Google keeps its algorithms under tight wraps, so not much is definitively known about how these elusive algorithms work. Over time, it has been possible to learn a bit about some aspects of the algorithm, like RankBrain. Lately, Google has made several announcements regarding a newer technology, Neural Matching. There are still a lot of unknowns surrounding Neural Matching, but after conducting some research it’s possible to understand more about this emerging technology. Armed with knowledge about RankBrain and Neural Matching, you can determine how these algorithms impact site ranking.
Understanding the Pieces of the Puzzle
Google’s content algorithms are vast and diverse. There are different components designed to do different things. However, all of these different pieces work together to deliver the results users have come to expect from Google. All of the content algorithms employed by Google are a work in progress, and changes are routinely introduced. This regular change is part of what makes it is difficult to understand exactly what the algorithm looks like and how it functions.
What is RankBrain
RankBrain is a long-established element of Google’s core algorithm that puts machine learning to use. Essentially, RankBrain uses an interpretation model that includes a variety of factors to deliver more accurate results intended to fit each user’s needs.
An example of this machine learning in action would be if a user searches about a hurricane. Before RankBrain, pages about hurricanes that had received a lot of traffic most likely rank very high. However, thanks to RankBrain, during an active hurricane, a page tracking the storms progress that was recently updated would rank higher than pages that have a proven track record with the search term. RankBrain can identify and prioritize different pages at different times. This is an attempt to give users relevant results that fit their needs at that moment.
How RankBrain Impacts Your Ranking
Consistency is important in a RankBrain landscape. You need to decide how your site will serve your users and ensure your site continually fulfills this purpose. Over time, RankBrain will get to know your site and understand your reputation. When a user searches for something in your wheelhouse, RankBrain will ensure your site ranks.
What is Neural Matching
Neural Matching utilizes Google’s AI technology to better understand language. The intent is to provide more diverse results and as well as results when the user didn’t search for the exact term. With a better understanding of what words mean, how words form sentences and how these sentences build paragraphs, Google can provide more accurate results that align with the search terms.
Neural Matching elevates the use of synonyms, meaning a user can find results for a topic without using the actual words. For instance, if a user poses the question to Google, “why does my TV look strange,” Neural Matching can deliver results concerning the soap opera effect. Neural Matching can decipher the words used in the search to determine the full meaning and the actual question. Google’s content algorithm has long been able to understand synonyms, but Neural Matching significantly builds on this previous ability.
Creation of Neural Matching
Neural Matching is still new, so all that is known is what Google has shared, but sharing is not Google’s thing. It’s likely Google has been building and refining this technology for a while. Looking at various research papers (like Contextual LSTM (CLSTM) Models for Large Scale NLP Tasks, A Hierarchical Attention Retrieval Model for Healthcare Question Answering and aNMM: Ranking Short Answer Texts with Attention-Based Neural Matching Model) and patents it is clear that Google and others have been working to create algorithms specializing in:
- Word prediction
- Next sentence selection
- Sentence topic prediction
Neural Matching is diverse, so it is likely that this technology isn’t just one algorithm, but is several different algorithms working together to understand better a user search to provide very specific and dialed in search results.
Working With Neural Matching
From an SEO standpoint, Neural Matching provides an opportunity to custom tailor a site to better reach a niche audience. Google’s algorithms can now understand exact search terms, super synonyms, and question-based searches.
The main takeaway from all of this research is the importance of structure and meaning. Taking the time to plan and write content that is concise and focused and provides value will deliver results. Through understanding words, sentences, and paragraphs, these algorithms can find the ideal match for any given search. Google is working to serve better their users, which means websites need to keep up to help Google help users. Creating content that addresses these algorithm advancements will positively impact site rank.