Google’s latest algorithm has implications for SEO. Here’s what you need to know.

Have you ever tried searching for something on Google but you can’t quite come up with the right words? You sorta know what you want, but you don’t actually have the right key terms. Or maybe you don’t know at all, and you’re wishing something could jump in and finish your sentence for you.

Sometimes you don’t know what you don’t know, and it’s hard to find what you’re looking for if you aren’t sure what it is. That seems like an intractable problem, right?

Fortunately, Google has been hard at work to solve this for many years. Back in 2015, their RankBrain algorithm was designed to deliver more relevant results based on user data. Now, neural matching is designed to fill the gaps to deliver the answers to questions users didn’t know how to ask.

One of the driving forces behind these updates has been the growing popularity of voice search. If you’ve ever heard someone using a voice search—or if you’ve ever used one yourself—you’ve noticed that no one yells keywords at their phone. Instead, we speak in complete sentences, and those sentences tend to be comprised of shorter words rather than dense key phrases. We might search for “automotive repair,” but we’ll say “car shop.”

Google’s mission with neural matching is to recognize not just simple synonyms (“automotive” and “car”), but so-called “super synonyms” as well—even when they’re comprised of long, complex sentences. Sounds complicated, right? Let’s try to clarify the situation.

What is the difference between neural matching and RankBrain?

As we were just describing, RankBrain uses user data, such as location or search history, to deliver more relevant results to users. This algorithm doesn’t override other ranking signals, such as keywords, link strength, or content depth. Instead, it organizes results based on those factors, then applies RankBrain to that data set to determine if some of those results are more relevant to the search user.

For example, if you’ve recently looked up a number of articles on how to follow a vegan diet and then start searching for recipes, it may rank vegan recipe results higher than others, even if you don’t use the term “vegan” in your search query.

On the other hand, Google’s neural matching algorithm effectively works like a dictionary in reverse. Instead of typing in a term and receiving a definition, you can search for the definition and receive the term you were looking for. This works because, rather than strictly relying on keywords and close synonyms, neural matching relies on AI to discover super synonyms that can better relate words to concepts.

Currently, whereas Google uses RankBrain on all queries, neural matching only applies to approximately 30% of searches. This means that Google will still prioritize keyword matches and RankBrain relevancy first, but pull in neural matching any time the search seems ambiguous.

What does neural matching look like in action?

All of this is very interesting, but what does it actually look like in practice? Well, let’s say I search for the phrase: “What is the psychological term for recognizing other people have thoughts and emotions that differ from your own?”

The answer I’m looking for is the “Theory of Mind.” If you’re only thinking of Google as a term-matching machine, you’ll notice that neither “theory” nor “mind” appear in my search query. It’s unclear whether RankBrain would help with this either, unless I had prior search history that might shed some light on my intent.

Overall, this search query is fairly complex, and while any English speaker could understand what I meant by the phrase “the ability to recognize that other people have thoughts and emotions that differ from your own,” jumping from there to “Theory of Mind” is quite a leap.

And yet, in my Google search results, the Wikipedia article on “Theory of Mind” is ranked third:

Search results showing Wikipedia's "Theory of Mind" article ranked third.

The first and second results are also clearly related to Theory of Mind, which is largely concerned with the ability to empathize with others.

What does this mean for your SEO strategy?

We’ve talked in the past about the value of long-tail keywords, and how being able to perfectly match a long, detailed query is likely to result in higher rankings. To give a classic example, if you try to optimize your page for the word “shoes,” you will struggle to rank for that term because the search volume is so high. Even worse, you will have no idea if you’re attracting visitors who want to buy baby shoes, high heels, or work boots.

A term such as “best orthopedic shoes for women” has lower traffic volume, but that makes it easier to rank for. Even better, people searching for this term are probably more interested in purchasing orthopedic shoes from your store.

All this is still true, but it applies mostly to traditional keyword strategy. What neural matching means is that Google is becoming better at delivering relevant pages, even when the searcher doesn’t know what keywords to use.

This does not mean keywords are unimportant.

Even if the searcher doesn’t know what they’re trying to find, Google does. The job of an SEO is still to optimize a page so that it communicates the page content accurately and effectively to both the reader and Google. If a searcher knows what they want, good keyword implementation will help them find it. If the searcher doesn’t know what they want, Google can figure out what they mean, and good keyword implementation will help Google deliver it.

Algorithm updates put the primacy on searcher intent and content quality.

High-quality, relevant content is more important than ever. As Google becomes more effective at weeding out low-quality posts, the best will have to compete more rigorously against each other to come out on top. To satisfy both Google and the user, content has to be laser-focused on searcher intent.

Keyword rankings will still be a reasonable measure for understanding how well a website is performing in their market. However, keyword research will need to move past identifying what users are looking for and instead focus on discovering why users are searching for specific terms. And on-page optimization strategies should focus less on a one-keyword-one-page strategy and instead pour more resources into long-form pages that can be a single, authoritative source for a search query.

Businesses that focus their SEO strategies on delivering high-quality content will continue to have an advantage over those who don’t. As Google’s algorithms improve, it will only get better at matching the best articles to searcher intent. Write content your readers want to see, and Google will make sure it gets to them.

Published 12/13/19 by Laura Lynch