Join the discussion

and subscribe

Cross Device Tracking: An Introduction

By 3 June, 2016
  • Thinking
  • Thinking - Features


For 15 years the cookie has represented the principle of digital marketing identification.  However, its limited utility in a multi device, mobile, app driven environment has seen its throne toppled by a new wave of digital ID technologies. With users (according to a recent study by the OMD) switching from laptop to smart phone to tablet an average of 21 times in a single hour it has been increasingly important for marketers to able to connect location and behaviour on different devices.  By implementing cross device tracking, Michael Griffin, CEO of Retail search agency Adlucent, has argued that on average, 30% more search conversions could be attributed to smartphones.


There are two approaches used to solve the problem of cross device tracking; deterministic and probabilistic. Deterministic tracking is where a user is identified, across several devices, because they are logged into a multi-channel platform such as Google, Facebook, Yahoo or Twitter. By importing first party data that is consistent across devices like an email address, customer ID or login, deterministic tracking solutions are then able to create joins between mobile, desktop and tablet. By providing a unilateral road map of user behaviour deterministic tracking is considered as the more accurate solution as long as the journey is contained within the ‘walled gardens’ of the major platforms. Despite its merits many have accused deterministic tracking of suffering from a limitation of scale as it is unable to provide a total view of user behaviour. Privacy concerns have also been raised over its use of its use of personally identifiable information to build customer profiles.

The alternative is probabilistic tracking which involves ‘fingerprinting’ devices using a variety of attributes, such as number of cookies stored, device ID’s, public data such as IP addresses and behavioural data (geographic location and movement of devices geographically across time) which are then used to connect disparate devices to the same user. Although this predictive model which employs algorithms, analysis and probabilities to match devices with consumers is a less accurate solution than a deterministic approach it can be argued that probabilistic tracking offers a more expansive view of the entire market without compromising privacy.

Recently there has been a move to marry these two solutions and provide an even more accurate, composite sketch of user behaviour across multiple devices. Agencies are now using accurate deterministic data to validate the expansive joins made by the probabilistic approach.


The advent of this new form of direct marketing has prompted a new wave of regulations with privacy concerns being raised by the Digital Advertising Alliance (DAA) and Centre for Digital Democracy. In 2014 the DAA issued a ‘compliance warning’ to clarify that its Self-Regulatory Principles for Online Behavioural Advertising (OBA) were applicable ‘irrespective of the technology employed to collect and use consumer web-surfing activity to serve interest-based ads’. The DAA believed that that the cornerstones of their OBA principles (consumer control and transparency through real-time notice) were just as applicable to cross device tracking as to cookies. However, it can be argued that while cookies are relatively easy to identify and delete through common browser options that isn’t the case with new ID techniques. In a statement submitted to the Federal Trade Commission in the United States the Centre for Digital Democracy argued that ‘current self-regulatory approaches are ineffective and do a disservice to consumers by falsely claiming to provide privacy protection and user control’.  To combat the concerns over privacy an argument has developed to make cross device tracking ‘opt-in’ meaning consumers would automatically be opted out unless they decide otherwise. However, if this would be implemented it would signal the end of targeted online advertising. Lack of customer profiles and road maps would lead to irrelevant adverts being sent, multiple times across a variety of platforms, to the end user diluting brand messaging and effective marketing strategies.


Cross device tracking solutions provide marketers with the technology to fix broken user journeys and ultimately improve the conversation between brands and consumers. From an acquisition perspective cross device tracking can help marketers predict consumer trends and send targeted ads. Cross device tracking is also incredibly important when improving customer retention as it allows you to develop customer profiles across mobile, desktop and tablet and ensure that your re-marketing is relevant and efficient. Finally, the ability to track cross device behaviour and sales is essential when readjusting your attribution model. For example, while the final conversion may be made on a desktop the initial engagement might be attributed to an ad served or organic search made on a tablet or mobile. The ability to identify and understand the genesis of every conversion can help you allocate the adequate resource to the relevant channel or device ensuring your marketing operating is as cost efficient as possible.



Alex Bugle