Google Knowledge Graph
What is the Knowledge Graph made up of?
In 2012, Google added an expansion of search results with a widget on entities in the Knowledge Graph. In the Knowledge Graph people, places or the circumstances of various facts, images and associated search queries are displayed in a dedicated area if Google recognizes or suspects a semantic entity behind a search term. The information leveraged by the Knowledge Graph used to partially come from Freebase, whereas today it is mainly gathered from Wikidata, enriched with enhancing links or related search terms.
On desktops and tablets the Knowledge Graph is displayed on the right next to the organic search results; on smartphones the display is above the search results. The following elements can be included in the separate info box:
- Data on a company, a person or a location
- Image(s) with a link to Google Image Search
- Text extract as well as a link to the source
- Prepared, structured data with details on the search query
- Information on similar search queries “others also searched for”
Which sources does the Knowledge Graph use?
Google either uses serious, publicly-accessible data or data from its own inventory for the display of the Knowledge Graph. Current sources, alongside Wikidata, also include the USDA (US Department of Agriculture).
The Google Knowledge Graph also uses content from websites, which the company trusts. This is above all the case for definitions. Furthermore, there must be structured data for certain search queries (rich snippets), which are identified with markups. It is only in this way that this information can be read out and displayed by Google.
When is the Google Knowledge Graph displayed in the search results?
The display of the Knowledge Graph has also been triggered by Google with precise questions since the introduction of the Hummingbird algorithm and can be displayed both in desktop search or mobile search. Depending on the search term or question, there are various triggers for the display of the Google Knowledge Graph. Ten possible search queries that trigger the inclusion of a Knowledge Graph in the search results are:
- Questions on famous people. Example: Who is the president of Germany?
- Questions on companies and their people. Example: Who is the CEO of Apple?
- Questions on relatives of famous people. Example: What is the name of Barack Obama’s wife?
- Questions on recipes in mobile Google search. Example: Apple pie recipes
- Questions on artistic works such as movies or novels. Example: Which books has Paul Auster written? or Who is the director of Star Wars?
- Questions on food. Example: Nutritional value of oranges.
- Questions on distances. Example: How many kilometers is it from New York to San Francisco? In this case Google uses its Google Maps service as a source.
- Conversion of measurement units or calculator.
Example: How many yards make up a mile?
- Questions on celestial bodies. Example: How big is Mars?
- Definitions of a topic. Example: What is a search engine?
The Searchmetrics study on universal & extended search has found that a Knowledge Graph is found in eight percent of all desktop search results of Google.de; the share in mobile search results was 14 percent of all search results.
What effects does the Knowledge Graph have on SEO?
The Knowledge Graph has both advantages and disadvantages to websites and search engine optimization (SEO). As users usually do not require a further click to find the desired information, websites with a high information share can lose traffic in general search queries. This traffic loss can also result in losses in advertising revenue. However, the opportunity to be listed by Google in a Knowledge Graph with a text or image link generally seen to as a positive.
Content should, for example, be given markup language to increase the likelihood of being displayed in a Knowledge Graph. An opportunity to label data is offered by schema.org. This data must also be correct and up-to-date. Note that there is no guarantee that your website will be displayed on the graph. Google’s algorithm decides on the display. Google itself has not given any further information on this.
Overview of Schema.org:
Google Inside Search via the Knowledge Graph: