Google recently made another one of its interesting yet ambiguous announcements: the search company is now relying on the use of an artificial intelligence system known as ‘RankBrain’ to help monitor its search results, and eventually guide the development of their core search algorithm. In truth, over the last several months a substantial percentage of all Google queries were actually handled by RankBrain. To the average user (i.e. you and me) we haven’t noticed anything especially different. However according to Google’s plans and available data, this new move to RankBrain will make the web an easier, better place to search.
Nevertheless, if we haven’t noticed much of a difference, then why is RankBrain such a big deal? What is it truly doing, and why should that be relevant to the average user?
The Basics of RankBrain
RankBrain is an artificial intelligence system being applied to Google’s current search engine algorithms to try and provide better results to user queries. It is rooted in machine learning, and will use mathematical processes and advanced understanding of language semantics to gradually learn more about why and how people search, then apply those calculated conclusions to future search results. Instead of being pre-programmed to respond to certain situations in a pre-determined way, it can actually update itself over time. Think of it as a robot that’s continually seeking for upgrade parts itself rather than relying on a human for maintenance.
What RankBrain Isn’t
To clearly understand what RankBrain is, it’s helpful to spell out exactly what it isn’t:
It isn’t an algorithm update – Don’t think of RankBrain the way you would think of Panda, Penguin, or Hummingbird updates. This is not a new algorithm, nor is it any standalone update to assist that algorithm. Rather, it’s a new modification, working in conjunction with Hummingbird (the semantic search algorithm) to produce more meaningful results.
It isn’t a new Knowledge Graph – Yes, the Google Knowledge Graph is an artificial intelligence program also, but this isn’t a new version of the Knowledge Graph, nor is it directly affiliated with the feature.
It isn’t a new type of search – RankBrain is not a new type of engine, and it isn’t intended to make any significant updates to the process a user relies on to get results, meaning SERPs will remain indistinguishable for the typical user.
It certainly is not a robot – Hopefully, this one is obvious, but the term “artificial intelligence” combined with robot examples (like mine above) could lead people to a misunderstanding of what RankBrain truly is. It is not a robot, nor is it conscious. It’s just a series of mathematical equations that can learn and improve over time.
Ranking Signals Today
RankBrain can understandably be considered as a new ranking signal. And as far as ranking signals go, this one is now said to be the third most important for determining the eventual rank of a given page. However, Google is very secretive about the exact mechanics of its ranking algorithm, so we have no idea of what the top two ranking signals really are, but we do have an understanding of the ranking signals that exist today.
There are actually hundreds of different ranking signals (and possibly thousands of sub-signals). The point here really is about RankBrain’s significance – for it to displace all other ranking signals except two, then it must be seen as important to the future of search.
How RankBrain Will Improve Search
The prominent improvement seems to be the analysis of ambiguous, unrefined, or otherwise difficult-to-interpret queries. An example previously given is a complex and vague query like; “What is the title of a consumer at the very highest level of a food chain” compared to a more concise query with the same intention, “top level of the food chain.” Now, the latter is sufficiently handled by Google’s human-programmed search algorithm, but it can only do so much with the former query.
Over time, RankBrain will begin to learn how to appropriately manage these types of queries, correlating certain ambiguous phrases with the more concise terms to better understand user intentions.
What RankBrain Is Doing to SEO
Despite its significance to the future of search and status as a ranking signal, RankBrain isn’t going to have much impact on SEO practices. Its effects primarily deal with exceptionally complicated, ambiguous, or poorly worded long-tail queries. This may lead to some changes to ranking on a very small scale, but otherwise, traditional best practices will remain the same. For instance, if you’re optimising for queries related to flip-flops, RankBrain is going to help ambiguous, relatively rare queries like “the things you wear on your feet at the beach” to more concise, relevant, appropriate terms. It could be possible that Long-tail keyword strategies might be influenced slightly, but for now, it doesn’t seem like RankBrain necessitates any significant changes to your current SEO campaign.
As with any new Google feature, RankBrain will expectedly continue to develop over time. As a consumer, you might not notice these changes, but do note that they will have significant relevance to a search results page, and may even dictate the future development of search algorithms. For now, don’t make any drastic changes to your strategy, but as always do pay close attention to your rankings.