|SEO & Named Entities; what can we learn?|
|Written by David Harry|
|Tuesday, 17 January 2012 13:52|
In the world of search entities and their associations have become more and more important over the years and it's an area of study I advise all SEOs to get into. Today, more than ever.
Last week we looked at some of the what, where and how of named entities and their applications to the modern world of search. As promised we're now going to dig in a little further and see what this means for SEO and how it can be incorporated into your thinking.
Part II - Where do entities fit into SEO?
It was certainly an interesting week for this topic. Not only did I end up writing about it, but there were also posts from Bill and Justin Briggs on it as well, (links at the end). What gives with that?
And it didn't end there. The very same day the first part in this series was published, Google rolled out it's new Search Plus Your World features. What's so special about that? It's all about entities. People, the social graph and their relations with the web, (where they go, what they talk about and who they're connected with).
Entities are everywhere
Once you've wrapped your head around the concepts, it becomes apparent that these damned things are everywhere. Let's look at just a few...
On and on it goes. You see what I mean here? It's an overlying concept that is pervasive throughout the information retrieval process. Now start to wrap your head around how pages can relate to each other and the end user.
Play around in some of your own query spaces and watch for the named entities. In some query spaces they're prominent, while in others... not so much. Know thine query space.
Strong entity associations; the new keyword research
Earlier I juked past the topic of query classification, so let's go back there. These are better known in SEO circles as query types; informational, transactional and navigational. In each area you will find different types of entities that are relative to the space. Find them.
As part of the process you can start to collect entity associations as you would with modifiers. This is somewhat reminiscent of collecting semantic baskets of terms to support main concepts (in your content). Except we're after entity associations.
In the 5 minutes it took me to snap those screens, it wouldn't take much to take those and craft a piece of content that carries a ton on entity associations with the query space we're targeting.
Keyword research is focused on matching user intent with our (ranking) targets. Not only are the named elements of your target terms in play, but the actual demographic of the user/consumer that you're seeking to attract as well.
If we are to consider the social graph, (more on this in a bit) and it's potential effect on search, then keyword research must be inclusive of demographics to understand the user entity profile. Don't even get me started on the analytics side of that, have more than a few new custom reports since this journey began. This is why it pays to get your head around it.
The concepts can be used to deepen your existing keyword targeting efforts. It's not a replacement.
Another aspect to all of this are links. If you want your brand or offering to be synonymous with a query space, then building associations. Is a link from the NYT gold? Sure. Why? Because of not only the equity (PageRank), but the authority as well, (trust).
This can be the same for non-link citations. Having your entity mentioned along side of other common known named entities, you have now an association. We've seen the power of citations for local and there's certainly potential for this kind of signal to grow in value moving forward.
Also consider the social graph. The user entity aspect of following other user types and sharing of content. It is not the link itself now that becomes the direct value, (most are nofollowed anyway, thus inhibiting PageRank flow). The sharing, (an implicit link of sorts) becomes the vote.
Links themselves are also still great given that regardless of state (nofollowed or clean) they can still pass a signal. Have you linked up all your entities in your own personal social graph? Allow the search engine to get a clear view of your associations. Link them up!
Rethink your link building. Seek metaphoric links not just the href variety.
Taming the Verticals
Another area that would also come into play are the various verticals that make up the universal search elements in a given query space. These can include;
How can your efforts tie into these? Are you building a front-facing staffer? Promoting an event? Seeking to be an authority in a space? Publishing timely topical matter?
Looking at your related target terms for universal elements can often lead to more opportunities. The more associations you build, the greater the over-all body of work will be. Taking a SERP out approach and establishing entity presence is certainly something to consider.
Next time someone says they don't 'need' to get into Google places. Don't need to create a video. See no point in a PR stunt. You have a new weapon.
Entities and the Social Graph
Seriously... the title says it all. We can't talk about entities in search these days without considering social media/networking/bookmarking. People are walking, living breathing entities. Not the inanimate variety. Given ol Googly's never ending desire to play with these personalization, it's a safe bet this becomes more and more important.
What I have been looking at, via analytics, are behavioural data sets. Segment the social traffic, add some engagement metrics and performance ones, get a sense of the user types form various platforms. Then create a (not provided) report to compare (more on that in a bit).
Google will most likely still be grouping user types. This type of entity grouping will create natural patterns for users. If you're using actual staffers or persona as forward-facing entities, be sure to once more take care in considering the associations.
Right away you want to start crafting a social graph that is inclusive of the demographic considerations we talked about earlier. To get a sense of how Google might view your connections, play around with the Social Graph API examples;
What we want to do here is ensure that the social graph we're giving the engnie has a well thought out entity association tied to it.
The (not provided) connection
I mentioned the tracking and analysis of (not provided) data. For more on the concept, please see last week's post; the Hidden value of Google (not provided) for SEO
The short and sweet of it is that this data point means visitors are logged into Google. If they are logged in, there is a good chance they're seeing personalized results. There is also the possibility for socially annotated and re-ranked results.
Given how social graph entities are changing the SERP landscape, seeking to analyze this data and establish potential user types is key (compare data against known keyword data and segmented social site referrer data).
All of that of course can play into how you're crafting the social graph. The ultimate entity is the user. Entity construct and the social graph has become part of the SEO domain. Because social entities now litter the SERPs, (I knew I could become a social media guru someday if I tried REAL hard!!).
Attack of the brands!
One area I touched on last time was the ever-popular scratching post that is the wrath against Google from SEOs about them treating brands preferentially. For the record, I have never seen a patent nor paper that says, “Give X site increased scoring”. The reason this isn't sensible is that brand popularity fluctuates from region to region and over time. That's not a scalable algorithm.
What you ARE seeing is the treatment of entities. Based on user activity (queries, query refinements, click through rates, bounce rates etc..) the search engine starts to learn which entities are more sought after in various query spaces. It stands to reason Walmart is going to be searched more than your local mom and pop. And if you happen to have a Walmart in your town, guess who's likely to show up in related searches? Not the brand, but the entity.
We must not be swayed changing tides of search and pander to the common ignorance. Seek to better understand the search engines, why it does what it does, then adapt. I hope these two posts helped you in that direction.
Note; I wanted to also get into RDFa and semantic data, but again the post grew out of control. We did get into authors and authority last time, it'll have to do. Obviouly author markup is important. More on this stuff soon...
Posts from last week;
Entity Search Results: The On-going Evolution of Search - Justin Briggs
(see? Linking out is great for associations.... ;)
|Last Updated on Tuesday, 17 January 2012 14:19|
Home - all the latest on SNC
SEO - our collection of SEO articles
Technical SEO - for the geeks
Latest News - latest news in search
Analytics - measure up and convert
RSS Rack - feeds from around the industry
Search - looking for something specific?
Authors - Author Login
SEO Training - Our sister site
Contact Us - get in touch with SNC