This Week In SEO 29
Google Machine Learning, New Schema, Pinterest Visual Search
Google Open Sources Machine Learning System
Last week we brought you some info about RankBrain, an artificial intelligence algorithm that Google uses to process never before searched queries, among other things. This week, Google announced the open-sourcing of TensorFlow, a machine learning system.
Just what does that mean? Here’s an explanation from the Google blog:
So we’ve built an entirely new machine learning system, which we call “TensorFlow.” TensorFlow is faster, smarter, and more flexible than our old system, so it can be adapted much more easily to new products and research. It’s a highly scalable machine learning system—it can run on a single smartphone or across thousands of computers in datacenters. We use TensorFlow for everything from speech recognition in the Google app, to Smart Reply in Inbox, to search in Google Photos. It allows us to build and train neural nets up to five times faster than our first-generation system, so we can use it to improve our products much more quickly.
Google seems to be going hard on machine learning/artificial intelligence lately. This should serve to reinforce the increasing importance of voice search and casual queries in search, and how long, high-quality content that takes advantage of related words and phrases will continue to be essential.
Here’s one example of how machine learning relates to search engine queries:
Machine learning is still in its infancy—computers today still can’t do what a 4-year-old can do effortlessly, like knowing the name of a dinosaur after seeing only a couple examples, or understanding that “I saw the Grand Canyon flying to Chicago” doesn’t mean the canyon is hurtling over the city.
TensorFlow Video from Google:
What’s New at Schema.org
From local businesses to big ecommerce stores, schema mark-up is an essential part of helping search engines structure the information on your site. We are huge proponents of using Schema markups on your websites for improved rankings. Since that’s the case, keeping a close eye on new improvements to Schema.org is essential. Here’s are the highlights:
- We made a number of improvements relating to the description of services, including the addition of providerMobility to indicate dynamic locations, OfferCatalog for hierarchical collections of offers, as well as introduced the notion of a GeoCircle to make it possible to describe service availability in terms of distance from a point or postcode.
- A new type: ExhibitionEvent for describing exhibitions (e.g. in museums, galleries), alongside a property workFeatured that indicates a CreativeWork featured in an Event. This is quite a typical schema.org change: it generalizes existing vocabulary – workPerformed, workPresented – to cover more scenarios with less terminology.
- Added an inverse of the makesOffer property: offeredBy to simplify the description of not-for-profit offers (e.g. library book lending).
- Improved our support for feed-oriented structured data, by adding DataFeed and DataFeedItem.
- Introduced a new type to represent barcodes.
Pinterest Introduces Visual Search
Google is, not surprisingly, at the center of a lot of our discussions here on SupremacySEO.com. But search engine optimization has value beyond Google. Pinterest is especially of interest to ecommerce sites, and recently, they introduced a big new feature: visual search.
When you spot something in a Pin that you want to learn more about, tap the search tool in the corner. Then select the part of the Pin you’re interested in, and we’ll show you Pins just like it. You can even filter your visual search results by topic so you find exactly what you’re looking for.
So if you zoom in on that lamp, you can discover what it’s called (“Antiqued Metal Funnel Pendant”), and where you can find it (Restoration Hardware). If you want to know more—like how to get a table like that—just resize your selection and move it around to instantly see more Pins.
Is Yahoo Really Trying to Beat Google
A little bit of an inside look at what Yahoo! is working on as it relates to their search engine results — or at least, the Yahoo! homepage. A former Googler, Muñoz is working with Mayer and Yahoo! to, according to the article, grab a chunk of search real estate from Google.
…experts in the field think he may actually have found an opening here, an Achilles’ heel in Google’s crazy dominance — and it’s in the mobile, rather than desktop, space. So far, Yahoo spokespeople say they’ve got 600 million–plus mobile users, and growing. “It’s an uphill battle,” in general, says Anindya Ghose, professor at NYU Stern School of Business, but that’s because no one’s cornered mobile yet, not even Google.
This story, coming so quickly after the recent announcement of the Yahoo!/Google search partnership seems odd, but maybe it’s not hard to believe Yahoo! is not looking to be left hanging on to whatever scraps Google throws their way. In any case, it will be interesting to see how this story unfolds.
The Price of Changing Domain Names
This is a great piece on the risks involved with moving from one domain to another (specifically for an established site with a lot of pages and authority).
Both the domain name change for Theguardian.com and Dunelm.com make it clear that Google may not immediately transmit all the quality signals from the old domain name to the new. It may come as a surprise that Google actually needs some time to move all the important quality signals, such as authority, reputation, trust, user signals and the internal PageRank budget for crawling and indexing, between the two domains. This is, in part, due to the fact that Google has to crawl all internal links and pages on the old website once again to identify all redirects, in order to transfer the quality signals to the new domain.
Check out the whole case study, there’s a lot of good data and some more interesting insights.