX (user experience) and crappy, depth-lacking content especially.
Quick facts achieved. Depth time.
RankBrain in 2018: Status
What do we know about RankBrain and how we can work with it in 2018? What can we do to adapt to what it looks for?
What Do We Know?
Since we’re feeling mighty list-y today, here’s more information that has been gathered on RankBrain up through 2018:
- RankBrain is machine-learning artificial intelligence (look out, Skynet!)
- This information has been confirmed by Bloomberg, Search Engine Journal, and Google themselves (to Search Engine Journal and Bloomberg both).
- “RankBrain uses artificial intelligence to embed vast amounts of written language into mathematical entities — called vectors — that the computer can understand.” – Bloomberg
- RankBrain processes what Bloomberg called a “very large fraction” of search queries. We’re guessing that means most search queries.
- According to Search Engine Journal’s report (linked above), shortly after being deployed in 2015, RankBrain quickly became one of the most important factors in ranking websites at Google. As mentioned, it even made top three.
- RankBrain is less likely to exist as a ranking signal and more likely to function as a query processing tool, posits Search Engine Journal.
- RankBrain is also exceptional at figuring out implied meanings. To elaborate, if you type “Benedict,” based on your previous contexts (Marvel, Doctor Strange), it’s likely to understand the implication that you’re searching for Benedict Cumberbatch instead of Benedict Arnold.
- A third function of RankBrain (although we’re sure it’s got waaaaay more than three functions) is that it measures user satisfaction based on interaction with links. If the user stays on the page that was suggested, that page’s rank gets upgraded. If the user leaves the page (is dissatisfied), then RankBrain will try another page next time.
How Do We Adapt to Google’s RankBrain in 2018?
One of the vital takeaways here is that TF*IDF (entire article on this from us to come soon, btw!) is how you can give RankBrain the data it’s looking for. RankBrain judges the weight of your words, as should you. What’s TF*IDF? For a quick summary, TF*IDF is essentially a mathematical/statistical equation that weighs the weight of words that have greater magnitude (nouns, verbs, etc.) against emptier words such as “the” or “a.” The outcome is a ratio which tells RankBrain how meaningful your content or site actually is. Thankfully, there are a plethora of TF*IDF tools available, such as the tool in WebSite Auditor. Complicated math unnecessary. Basically, make sure your words matter. Concise ftw. Secondly, stop worrying about long tail keywords. With RankBrain, they’re no longer a concern and are obsolete. This means that searching for “recipe cheesecake” and “recipe for cheesecake” will produce almost identical search results. You should, however, use medium tail keywords. These are not single- or double-worded keywords, but often three or four words long and provide a more stelwart focus. As Backlinko uses in their lovely example, using “Paleo Diet Recipes” provides drastically heightened search volume than their single- or double-worded counterparts (“Paleo” or “Paleo Diet”). Hone your keywords to this proven principle. Keep keywords more specific (but not too specific). Medium tail is your friend, tried and true (and proven).
RankBrain and SEO in 2018
RankBrain is only one of eons of SEO factors to consider, albeit a decisive one. We haven’t even covered all of the possible angles of Google’s algorithms, so you can only imagine how deep that knowledge vein runs. Guess what, though? That knowledge vein courses through our office. It’s like a damned SEO river in here. We’re here to help you cross that river.
Contact us today and gain access to all of our knowledge. Join forces with us and give your site and brand the ranking it deserves: absolute domination.