fbpx

GOOGLE SEARCH MEETS AI: THE BEST SEO PRACTICES FOR 2020

Something huge happened in the search engine industry in the October of 2019 that could potentially change the entire SEO landscape. Google has rolled out an algorithm update unlike any other that it had rolled out before – the BERT update. Yet, not so surprisingly, it could mean a great relief or a big blow to you depending on what you’ve been doing with your SEO so far. But the fact is, 2020 is going to be different.

Google as a search engine has evolved by an unimaginable magnitude since its inception roughly 22 years ago. What started as a simple PageRank algorithm that mathematically ranked web pages based on the links that point to them (a link-voting system, if you will) has now turned into an AI beast that actually tries to imitate how humans understand written/spoken language, and return them results ( or answers, in most cases) with stunning accuracy. This is big.

SEO, on the other hand, became a thing around the same time. It actually started a year before Google was founded, if we consider the year the term “Search Engine Optimization” was coined. All this while, the core objective in SEO has been to try to reverse engineer what we call are the SERPs, and try to unpack the various factors that Google (or any other search engine) might use to rank web pages. In other words, SEO has been the case of humans trying to imitate how machines understand written language.

That’s ironic, of course, but within that irony lies the key to be successful in the current SEO landscape. First, let’s see why BERT is particularly different from all the other algorithm updates that came before it.

From Keywords to Intent

There is so much that BERT does, but this heading, in four simple words, pretty much sums up what makes BERT different from its predecessors, though it doesn’t help us understand the gravity of the change.

Google made its first significant move from string processing to semantic analysis in 2013 with its Hummingbird algorithm. It was not like other updates it had rolled out before (like Panda, Penguin, Pigeon, etc.), but was a complete overhaul of its entire algorithm. It was geared towards what Google called “conversational search”.

In simple terms, Google tried to move further away from string-matching, and towards understanding the meaning behind what users typed in Google’s search box.

About 2 years later, in 2015, Google made its first shot at AI-assisted search when it came up with the RankBrain update. This supplemented Hummingbird greatly by identifying various ways in which a given query can be interpreted, and trying to improve the quality of results accordingly.

Fast forward 5 years, and Google has now rolled out BERT – what we at BGMG would consider as the AI update because of its significant leap. BERT is significantly different because it enables Google to understand the nuances of human languages a lot better – things like what the users could be looking for when they use certain combination of words (including the so-called stop words). In simple terms, Google not only tries to understand the meaning of a user’s query, but it now better understands the intent behind the query.

What Does It Mean for SEO?

The first takeaway is that the most treasured part of SEO, the “keywords”, are now slowly becoming obsolete. Hence, if you’re still having a long list of keywords you’re desperate to rank for, it’s time to reconsider your approach.

Best Practice #1:

Focus on answers, not keywords

Let us first clear an age-old misconception among many people: What people enter in the Google search box are “queries”, not “keywords”. Unfortunately, the difference between the two has blurred in haste as the practice of SEO became more and more popular. The result was a cult-like obsession towards “ranking for a list of keywords”, which may or may not always translate into relevant website traffic.

Now with BERT in the picture, the obsession becomes even more irrelevant. What businesses instead should focus on is trying to identify the intent of their customers when they search online and answer their questions and needs with the content, not trying to insert more keywords in future content.

Best Practice #2:

Mobile, then Desktop

Google has been pitching for mobile-first approach to building websites for a few years now. The advent of BERT makes it all the more critical. Make no mistake, we’re not talking about a “mobile-friendly” approach – that’s something you should have taken in 2015 or so. Now we should be looking at building websites for mobile first, and then make them compatible for desktop.

The reason BERT plays a key role in this change is that it is intent-based, which means the context in which the users search for specific queries has a lot of impact on the results. The context, more than ever, is a lot more dynamic nowadays because a majority of searches are happening in hand-held devices, which are mostly smartphones. If you do create content that answers your customers’ questions, but your website doesn’t do well on mobile, then you’re still at a disadvantage

We’re not saying you should discard your desktop-thinking altogether. The priority just needs to go to mobile.

Best Practice #3:

Get Comfortable with Entities & Knowledge Graph

Knowledge Graph results in Google search are not anything new, but it would be disastrous to not do anything about it when it comes to your content. Google for quite a while now has been relating queries to known entities (places, people, companies, currencies, and other real world “entities”), in an effort to better understand the meaning behind certain combinations of words.

This means that it can now consolidate information about your brand from the world wide web if it has enough reliable information online. It is up to you as a business to manage your reputation online through your content (by building reliable and authoritative content) as well as the content your customers generate with your brand’s name. This goes way beyond doing SEO from behind your desk, and into focusing on seemingly unrelated areas such as customer service, UX, etc., which brings us to the next point.

Best Practice #4:

UX and Technical SEO

Google has been pushing hard on the technical side of SEO for a good while now. It offers at least three standalone tools for webmasters – web.dev, Test My Site and LightHouse (this one resides right inside Google Chrome’s developer tools and is also the tech that powers the other two tools). It is no coincidence that Google pushes technical SEO this hard. Take care of your site speed, plan on systematic rendering of resources, use CDNs for images, etc. The bottom line is that excellent UX will soon play a key role in your SEO.

Best Practice #5:

Prepare for the Future (where there’s no text)

If you’re conscious of the pace at which AI technology is evolving, it should be of no surprise to you that future searches will happen more and more without the users having to type in their queries (think Siri, Alexa, Google Assistant for now), and you’ll have no more “first page results” to worry about.

The underpinning for these shifts comes from how these search engines are trying to understand content, information, and entities on the web, which ties back to #3. One specific way search engines are able to extract semantic information from web pages is known as using “structured markup data”. It is an additional layer of HTML markup that gives search engines (certainly not only Google) contextual information about your content. Utilize structured data to feed your content with context to any search engine or voice assistant, so they can better process your data and include you in their answers to queries.

Best Practice #6:

It’s Not Link Building Anymore, It’s Authority Building

Link building has always been one of the most misunderstood practices in SEO, for obvious reasons. About 10 years ago, link building was the primary tactic abused by the so-called “black-hat” SEOs who tried to manipulate Google’s algorithm into believing their websites were trust-worthy by building huge piles of questionable links (social bookmarking, article directories, link networks, and what not). This had worked for a while, because Google at the time was so much dumber, relying largely on backlinks for its PageRank algorithm and without making much headway into semantic search.

That stopped working, thankfully, about 5 years ago at least, and if you still get SEO pitches that include some of these tactics, tread carefully. What you instead focus on is to build brand authority in your niche with content – both created and earned. Publish original research pieces, expert opinions in your niche, write on known publications in your niche as a guest writer, and seek out editorial coverage, etc.

The more authority your brand gains, the more chances for you in making your brand future-proof in search.

It is not just the search engines that evolve with technology. SEO evolves along with them as well. The onus is on us to adapt to the changes. The one good news we have is that, if your SEO strategy is user-centric, it is without a doubt a working strategy. It hasn’t changed in years, and won’t change in future.

Have a question?

We are ready to listen and help your organisation.