Difference hummingbird and rankbrain
-
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,
-
I totally agree, thank you for your detailed explanation.
-
You can't really "optimize" for Hummingbird, but understanding RankBrain can certainly help you do keyword research and write relevant content better.
-
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.
-
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,
-
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.
-
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.
-
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
-
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.
Got a burning SEO question?
Subscribe to Moz Pro to gain full access to Q&A, answer questions, and ask your own.
Browse Questions
Explore more categories
-
Moz Tools
Chat with the community about the Moz tools.
-
SEO Tactics
Discuss the SEO process with fellow marketers
-
Community
Discuss industry events, jobs, and news!
-
Digital Marketing
Chat about tactics outside of SEO
-
Research & Trends
Dive into research and trends in the search industry.
-
Support
Connect on product support and feature requests.
Related Questions
-
Difference keyword and co-occurence
Could someone explain me what the difference between a keyword and a co-occurence is ? Thank you,
Intermediate & Advanced SEO | | seoanalytics0 -
Our Web Site Is candere.com. Its PA and back link status are different for https://www.candere.com, http://www.candere.com, https://candere.com, and http://candere.com. Recently, we have completely move from http to https.
How can we fix it, so that we may mot lose ranking and authority.
Intermediate & Advanced SEO | | Dhananjayukumar0 -
Google indexing only 1 page out of 2 similar pages made for different cities
We have created two category pages, in which we are showing products which could be delivered in separate cities. Both pages are related to cake delivery in that city. But out of these two category pages only 1 got indexed in google and other has not. Its been around 1 month but still only Bangalore category page got indexed. We have submitted sitemap and google is not giving any crawl error. We have also submitted for indexing from "Fetch as google" option in webmasters. www.winni.in/c/4/cakes (Indexed - Bangalore page - http://www.winni.in/sitemap/sitemap_blr_cakes.xml) 2. http://www.winni.in/hyderabad/cakes/c/4 (Not indexed - Hyderabad page - http://www.winni.in/sitemap/sitemap_hyd_cakes.xml) I tried searching for "hyderabad site:www.winni.in" in google but there also http://www.winni.in/hyderabad/cakes/c/4 this link is not coming, instead of this only www.winni.in/c/4/cakes is coming. Can anyone please let me know what could be the possible issue with this?
Intermediate & Advanced SEO | | abhihan0 -
Duplicating a site on 2 different ccTLDs and using cannonical
Hello, We have a site that sells a certain product on www.example.com. This site contains thousands of pages including a whole section of well written content that we invested a lot of money in making. The site ranks on many KWs both brand and non-brand related. SERPs include the Homepage and many of the articles mentioned. We receive traffic and clients to this site from around the world, BUT our main geo-targeting is UK. Due to lack of resources and some legal needs we now have to create a new site - www.example.co.uk that all UK traffic will be able to purchase the product only from this site and not from the .com site anymore. We have no resources to create new content for the new .co.uk site and that is the reason we want to duplicate the site on both domains and use a canonical tag to point the .co.uk site as the primary site. Does anyone have experience with such activity? will this work across the whole site? We need to have a fast solution here, as we do not have too much time to wait because of the legal issue I mentioned. What is the best solutions you can offer to do this so we do not lose important SERPs. On the one hand since our main market is the UK, we assume the main site to promote will be www.example.co.uk but as said earlier, we still have users from other parts of the world as well. Is there any risk that we are missing here? Thanks James
Intermediate & Advanced SEO | | Tit0 -
Is a different location in page title, h1 title, and meta description enough to avoid Duplicate Content concern?
I have a dynamic website which will have location-based internal pages that will have a <title>and <h1> title, and meta description tag that will include the subregion of a city. Each page also will have an 'info' section describing the generic product/service offered which will also include the name of the subregion. The 'specific product/service content will be dynamic but in some cases will be almost identical--ie subregion A may sometimes have the same specific content result as subregion B. Will the difference of just the location put in each of the above tags be enough for me to avoid a Duplicate Content concern?</p></title>
Intermediate & Advanced SEO | | couponguy0 -
Canonicals: use when page has same listings, but displayed very differently?
Say you have a listing of movies. In that listing, there are 5 different view types. One has the scenes broken out. Another has only the box covers. Two of the views have movie descriptions, but others don't. Still, the listings themselves are the same, and you only want the default view to be indexed. Is it appropriate to use canonicals in this case? The alternative is to noindex the other views, but the site already has rankings and deep links. If Google does see the pages as unique and we apply a canonical, could we be penalized or would they merely ignore it?
Intermediate & Advanced SEO | | LahomaManagement0 -
Optimal ratio of different categories of link
Hello, I know this probably depends on market sector, etc., but are there broad guidelines on the optimal ratio of different types of links - blogs, directory links, citations, etc. I'm working for a tourism business and see nearly all of their links are from directories - both generic and tourism specific. I'm thinking about re-balancing that with guest blogs, PR work, etc.
Intermediate & Advanced SEO | | McTaggart0 -
How Come Meta is different based on different query?
We have a site we added a number to in the meta description. Once we did that we did a fetch as google to hopefully recrawl the page quicker. A few days later and we cleared W3 cache on WP and clear computer cache, then did search on common search for the site/page. WidgetA for example. The url is OurClient.com/widgetA/ - on organic in meta on SERP and we see our new meta with number. We then do a search on a similar term WidgetingA for example and the same url shows: OurClient.com/widgetA/ BUT THE meta description is different on SERP! It is the old meta. When we look at the element using mozbar, it shows the new meta as it should same as when we look at it under the original search term. So, search for WidgetA, get new meta in serps and search for WidgetingA (which returns same url as WidgetA) and we get the old meta. Thoughts???
Intermediate & Advanced SEO | | RobertFisher0