Future SEO: String Entity Optimization

Imagine the future of SEO — a future in which you forget about using keywords or their synonyms multiple times on a page. In the future, this will be obsolete. Search engines of the future will provide users with answers to their queries by internally verifying validated data that link to trusted documents. To optimize […]

Chat with SearchBot

Imagine the future of SEO — a future in which you forget about using keywords or their synonyms multiple times on a page. In the future, this will be obsolete. Search engines of the future will provide users with answers to their queries by internally verifying validated data that link to trusted documents.

killer seo new1To optimize websites for search in the future, SEOs will need to create relevant, machine-recognizable “entities” on webpages that answer well-refined, focused or narrowed queries.

To create these entities, SEOs will use semantic Web technology and structured data. This allows search engines to better understand the page content and thus display valid search results/answers for each query.

What follows is the why and how of future search engine optimization. But first, let me digress.

Discussing The Future of Search & SEO

While discussing the future of Search and SEO last week in view of recent Google innovations such as Google Glass and Google Now, Barbara Starr, a good friend and business associate, agreed to share her insights in this article interview.

Barbara’s history as a semantic strategist is quite amazing. Earlier in her career, she worked on the High Performance Knowledge Bases (HPKB) project — a DARPA research program to advance the technology of how computers acquire, represent and manipulate knowledge. Barbara also worked with DARPA on the PAL program (which evolved into SIRI) as well as many other projects of that nature.

When I met Barbara in 2009, we started talking about how SEO ties into search engine technology. Over time, this led to a discussion of the historical facts and future possibilities of Search and the Semantic Web. This progressed into recent discussions around the concept of why we use keywords in search engines, which led to the concept of entity search and string entity optimization.

Google and other search engines are highly sensitive to their IP, signals and algorithms; therefore, it’s always been a bit of a guessing game for SEOs to anticipate what’s coming next in Search. Today, I’m going share a very qualified ideation Barbara and I have been brainstorming. Granted, we can’t predict what Google will do in the future; however, there are fundamental concepts about search and semantic technology currently being deployed by Google and other search engines that lay out some viable possibilities for consideration.

Why Do We Use Keywords In Search?

Paul: Barbara, why do you think we‘ve been using keywords in Search for so long?

Barbara: Keywords are an artifact that evolved during the history of search. They exist due to inefficient input mechanisms like keyboards, as voice search or other mechanisms such as gesture search had not yet matured. There’s much more work to do in perfecting these behavioral results from machines.

I believe it is part and parcel that keyword-only queries will ultimately virtually die out. They existed because it was a pain to type the entire query out in full, and there was no effective technology for voice recognition, touch screens, etc. However, at this point, search engines prefer full sentences or meaningful phrases as they give more context and information about user intent in a query (directly from the “explicit” component of the query). I would actually break down “implicit” to several different subcategories based on “situation,” such as device, geo-location, etc.

Moore’s law also comes into play on the hardware side. We did not have the raw compute power and Search Engines still needed to be aware of performance and more, so we ended up with this weird keyword system as it was better than anything else at the time, not to mention Google was great at marketing and monetizing keywords!

delivery search

Search [delivery]

No one in their wildest dream would think of an “answer engine” answering “keywords.” Without any context and out of nowhere, if you walked up to someone and said, “delivery” — they might think you were nuts. However, within existing Search Engines today, if you give it context such as search history and geo-location, in Google, you will get a restaurant carousel that delivers food-related results, and if you do not find that relevant, it will activate a second algorithm that provides e-commerce type delivery results, also using its Knowledge Graph entities, whereby Google derives the “context itself.”

Paul: Do you see a day where keywords are less emphasized?

Barbara: For SEOs, the paradigm of not relying so heavily on keywords is very disruptive; however, not so much for someone with my background, who has spent 10 years building answer engines where many searches would have queries which timed-out after an hour.

As a great example, think of search engines like Sindice.com. It crawls the pages on the Web, extracts the structured data metadata from the HTML, and throws away the rest of the page, keeping the structured data internally to search directly on it as an “entity search.” Therefore, keywords on the page in entity search, in this case, are beyond irrelevant. They are literally discarded by the search engine!

Google is using its other Signals to determine trust, to verify and validate its own internal entity graph, while a huge and natural focus on the Semantic side is looking at “Signals” from both the perspective of understanding the user’s intent, and also at “Signals” for returning relevant results for a particular user in a specific situation. If you are not defined as an entity, clearly you cannot be found in an entity search. We could really expand on the entire “situation” aspect here, and also the “social aspect,” which is key with Google Plus.

Entity Search & String Entity Optimization

Paul: These are concepts SEOs may struggle with, can you explain how SEO fits within “entity search”?

Barbara: This obviously applies to SEO when you begin to realize that Google’s Knowledge Graph is an entity graph. Therefore, if you are not defined as an “entity” in Google’s terms, you do not exist in their Knowledge Graph (or carousel), and you can certainly never become a result as the answer to a query if you do not exist in their Knowledge Graph, e.g., Google Now and/or Google Glass. You simply do not exist from the context and perspective of the machine.

So, as I mentioned, you have to make sure your brands/clients, etc., exist as entities for Google, Bing and other emerging Semantic Answer Engines, wherever you want them to be found.

SEOs can get an idea of entity ranking from Trending Entities, an initial indication of Google’s shift to entities in terms of metrics, which we may begin to see as new semantic Web tools are created.

google trends

Example Of Trending Entities

Another small example of early metrics is Structured Data results in Google Webmaster Tools, which is still in its nascent phases compared to what we could see in the future — just as the Knowledge Graph is still in its infancy as well.

Paul: What changes do you expect to see in search users and engines over the next 3 years?

Barbara: Google and the other search engines are migrating toward their goal of the Star Trek computer of the future. Search will become increasingly semantic and leverage many concepts innately thought of as relating primarily to artificial intelligence.

Paul: What considerations will be involved with search engines and SEO?

Barbara: A stronger semantic shift and a move to utilizing structured data, both in search results and in user queries.

  • With the effects of voice search and the enormous improvements in this direction, queries are going to be composed less and less of keywords and will become full sentences of questions.
  • A focus on understanding the query to better display valid search results or answers.
  • The release of many devices, perhaps wearable or possibly even one day embeddable; that can access search from wearable items such as Google Glass to other wearable items such as a watch.
  • Choice of the ultimate “device” to access all your displays, such as a home entertainment system, TV, etc. Search across all of these categories, like music, TV channels, available movies, etc., initiated presumably by your device of choice.
  • Closer interaction with the Internet of things as signals being “pushed” to user devices for Search Engines.
  • Leveraging Social Profiles such as Google Plus, a much deeper discussion.

Paul: What do you think is coming next in String Entity Optimization (i.e., Semantic SEO)?

Barbara: This brings us back to the move toward the Knowledge Graph, entity searches and Google’s associated shift from “indexing” to “understanding.”

Currently, entity search is being leveraged by the majority of search and social engines; therefore, understanding how user intent is leveraged and how this relates to entity graph searches is key. If you are not defined as an entity, clearly you cannot be found in an entity search.

Paul: What can we do as SEOs to improve our code before the engines move full force in this direction?

Barbara: Questions for the SEO then become: how can I create relevant entities on my pages that answer a well-refined query that is focused/narrowed to seeking specific results/entities as an answer.

The mapping of strings to things needs to be understood by SEOs.

The understanding that an entity found and displayed as a result and coming from structured data sources within the search engine, defined in the Schema of the entity is a stronger movement. So, stop worrying about keywords on a page. Worry about what the page “says” to a human and how it relates to the entities or “things” that search engines are interested in. The engines are interested in those “things” for a reason.

Remember my example, Sindice.com. It crawls the pages on the Web, extracts the structured data metadata from the HTML, and throws the rest of the page away, keeping the structured data internally to search directly on it (entity search).

Again, forget about using a keyword or synonym multiple times on a page; it’s useless. Once the engine has internally verified validated data that can link to documents providing the user with answers to their query, it will use that as a trusted and proven source for answers.

Paul: Thanks for sharing your insight with us, Barbara. As always, it appears change is on the horizon, and understanding semantic markup is becoming more and more a best practice for SEOs. For more elaboration on the new SEO: String Entity Optimization, Barbara and I may collaborate on a column in the future. In the meantime, see the semantic SEO resources below for more information.

Semantic SEO Resources

Google Blog: Introducing the Knowledge Graph

Search Engine Land: Used To Searching For Content? Content Searches For You

Aaron Bradley’s Blog: SEO Skeptic

David Amerland’s book: Google Semantic Search

Semantic Web Technology and Linked Data: SemanticWeb.com

Linked Data: LOD Cloud

Moz Blog: From Keywords to Contexts, The New Query Model

Google+ Community: Semantic Search Marketing


Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.


About the author

Paul Bruemmer
Contributor
Paul Bruemmer is Managing Partner at PB Communications LLC. Specializing in SaaS solutions for Enterprise Store Locator/Finders, Semantic/Organic/Local/Mobile and SEO Diagnostic Audits for increasing online and in-store foot traffic.

Get the must-read newsletter for search marketers.