Search engines are leveraging artificial intelligence to create more relevant responses to search queries.
6 min read
Opinions expressed by Entrepreneur contributors are their own.
Understanding SEO is the first rule for being able to optimize your site and its content so that would-be customers can actually find your company online. When your SEO is on point, you are more likely appear in search engine results when someone looks for the type of products or services you sell. However, to be able to truly understand SEO, you need to comprehend how modern search engines work — and that means understanding artificial intelligence, or AI.
What is AI?
AI is a technological advancement that enables a combination of hardware and software to function like a human brain — minus the inherent flaws in logic and the relatively small memory capacity. It makes it possible to not only analyze large amounts of data but to draw meaningful insights about the information. In some cases, these AI-enabled insights are more insightful, quicker or on a much larger scale than would be possible with human involvement — usually all three.
And those insights can help build better search engine results. “If you feed enough photos of a platypus into a neural net, it can learn to identify a platypus,” writes Cade Metz, formerly of Wired. “If you show it enough computer malware code, it can learn to recognize a virus. If you give it enough raw language — words or phrases that people might type into a search engine — it can learn to understand search queries and help respond to them.” In other words, AI learns over time.
In general, there are three types of artificial intelligence:
- Artificial Narrow Intelligence (ANI): This type of AI focuses on one function (e.g. chess, spam filters).
- Artificial General Intelligence (AGI): When AI has a general, overall function. Usually, this is equated to the ability of a single human.
- Artificial Superintelligence (ASI): This is AI that is beyond Artificial General Intelligence. The neural network operates at a much higher capability than AGI.
Google’s RankBrain is a type of ANI. Specifically, it is a connection-based system that mimics the way people learn. Also called deep learning, connectionist ANI uses back-propagation to identify flaws in the output of the system (e.g. misidentifying a duck as a platypus) and uses that information to inform future query results.
Google RankBrain (and how it differs from previous updates).
Google RankBrain is the current leader. The rollout of RankBrain began in 2015. It is a search engine system that uses artificial intelligence to create more relevant responses to search engine queries. Previously, Google’s focus had been on algorithms designed by humans.
It took time, but RankBrain slowly started to replace the use of algorithm-based technologies. Today, it is the prevailing search engine technology, and it is changing everything.
Marketing is already shifting directions.
Already, RankBrain has changed SEO, but the result has not been linear. The algorithms that Google used before still exist. What has changed is the way those methods are applied. RankBrain uses deep learning to figure out which combination of those algorithms will produce the best search engine results.
“For instance, in certain search results, RankBrain might learn that the most important signal is the META title. Adding more significance to the META title matching algorithm might lead to a better searcher experience,” explains John Rampton, founder of the online invoicing company Due. “But in another search result, this very same signal might have a horrible correlation with a good searcher experience. So in that other vertical, another algorithm, maybe PageRank, might be promoted more.”
The exact algorithm mixture varies considerably each time. RankBrain isn’t trying to figure out which combination is best, per se. It is determining which sites are good and bad.
What makes a site good.
RankBrain learns which sites answer search engine queries best and analyzes them for common features. This includes the keywords used and the density of those words, but that’s just scratching the surface. RankBrain also looks at the structure of the site. For instance, if a website is overly generalized, it could flag as “bad” and be ranked lower than a more niche site. RankBrain also looks at the backlinks. For example, a shoe website linking to a clothing website is perfectly normal but linking to a computer website? Not so much.
To avoid misclassification, you need to follow the parameters that RankBrain has found to identify “good” sites.
Adapting to the new normal.
Good sites are becoming the new normal, and that fact is changing the way we SEO. Sticking to a niche and using quality backlinks helps, and that’s just the beginning. You can adapt to the new normal in SEO marketing by understanding how RankBrain works.
Start by prioritizing quality. Whether the issue is graphics or content, quality matters. It creates more engagements amongst users and earns more links amongst your peers. RankBrain pays attention to this information and adjusts your search engine ranking accordingly.
User experience also matters. Ease of navigation, quality graphics and overall usability make a real difference. RankBrain isn’t going to rank a “crummy” site very high. Having a similar structure to the sites that leaders in your industry have helps.
Relevant content experience.
Content relevance counts too. When someone comes to your site to buy a sweater, you shouldn’t send them to a blog post about Fall trends. Likewise, links should be anchored in relevant keywords and connect to sites that make sense with those anchors.
SEO is becoming a game of cat and mouse, with Google constantly changing the game. AI and, by extension, big data is a huge part of that. Help your company stay on top by adapting to the new normal in SEO.