Writing for SEO in the age of Google BERT

« Ecrire pour le SEO à l’heure de Google BERT »
Rolled out in France since December 9, 2019, Google BERT (Bidirectional Encoder Representations from Transformers) is a fundamental step forward in how search queries are understood AND interpreted. In this context, how should you write for SEO in 2020?
BERT is a natural language processing (NLP) technique based on neural networks. Google released it as open source in November 2018.
Before being integrated into the search engine, it was used by other applications (including Facebook and Twitter).
A few examples of how BERT is used:
-
Classification and sentiment analysis
-
Natural language inference
-
Categorizing movies through prediction based on sentiment analysis (IMDB).
In this context, what does writing for SEO mean today?
What is at stake for content, and how do you produce high-performing content that drives ROI?




