Smart Analytics Reporting: In-Depth, Names-Based

Names-based reporting uses the automatically identified names of each content published in order to provide an in depth statistical collection of your website visitors readership habits and your editorial team’s publishing ones.

Main Purpose

Knowing that the majority of subjects of interest people read about can be boiled down or related to Names of things, this dashboard provides you with the ability to dissect your views and published content based on these names to identify:

• Main types of names of interest. For example: political parties, political figure, political events.
• Main Names of interest. For example: Donald Trump, iPhone, Games of Throne, Ebola virus, etc
• Outliers during a period of time. For example: there could be certain articles that received huge number of views and that do not belong to a particular type of news you usually publish about. One example is evergreen type of articles, or even an article about a historical figure. Such examples provide new insights you never thought you should consider. Ex: Elvis Presley, World War II, a particular homicide victim or criminal, etc

Please check the last section: “Sample Questions Magnet Smart Analytics can help your team answer” for further examples of how you can use this interface.

Types of Names

Magnet identifies 45 top-level types of Names, 100s of secondary level types, and 1000s of leaf-level types of Names. For example: People > Athlete > Basketball Player > Shooting Guard.

The main types of Names you can filter the reports by are listed in the below table:

Figure 1: Type of Names

Interface

Figure 2: Names-based Reporting - Interface

As illustrated in Figure 2, the “Entities” or Names interface allows publishers to retrieve reports based on the following filters:

• Period: a date period for which the data will be retrieved.
• Published vs. Consumed vs. Avg. Page Views:
  1. Published refers to all published content. Selecting this checkbox will retrieve the publishing statistics for the selected period based on the Publishing category(ies) of each content piece.
  2. Consumed refers to every pageview recorded. Selecting this checkbox will retrieve the pageview statistics for the selected period based on the publishing category(ies) of each content piece.
  3. Page Views refers to the average pageviews each content piece generated per category.
• Category: this dropdown list allows you to choose between: Politics, Sports, Law Enforcement, Business, Entertainment, Health, Technology, Science, and General.
• Referral Channel: this filter allows you to choose one or more referral channels (sources) from which the visitors arrived to your website. For example: filter the report to show data from views that reached the website from Facebook Mobile or Google.com.
• Device: this filter allows you to choose one or more device types (Desktop, Mobile, Tablet, other) from which the visitors accessed your website. For example: filter the report to show data from views that reached the website from Desktop Views only.
• “Count Docs Where Entity is:” this filter allows you to select the importance of the Name within context of the articles to be counted in the resulting report. The importance is scored as: VR (Very Related), R (Related), and SR (Somewhat Related). For example, you can select to view a report where “Companies” (Names type) were “VR” only + other filters.
• Min Articles Count: this filter allows you to define the minimum number of content pieces the Name should have appeared in before being accounted for in the report. For example, if you selected to see “Companies” with Min Articles Count = 3 + other filters, then the resulting report would not account for “Tesla Motors” if this Name was not featured in at least 3 articles.
• Language: this filter is used when you publish in more than one language. Selecting one or more language will filter-out all others.
• Entity Type: lists the Names type you can select from. You can select one or more of course.

While the unit of the report can be in:

1. Percentage (%): percentage of total content published (when Published is selected) or percentage of total pageviews viewed (when Consumed is selected).
2. Numbers: actual number of content published (when Published is selected) or pageviews viewed (when Consumed is selected).

As for the sorting of the resulting charts, the interface allows sorting by: Published, Consumed, or none depending on the other selected criteria.

Figure 6: Sample Names-based report showing Top 20 Published vs. Consumed People Names based on Pageviews generated

Interested in learning more about Magnet’s Smart Analytics tool or have a question?
Please contact our team at [smart at klangoo.com].