Understanding the Customers’ Device Graphs is the key for Targeted Marketing

By Logicserve News Desk

  • August 4, 2017,

In the purview of the growing usage of the different mobile devices among the consumers globally, cross-device identity matching has gained a huge significance among the marketers. The modern day marketers are taking the device graph serious since it provides them a solid foundation for profiling the different users based on their mobile device usage for different purposes. Today, it is nearly impossible for a business owner to get a complete picture of their existing or potential customer unless and until they track their daily activities across all devices.

There are two major players specializing in the Device Graph, vis-à-vis Tapad and Drawbridge. There are also other companies like Criteo, which specializes in retargeting that develops Device Graphs as essential marketing tools. The big cloud companies like Adobe, Salesforce and Adobe have either developed their own tools or they have a partnership deal with vendors who develop device graphics as an integral part of their platforms.

Mike Murphy, the Senior Marketing Manager at Drawbridge suggests that all the Device Graphs are built on a single fundamental idea, however, each developer has a different formula in terms of providing weightage to a certain attribute. Typically, there are to major techniques, which are probabilistic and deterministic and most Device Graphs use both techniques in some way or the other.

The Deterministic technique are considered to be a gold standard since they  provide definitive results on which device the customer uses the most. This data is usually collected through the logins of the users. For instance, if the customer regularly uses his smart phone or the tablet device to log onto a particular website or portal, the brands can easily conclude that the user is most associated with that particular device. The Probabilistic technique on the other hand user the most common probable data points such as the time of the day when the device is used the most by the customer, the physical location of the device, which is derived by the IP address and the browsing patterns of the user. Murphy stated that Drawbridge has the potential to identify the device usage of an individual even if the same device is shared by other members in a household for different purposes.

In the video below, Tapad shows the Device Graph process which are largely derived using the probabilistic technique:


More often than not, Device Graphs are made using both deterministic and probabilistic techniques. For example, the user’s multiple device usage can be determined with a high degree of accuracy only if one the devices was used to log into the brand’s site. That one particular device, which is used by the customer becomes the definitive association and the other devices, which are placed at the same location may be used for visiting the same website on a daily basis. This and other attributes of the device usage pattern can give the Graph-maker sufficient data to conclude that the devices belong to the same users even though he/she has used only one device to login.

Whilst it is known that the deterministic-based approach yields more accurate results, there are many vendors who claim that the probabilistic approach has provided results that are 80 to 90% accurate. Retargeter Criteo has claimed that it considers Device Graphs to be a vital tool for its businesses and it has built the largest global scale mapping of email addresses to Cookies.

Murphy considers the usage of Device Graphs will continue to rise in the near future and it will continue to use a combination of both deterministic and probabilistic techniques. However, one of the possible exceptions would be the usage of biometric identification for logging into the aps or the website.

Related Blogs