Good morning search marketers, the machines are catching up to us.
Google welcomed BERT into the search system fold last week. The company's latest search algorithm rolled out for English queries last week and will expand to other languages in the future. BERT will impact the search results for about 1 in 10 queries, Google said.
BERT, which stands for Bidirectional Encoder Representations from Transformers (I will never get that right in trivia, so bet against me now), is designed to help machines better understand the nuances of natural language. As voice queries, in particular, grow, BERT should help Google surface more relevant results.
BERT's fundamental benefit is in helping close the gap between what machines understand and how humans speak and write.
Whereas so much of early SEO was about striking a balance between writing for machines (to gain search visibility) and writing for readers (to gain, well, business), advances like BERT will enable marketers focus more on their customers — and further shift SEO priorities away from on-page content optimization. Google said it's the biggest change since Rankbrain, and, as with Rankbrain, you can't "optimize" for it.
I asked Google if BERT was being used for ads, yet. It's not, but it's easy to see how it could help solve some of the bad close variants matching we currently see.
In other news, Amazon's ad business grew by more than 45% year-over-year in the third quarter. It is focusing on machine learning to deliver high ad relevancy and building out its video and OTT advertising capabilities. The company said it's expecting a record holiday season, despite it being 6 days shorter than last year's span between Thanksgiving and Christmas.
Read on for a Pro Tip on using Google Ads' new audience options and more.
Ginny Marvin
Editor-In-Chief