Google addressed some questions going around the SEO community as of late related to neural matching and how it’s used in search.
Danny Sullivan, via Google’s Search Liaison account, published a series of tweets explaining the difference between neural matching and RankBrain.
Here is an overview of what was shared by Sullivan.
How Google Uses Neural Matching
Neural matching helps Google better relate words to searches. It’s an AI-based system that has been in use since 2018.
Sullivan describes neural matching as a “super-synonym” system:
“For example, neural matching helps us understand that a search for “why does my TV look strange” is related to the concept of “the soap opera effect.” We can then return pages about the soap opera effect, even if the exact words aren’t used…”
In September 2018, Google stated that neural matching is used in 30% of searches.
It’s not known how widely used neural matching is right now, though it would be reasonable to assume its use has only expanded.
Sullivan’s description of neural matching closely resembles what my colleague Roger Montti wrote about it last year: What is Google’s Neural Matching Algorithm?
What is RankBrain?
RankBrain helps Google relate pages to concepts, even when the pages do not include the exact words used in a query.
It’s also an AI-based system which has been in use since 2016, two years before Google implemented neural matching.
There are theories which suggest RankBrain also takes into account user behavior signals, but those theories have been debunked.
So, to sum up, RankBrain relates pages to concepts and neural matching relates words to searches.
