Difference hummingbird and rankbrain

[seoanalytics]

From my understanding hummingbird is the fact that google is able to parse sentences and link entites to understand the meaning of content in a better way than with just keywords and rankbrain is about user intent, google understands that they are various ways to mean the same thing.

Is my understanding correct ?

Thank you,

[Seenlyst]

Hello,

"RankBrain is an algorithm learning artificial intelligence system" - Wiki

Hummingbird is an update of Google's ranking algorithm.

They aren't something comparable because one is a system and one is a codename of an update to a system.

Hope this answered your question.

[MiriamEllis]

Hi There!

We have two nice resources here on Moz that should help you feel totally clear on Hummingbird vs. Rankbrain:

https://moz.com/learn/seo/google-hummingbird

https://moz.com/learn/seo/google-rankbrain

Hope these help, but please let me know if you have any questions remaining after reading through those! I'm pretty sure I wrote both of them, so if anything isn't clear, just ask

[seoanalytics]

Thank yo. I read it and rank brain is clear hummingbird a little less but I think it is about the knowledge graph and parsing from what I understand.

[Dr-Pete]

I'm afraid it's probably more complicated than that, and I'm not sure anyone outside of Google (and most of the people inside of Google) has a handle on all of the details.

Hummingbird was very broad. It wasn't just an update, but an entire rebuild of Google's core "engine." When Google launched it, they gave examples that make us think a lot of the updates were necessitated by natural-language queries (voice certainly created some of that pressure). So, it definitely changed how Google processed very-long-tail queries, but I think it also created a framework for much more (and may have even been a foundation for RankBrain). Danny's very early FAQ is still a good resource:

https://searchengineland.com/google-hummingbird-172816

RankBrain causes confusion because it gets conflated with ML in search in general, but I think RankBrain has a very specific meaning to Google. I've written about it quite a bit and have had a handful of private conversations with Google employees, and still don't feel like I have all the facts. Here's what I'm comfortable saying... It is an ML-based approach to understanding query relevance, very likely related to models like Word2Vec. Best I know, it acts as a sort of re-ranking layer. So, Google returns results and then RB re-sorts them based on its understanding of relevance. So, truth be told, it's probably not as impactful as some folks think (ML in search could be much broader). It's most active for long-tail, natural-language queries, so there's some connection to Hummingbird, conceptually.

[seoanalytics]

Thank you for the information Dr Pete. It is a little more clear. If I understand correctly rank brain is really about user intent in rewriting the query and Hummbird seems to be about voice searches and parsing but it is a little blurry in my mind as you would say other that people at google nobody really understands it fully.

Thank you,

[Dr-Pete]

I know it's not very clear, but I think the important thing to remember about Hummingbird is that it was a complete rebuild of the core algorithm. I think natural language queries drove part of that rebuild, but Hummingbird covers a lot of ground and will be powering algo updates for months or years. As Danny said, it's like they put a whole new engine in the car. RankBrain is much more specific.

[Dr-Pete]

You can't really "optimize" for Hummingbird, but understanding RankBrain can certainly help you do keyword research and write relevant content better.

[seoanalytics]

I totally agree, thank you for your detailed explanation.