Artificial intelligence or AI is often depicted in pop culture as a kind of malevolent, self-aware technology born out of the arrogance of humankind. Think of the Machines in The Matrix, Skynet from Terminator 2, or HAL 9000 from 2001: A Space Odyssey. Many people have made the observation that these fictional computer programmes hell-bent (or was it code-bent?) on destroying their human creators are a commentary on the dangers of developing AI that’s too smart for our own good.

The reality is a slightly duller reflection of its Hollywood portrayal, though.

While humanity hasn’t had to battle an army of super-intelligent robots just yet, we do live in a world where people are scrambling to curry favour with a relatively clever and massively influential thinking machine: Google search. People ask Google billions of questions every day. At the same time, people are trying various ways to provide them with answers. Recently, however, the search giant revamped the way it produces search results by focusing more on certain qualities than others.

The Advent of RankBrain and Semantic Search



Ever since launching RankBrain as part of its sprawling and comprehensive Hummingbird search algorithm, Google has returned significantly smarter and more relevant results for queries. Alongside updates like Panda, Penguin, Pigeon, and Mobile Friendly, RankBrain’s effect on search has changed the way SEO works — most notably in the content marketing sphere. The days of peppering blogs, articles, and other types of copy with targeted keywords that had little to do with the content they’re sandwiched are gone.

This change had begun even before RankBrain’s introduction. It was at the time when the Hummingbird update started to focus more on interpreting searcher intent rather than simply scouring the internet for matching keywords. Anticipating a user’s intent purely through an algorithm demonstrated Google’s commitment to improving search results, slowly phasing out old link-building techniques in favour of semantic search.

To understand why it matters for SEO to optimise content for RankBrain and semantic search, let’s do a brief rundown on both.

How RankBrain Works



Launched in 2015, RankBrain is a machine learning system that was developed to refine Hummingbird’s core functionality: less focus on keywords and more on context and meaning. The distinction lies in RankBrain’s query analysis and ranking functions. When a user types in a search term, RankBrain’s analytic function attempts to associate it with more frequent searches to pull up better results. This is semantic search in action, where understanding a user or audience’s intent takes precedence over simply collecting pages that contain matching keywords.

The ranking function, meanwhile, compares pages in Googles’ index against parameters like click-through rate, time on page, etc., to determine which of those pages correspond best to the search term. This method of filtering turns up results that sometimes don’t even contain an exact match of the search term, yet still prove relevant. RankBrain is one of Google’s most fickle, yet rewarding ranking signals, just a step below content and links, according to Google itself.

Now that we have a basic idea of how RankBrain’s and semantic search’s cogs work, take heed of the following strategies to make content more relevant — and visible to Google’s enhanced search functionality.

Shifting the focus on your page



Focus, in this line of thinking, pertains to creating content that provides value to the searcher or user. There should be less emphasis on individual keywords and more on topics that address the user’s needs directly. While it still helps to include targeted keywords, they should occur naturally throughout the text and serve the content or topic rather than the other way around. Using semantic keyword groups is a much better approach, as RankBrain will still include your page as part of its results, provided that the content has a strong association with the search query.

Do as the Pros Do



Imitation may be the best form of flattery, and modelling your content style in a similar fashion to how the bigger domains do it can prove advantageous. If you’re running a tech-oriented site, try mimicking the style and format of sites that rank at the top of Google’s SERPs. Google will index anything that’s similar to its trove of authoritative pages as something that’s reputable. This should clue you in on the fact that it works the opposite way as well: if a site is badly designed or is flagged as being spammy, Google will hide it away from prominence on its SERPs.

Be Mindful of the Writing Style and UX Design



As important as the topic of your content is, the style it’s written in is nothing to scoff at either. It may sound counter-intuitive, but it’s a mistake to try and write content with RankBrain and other algorithms in mind. They’re not the audience, and RankBrain will be able to identify unnaturally-written content that’s obviously structured to pander to the system.

Instead, think about your intended audience and write in a way that they would understand and appreciate. This applies to the overall design of the site/page itself: good design is good design, regardless of your purpose for optimisation (search engine vs. audience, for example). It’s always better, however, to strike a balance between operating within RankBrain’s parameters and creating a user experience that perfectly serves your audience’s needs.

Tailor Ranking Factors to Your Industry



There isn’t a single set of ranking factors or signals from Google that applies to all industries. RankBrain appears to weigh a variety of factors depending on the industry, user intent, links, etc. that fits best with a given search term.

Do some research on what your industry’s particular ranking priorities are and apply them to your page or content.

Depending on the query, some ranking factors might affect your ranking score more compared to another site that caters to an entirely different audience. Sometimes fresh content lets you rank higher, other times it might be lengthier or more in-depth copy. And then there are times where it’s the level of engagement or the content authority that does the trick.

Optimising for RankBrain and semantic search is admittedly a tricky objective, but what SEO experts all agree on is that machine learning vastly improves Google’s search function with each new day. It might not be a singular factor to zone in on when implementing SEO work, but it’s still a worthwhile endeavour, especially if it means creating meaningful content and design.