Ahhh, great topic! This has been all over the news lately, and I'm interested to hear what folks around these parts think about it.
I think I'll move it over to Google Optimization where we can try and draw more eyes!
This to me sort of suggests targeting long tail search terms may become less effective. Ie the direct matching text on a page may help less than it used to. If google is deciding a very specific search phrase is actually being better answered by a more generic and common search phrase.
Live the moment
I've been following the RankBrain news with great interest. I find it funny to watch on Twitter as someone writes their thoughts about RankBrain and then someone else criticizes them for writing about a subject that they know nothing about. But in reality, I don't think too many of us are experts in machine learning.
I do have some theories about RankBrain though. I am wondering if this is perhaps why we did not see a rerun of Panda or Penguin for so long. If RankBrain started being used in early 2015, it would make sense that perhaps they didn't want to run Panda or Penguin while they were seeing what the effects of the new ML algo would be like.
I also wonder if RankBrain will impact Panda. I feel that in the past, a HUGE component of Panda involved technical site issues. But I feel that if RankBrain is able to learn more and more about what users queries are and what content answers those queries, it's possible that this data could be incorporated into Panda. In other words, if Panda is supposed to surface high quality sites, perhaps RankBrain will make it so that sites that truly are the best experience for users will get some type of a Panda boost.
If this is true, then Panda will be more and more about truly providing the best quality possible and a little less about trimming thin content and having sound technical SEO.
Hi Marie. Nice to see you posting. The idea of having AI incorporated into search results certainly suggests less needs for periodic updates as it suggests that the algorithm is learning and morphing on a more continuous basis. This is likely to make manipulation more difficult.
The 3rd most important… So I would guess 1.Relevance 2.Link juice 3. Rank brain…. So google is (according to quoted article) saying it is now trying harder to work out exactly what people are really looking for when they type a search query and not just return the results that most superficially relate to that query. This is a logical next step I suppose as in the past google has distinguished between navigational, informational and transactional search queries. Now they seem to be implying that people do not necessarily know how to use search queries to gain the desired search results and this AI will better interpret the search query to provide more relevant results.
Has anyone seen any evidence of the reduction in ‘direct matches’ to longtail searches as compared to more relevant/trusted overall content being provided? This could make it harder to capture ‘search phrases’ as less importance could be being placed on the actual phrase.
I'm quite new, but all I've noticed is words like cheapest being grouped together with the word cheap. if rankbrain is the third most important signal it's still in its infancy and doesn't really change alot.
From what I can gather machine learning is either highly advanced and or very basic, there is no inbetween. Google have the resources and the information. But do they have that much of either and who is going to sift through all that information. It's not rankbrain they need. It's a greater undertstanding of the nature of language and how it's used.
Is this new thing gonna affect the search listings? What needs to be done to improve listings? Any ideas?
I haven't seen too much chatter about it, lately. Seems to be one of those things that has become "wait-and-see," though I'm sure some bright bulbs are out there, working hard and performing experiments to see what's new. WebmasterWorld had a robust discussion back in late October, early November but it seems to have petered off for now.
Forbes had a belated write-up of it as well, back on Nov 12.
Aye, other than the common buzzwords and the knowledge that it's "the third most important ranking signal," I haven't heard much about RankBrain since it was announced. Partly, I think people are focused more on Panda, Penguin, and the Jan 2016 core algo update. Also partly, I think people either don't know enough yet...or they know more than they're willing to share.