Recently we have been hearing more about Google’s neural matching but the truth is, we don’t know much from Google on what neural matching is and how Google uses it.
What is neural matching? Google explained “Neural matching is an AI-based system Google began using in 2018 primarily to understand how words are related to concepts.”
“It’s like a super synonym system. Synonyms are words that are closely related to other words,” Google added.
The first time Google Search Liason Danny Sullivan spoke about this was in a tweet around its 20th anniversary in September. Sullivan said a “big change in search” is the ability to understand synonyms. “How people search is often different from information that people write solutions about.”
Sullivan said Google had been using neural matching over the last few months, so roughly since late spring or early summer of 2018.
How does neural matching work? Google said it helps better relate words to searches. The example Google gave us was neural matching helps understand that a search for “why does my TV look strange” is related to the concept of “the soap opera effect.” In this case, Google is now able to return pages about the soap opera effect, “even if the exact words aren’t used,” Google said.
How much is neural matching used? Google said in September 2018 that neural matching impacts about 30 percent of all queries. We asked Google if that has increased, but have not received an update.
What is RankBrain? Isn’t it similar? Google told us in 2016 that RankBrain (see our RankBrain FAQ) is also an AI, machine learning-based system that helps Google understand queries.
Google said a good way to think about RankBrain is as an AI-based system it began using in 2016 primarily to understand how words are related to concepts. “It’s like a super synonym system. Synonyms are words that are closely related to other words.”
So what’s the difference between Neural matching and RankBrain? Google put it this way:
- RankBrain helps Google better relate pages to concepts.
- Neural matching helps Google better relate words to searches.
Why it matters. The truth is, there isn’t much a search marketer can do to better optimize for RankBrain, as we said in 2016. The same seems to apply for neural matching, there doesn’t seem like you can do anything special to do better here. This is more about Google understanding queries and content on a page better than it currently does right now.
That said, it seems to indicate that search marketers need to worry a bit less about making sure specific keywords are on their pages because Google is getting smarter at figuring out the words you use naturally on your pages and matching them to queries.
We asked Google if it has additional recommendations around neural matching and RankBrain and were told its advice has not changed: Simply “create useful, high quality content.”