DeviceAtlas Data Explorer FAQs

DeviceAtlas Data Explorer FAQs

Welcome to the DeviceAtlas Data Explorer; we hope you find this a useful resource. We recommend reading these FAQs to assist in understanding the data shown, and if you have any further questions, don’t hesitate to contact support@deviceatlas.com.

There are two main sections of the Data Explorer; one section permits analysis of the attributes of the different devices within the scope of DeviceAtlas, and the other provides a perspective on the relative degree of usage of the devices on the web.

Important note:

- What the data represents is best described as the relative propensity of a non-desktop based browser (for example mobile phones and tablets) to be used on the web. This does not measure the device distribution or popularity, and there are three main sources of distortion, for which no attempt to compensate has been made.

  • Phones without browsers, and low end devices which are typically not used on the web, are not represented at all.
  • High end devices are over-emphasised relative to the overall mobile device distribution, for two principal reasons:
    • They provide a better user experience on the web.
    • They typically have generous data plans.
  • The data includes both phones and tablets. Since tablets are used more heavily on the web than phones, this results in their being more strongly represented in the web usage data relative to their actual percentage of distribution.

 

What devices are included in DeviceAtlas?

- The scope of DeviceAtlas is web enabled devices, or more specifically, any non-desktop device that includes a browser. Today this covers mobile phones, tablets, ebook readers, games consoles, TVs, set top boxes, media players, and more device categories are added as they emerge. Feature phones without browsers are not included, and nor are dongles or data modems today.

What does ‘rank’ mean?

- The term ‘rank’ is used to describe the relative web usage of a device in the country of interest, or in other words, where it falls in the ordered list of devices as measured by the degree of usage of those devices on the web.

What does ‘web usage’ mean?

- The term is used in the Data Explorer to mean the propensity of the device to be used on the web. This does not mean the degree of distribution of the device, and is weighted towards high end devices. Factors that influence this are the subscriber’s data plans for example, and the relative cost of mobile data in different countries. No weighting has been applied to correct for these factors.

Why is my country missing?

- If a country is not represented, this means that we do not yet have sufficient density of data for that country to be able to expose it. If less than 200 different device models are seen in a country, this is considered to be not statistically robust, and therefore the country will not be shown. (Some countries have several thousand different distinct device models active at any time).

Where does the data come from?

- The source of the usage data for the Data Explorer is a range of Afilias Technologies and partner web properties, covering much of the globe and with in excess of 200,000 sites.

What about desktop traffic?

- Desktop traffic is not included, and bots/crawlers are also removed. The dominance of the iPad in some territories is likely to be a consequence of tablets being more like desktop browsers in terms of their usage patterns.

Why do Apple devices dominate the statistics when Android devices are shipping in larger quantities?

- There are several factors at work here, as follows:

  • Apple device hardware versions are indistinguishable on the server side, and hence all iPhone versions are grouped into one single device titled iPhone. This results in a larger aggregated traffic level, artificially boosting it in the rankings. The same applies for the iPad.
  • iPhones are typically supplied with generous data plans, increasing their propensity to be used on the web relative to other devices.
  • The data includes both phones and tablets; most measures of OS distribution are based on smartphones only.

Recommendation: compare the OS, or vendor, to minimise the impact of these distortions.

 

How are proxies handled?

- Major proxies which affect the apparent geographical distribution of traffic are the Opera proxy, Nokia proxy, RIM proxy, Onavo proxy, and Silk proxy. Others may also apply. These are identified where possible as territories to avoid distorting the apparent traffic levels in the countries where the proxies are based.