The purpose of the Web Platforms Tracker is to highlight the share of Web use for Tablets, Smartphones, and Desktop and the operating systems powering them. To determine the respective share of different personal computing devices, Chitika Insights performed an analysis of the unique impressions seen originating from Tablets, Smartphones, and Desktops within the Chitika Ad network. To determine the usage share of each device, a user agent analysis is conducted across incoming impressions. The following paragraphs further detail the methodology behind the Web Platforms Tracker published by Chitika Insights:
Impressions:
Sample: This study was conducted across a full range of hundreds of millions of impressions from the Chitika Ad Network; coming out of the US and Canada.
Determination: The operating system of an impression can be determined through analysis of the user agent. All user agents of personal computing devices pass a string similar to this (in this case, we are using an Android impression):
Mozilla/5.0 (Linux; U; Android 2.2.1; en_gb) AppleWebKit/533.1 (KHTML, like Gecko) Version/4.0 Mobile Ninesky-android-mobile/2.2.0 Safari/533.1
Reporting:
Top Pie Chart: The percentages reported in the top pie chart represent an aggregate share of Web traffic, with a six hour delay to ensure maximum continuity within the data. These percentages are simply the count of each particular group (Tablets, Smartphones, and Desktops) divided by a total count of all Web traffic observed.
Bottom Pie Charts: The percentages reported in the bottom pie charts represent an aggregate share of Web traffic, with a six hour delay to ensure maximum continuity within the data. These percentages are simply the count of each particular operating system per device (Tablets, Smartphones, and Desktops), dividing by a total count of all Web traffic observed on the respective devices.
Line Chart: The line chart highlights the use of Tablets, Smartphones, and Desktops on an hourly basis. For each hour, the methodology is the same as in the pie chart, except shown using a time series. For example, if Smartphone impressions made up 10% of all mobile impressions on June 20th at midnight, and 11% at June 21st at midnight, the chart would highlight this growth (and show all of the hours in between).