I read a lot of patents, many of which may or may not apply to search engine optimization (SEO) or be used by Google at all. However, that’s not the case with the recently granted Google patent “Related entities.” I believe this patent is in use and offers significant insight into how Google identifies entities and the related entities people are searching for.
Let’s delve into some details I find interesting to understand the patent and its intent. Understanding how Google associates entities will help us integrate these connections into SEO.
### Related Entities
Let’s begin with an understanding of related entities, looking specifically at Google patent US 20180046717A1.
If you search the phrase “presidents of the united states,” here’s what you might encounter: The presidents presented are “related entities” listed because the general phrase “presidents of the united states” was searched. Different people are shown, but they all share a common denominator: having been President of the United States.
How does Google decide to show these particular people when a general phrase is queried? The patent explains this by discussing the selection and display of these related entities.
Let’s consider another example. Clicking on an image of Donald Trump on the page leads to a query for his name, appearing as: When searching his name without prior searches related to Presidents, this is the display you might see. The breadcrumb navigation at the top of the results, appearing since February 2018, shows how context carries through searches.
Searching for presidents presented a chronological carousel, with the context carrying through when you click an image—something absent when you search a president in isolation. This is key understanding how the patent operates.
### Entity Database
One significant takeaway is the idea of an entity database—distinct from the general search index, tasked with understanding various internet entities, their attributes, and interconnections. In this context, entities are more than people, places, or things, but also their characteristics connected via relationships. In patent discussions, entities are “nodes” and relationships are “edges.” Prominent entities and relationships involving Barack Obama include:
– Name: Barack Obama
– Position: President of the United States
– Birthplace: Honolulu, Hawaii
– Spouse: Michelle Obama
– Net worth: $12.1 million
These examples illustrate the separation of this database from the general search index.
### Determining Relatedness
The patent addresses determining relatedness, applicable beyond just voice search optimization. Google identifies relatedness primarily through the co-occurrence of entities in resources. In the presidential example, various presidents appear on the same pages, signaling relatedness to Google.
Entities in a carousel are grouped based on frequent co-occurrence or shared connection with the phrase “president of the united states,” despite infrequent common page appearances. This concept holds importance for content marketing and general SEO outside the examined patent.
### Determining Priority
Although this aspect of the patent is less applicable to general SEO, it’s still noteworthy. Google needs a mechanism to prioritize entities and relationships—for instance, highlighting Donald Trump’s presidential identity over his career as a businessman.
Similarly, Ronald Reagan is more prominently identified with his political career than his longer tenure as an actor. Topical relevance and click-through rates on related queries, combined with subsequent user queries, assist Google in determining priority.
### Overarching Factor
Authoritative sites related to specific subjects are prioritized in determining entity relationships. A Wikipedia page on Ronald Reagan’s presidency would strengthen the relationship between his name and “president.” Similarly, authoritative SEO content from companies in the industry enhances such relationships in technical SEO contexts.
This element functions like PageRank for entities, although without a visible rank indicator.
### SEO Implications
The patent has broad implications for SEO beyond displaying related search options. The concept of an entity database distinct from the general search index echoes in multiple Google patents. This database surpasses a mere record of web-based links and anchors by comprehending entity relationships.
For instance, a New York City hotel referenced on pages optimized for “hotel” strengthens the brand’s relationship with “hotel.” Pages optimized for “New York City” further reinforce this relationship, even absent direct links, enhancing relevance scores.
Entities appearing alongside other related entities enhance their association, much like piggybacking on already established relationships.
### Competing Brands
Continuing with the hotel example, mentioning a hotel alongside competing brands could boost the hotel-related strength. However, if the content also references dining and activities, the hotel relationship could weaken.
Entity association might not be binary or directly proportional; proximity and topical consistency could influence association strength. This mirrors PageRank logic, though here the presence of a link doesn’t diminish authority’s importance or influence.
While PageRank and links remain vital, the concepts here refer to relationship-building, unaffected by attributes like nofollow.
Wikipedia’s use of nofollow links exemplifies how content and structure enable strong signals that transcend nofollow limitations.
### Takeaways
This patent offers strategies for strengthening site or brand associations with specific terms and entities. It suggests opportunities to advance rankings through entity associations, enriching traditional link strategies.
While linking remains a priority, combining entity and link strategies yields dual benefits. Even on pages unrelated to search terms, focusing efforts where topical or geographic relevance aligns with desired entity attributes still constitutes sound marketing.
