The new Webtrekk device detection has been completely revised to meet our customers' business needs.

Nowadays, the very rapidly growing range of products and models of the consumer devices presents us with new challenges in terms of device detection. For example, since most websites have over 50% combined mobile and tablet traffic, it is vitally important to get mobile device data (or device detection in general) with precision and with up to date information so that you can have the ability to identify where improvements could be made. These are, for example, targeted campaigns and adapting mobile strategies accordingly.

Webtrekk has not only reworked the underlying infrastructure but has redesigned the Data Processing Pipeline. Our objective is to provide a stable and self-sustaining infrastructure with minimal manual maintenance.


Example: Device Class Analysis in Webtrekk Analytics


The improved device detection has been running in the background since March 29, 2018, parallel to the old device detection. Since July 2018, the improved device detection has also been used for analyses in Webtrekk Analytics.

The effects of the improved accuracy in hardware identification can be observed very well in the analyses in calendar week 13 – from this point on the new device classes are filled with data and you will see the already mentioned shifts in traffic distribution. As a rule, this ensures retroactive deviations in the single-digit percentage range.


In the following, you will find an overview providing the general outline of the status quo, the changes and the impact on the customers' end as a result of the respective change:

New & Improved FunctionsDescriptionCustomer's Benefit & Impact

Update Process Automated

The device update process is now automated.Previously, any addition of new devices requires manual updates and therefore manual work by the development teams.
  • New devices are instantly detected and recognized at first request or user agent hit.

  • Better convergence with the technical service provider for device detection.

  • Scalable if there is a need to add more device attributes in the long run.

Device Naming

Please keep in mind that you may need to update your saved filters based on the changes of device naming.

New device naming convention:

Vendor name + device family

Except for fallback device names, such as "any iPhone".

So far, device names are derived from the following information:

  • Vendor name + device model (e.g. Nokia 8400)
  • Vendor name + device family (e.g. Samsung Galaxy S7)
  • Vendor name + operating system (e.g. iPhone OS 10)



With the new device naming convention, a consistent, coherent and reliable structure is created.

An alignment of old and new device names is given through mapping of old and new naming conventions.

The following list of fallback device names is now applicable:

  • any Android
  • any iPhone
  • any iPod
  • any iPad
  • any HbbTV
  • any Desktop & Laptop

Before/ After Mapping Table

BEFOREAFTER
Fallback NameDevice ClassFallback NameDevice Class
Diverse PC & LaptopDesktopany Desktop & LaptopPC/ Laptop
Other AndroidSmartphoneany Androidunknown
iOSSmartphoneany iPhoneSmartphone
iPhoneSmartphoneany iPhoneSmartphone
iPhone OS XSmartphoneany iPhoneSmartphone
iPodSmartphoneany iPodother
iPadTabletany iPadTablet
HbbTVTVany HbbTVTV


The assignment of devices, which are not recognized, is now clearer because fallback names are more informative. This allows a reduced, but a more accurate list of fallback names.

Analysis of Device Models

The detection of specific device models is now possible.

Previously it was not possible to detect specific device models or to analyze devices based on their device family.


No model name will be shown in the device full name; however, we will add a new dimension (device model) within the next releases so that you can add the device model to the data table as a separate dimension for a better drill-down.


Device Classes

The following device classes have been introduced:

  • Console
  • E-Reader
  • Mediahub
  • Smartwatch
  • Other (for the ones that do not match these classes)

The new device classes provide a useful addition to the already existing classes:

  • PC/ Laptop
  • Smartphone
  • Tablet
  • TV

The device class "Unknown" is still in use.

By adding these device classes, more granularity in device class analyses is enabled.

Device Properties

Unused device properties are no longer supported.

The following device properties are no longer supported:

  • MOBILE_CONNECTION_TYPE
  • WIFI
  • POINTING_METHOD
  • HAS_QWERTY
  • WAP_PUSH_SUPPORT
  • XHTML
  • COLORS
  • COOKIE_SUPPORT
  • FLASH_LITE


Aggregation

While aggregation, the dimensions "mobile clients" and "resolution" are not supported anymore. Aggregates containing these dimensions will be removed.In the long term, the aggregation will be replaced by a more flexible system.

For customers using "Webtrekk Quick Insights", this means that device data (mobile devices and screen resolution) cannot be shown anymore.

For customers using "Webtrekk Q3", raw data can still be used when available.

New Device Dimension in Analytics

The following device dimensions have been added:

  • Device model
  • Device familiy
  • Device name

The above dimensions have been added to take advantage of the improvements in the new device detection.

In the Webtrekk Support Center, you will find the FAQs for the new device detection: FAQs Device Detection.




Release date: July 02, 2018