The future of SEO (Search engine optimization)

Suganthan Mohanadasan is a digital marketing Consultant, and we frequently use his expertise on SEO in Dekode. We have asked him for his SEO-prediction for 2021. Here you can read what he believes will be the most important aspects to focus on and what might change for SEO in the next few years.

AI solves problems

– Search engines are getting smarter by the day. We have seen significant changes to search engines in 2020, and my prediction is these changes will evolve and get more prominent in 2021 and beyond, says Suganthan.

His most significant prediction is around how artificial intelligence is starting to play an important role alongside more sophisticated algorithms (Machine learning/Deep learning) in recent years.

He continues by telling that Google posted a series of updates recently about how they’re using AI to solve problems and serve users with better results.

– Few things they discussed include better detection of misspellings (Using BERT language understanding), passage ranking in which Google might find a particular part of your content, answers a specific query and decides to show that instead of your entire document. They also talked briefly about «subtopics» and how Google can find sub-topics related to a parent topic. We may be already seeing this in action in the SERPS. 

Conversational search experience

Sugantan tells us that very recently, Google came up with an update which said they had deployed a conversational search experience.

– With this approach, the search engine will remember your initial search and will provide suggestions for follow-up searches and relate it to your previous search. We first saw this with their Google assistant voice feature where you can ask Google a question about someone. e.g., Who is Barack Obama? and then instead of repeating «How old is Barack Obama?» you can simply say «How old is he?» The new web search feature is similar to this. 

Predictive search

Suganthan says that Google not only has improved the predictive search in which users are not required to search something, but instead the recommendation engine, Google Discover, will proactively show content to the user. (Google Discover cards)

– For example, Google can use implicit and explicit input from users to personalise results. It makes optimising content and doing SEO difficult. Google discover is a recommendation engine which relies a lot on user inputs. We have seen Google doing so many small tweaks recently like adding a «Follow» button to entities, Like/heart buttons for story cards. Google Discover will get better, and we can start seeing Google adding additional features to the feed instead of just showing a bunch of related stories. 

He points out that we already see new features like «Recommended movies and shows» in the feed.

– It’s the same as watching Netflix and seeing shows and movies «Recommended for you» or Spotify «Curated playlists» for you. It means it will be difficult to influence these signals.

Using entities to understand

Suganthan predicts that entities will play a huge role as well, and Google will rely on this heavily in the coming years. 

– Search engines like Google can extract entities from your content, images and even from YouTube videos (Auto captioning YouTube video). They can use entities to understand something instantly as its machine-readable and even combine them to understand the context. 

As an example Suganthan uses an image of a person standing next to a Tesla car.

– Google vision can label the vehicle and the person, then understand that these are entities and get the context of «A person standing next to a Tesla car» without having to rely on older methods like «Alt text» or «Reading the text next to the image» to understand the context.

In conclusion Suganthan has to admit that Search engines are still not perfect. 

Want to know more about SEO?

Visit the home page of Suganthan Mohanadasan to get more information about Search Engine Optimization.