Searching is the most common activity we do online. When we are looking for something on the Internet, we start with a search. Searching became a natural part of our everyday lives and Google knows that. That’s why Google tries to make its search engine more and more natural every day.
A Natural Search Engine
It will be a good idea to start by explaining the concept of “natural search engine”.
In fact, we are all search engines. When someone asks us a question, we dig into our knowledge and find the best answer we have. This is a natural process for us.
Computers are trying to imitate the same process. When you type something like “What is the most efficient way for quitting smoking?” into Google’s search box, Google scans its knowledge base and tries to find the best answer.
However, our answering process is much more complicated and sophisticated than computers’. Our capacity of creating links between concepts is much more elegant than computers’, because we have linguistic skills that computers lack.
As human beings, we can understand the meaning of questions, even when they are formed with different words. The answer to “What is the most efficient way for quitting smoking?” can be the same as that of “How can I stop smoking permanently?”. For a human being, the similarity between these questions is obvious. But for a search engine, these are two different questions which probably have two different answers.
Thus, Google’s main challenge to create a more natural search engine is understanding language. And understanding language not only means understanding words and grammatical structures, but also being able to create links between concepts and contexts.
What is BERT?
BERT refers to “Bidirectional Encoder Representations from Transformers”. Thanks to this new technology, the question answering system of Google will be more sophisticated than before. Instead of evaluating questions word by word, the system will understand the relationship between words in a sentence and evaluate them based on the context. This new form of AI will learn and develop itself everyday.
How Will This Affect Your SEO?
BERT is one of the most important search algorithm updates in Google’s history, but it won’t affect all search queries immediately. BERT is launched merely in English for now, and only a small percentage of search queries are affected by BERT. But Google is planning to launch BERT in different languages soon, and the effect of the new search algorithm will be broader.
As Google hates talking about SEO, we don’t know how exactly BERT will affect search engine ranking positions (SERP) yet. But it’s possible to say that building SEO strategies only on focus keywords will turn out to be insufficient in the future. Creating original and valuable content will be more important than ever.