INside Performance Marketing
Why a 'Probabilistic Approach' Could be the Answer to the Cross-Device Crisis

Why a 'Probabilistic Approach' Could be the Answer to the Cross-Device Crisis

In the early 21st century, digital marketers had it relatively easy. Desktop browser cookies were the de facto method of reaching consumers with relevant messages, and the practice worked well enough. But as a result of the explosion of smartphone and tablet adoption in the past decade – environments that are primarily cookie-less – times are tougher as consumers use an increasing number of devices throughout their days. 

The Internet Advertising Bureau (IAB) recently found that the average UK household now owns 7.4 devices that are able to connect to the internet. In such a fragmented, multi-device world, it has become increasingly difficult for brands to seamlessly reach consumers as they work, play, research, shop and socialise across devices. Most savvy marketers realise this problem and combat it with digital campaigns across computers, tablets, and smartphones, as well as between web and app environments.

This has resulted in consumers becoming moving targets for programmatic marketers. Today, we are endlessly followed from site to site by ads that don’t care if we’ve already made a purchase, already viewed the ad and not engaged, or if we even fit the target market. Marketers know this, but too often choose to fight the problem by employing a so-called “spray and pray” tactic across every channel available to them – a silo-based approach that is far from ideal. These brands are still reaching devices, not consumers, and the only way to solve that problem is with a truly accurate and scalable cross-device solution.

The best interest of all involved

But there is another way to come at this cross-device conundrum. Facebook, Google, and other platforms that ask users to log in and provide information are known as deterministic. This approach to data collection gives an intimate picture of users across all our devices. Once we’ve agreed to Facebook’s terms of service, for example, we consent to marketers using our information in order to reach us across the entire Facebook Audience Network – which includes external sites and apps such as Instagram. 

Enhanced personalisation is the Holy Grail for marketers and brands wishing to target users in a more granular way, but we need to ensure this is done with both marketers’ and consumers’ best interests in mind. Deterministic solutions are potentially limited in scale to a solution’s particular user-base, and rarely do marketers get to own the customer relationship – it’s through the intermediary platform.

The viable alternative to these walled gardens is to take a probabilistic approach, which we see as a way of using technology to democratise cross-device data for the rest of the web. This involves collecting non-permanent, user-resettable identifiers – such as browser cookies and device IDs – then correlating the different data points to predict device ownership, demographic information, interests, and other attributes. Other areas such as stock markets (predictions), meteorology (weather forecasting), and healthcare (how certain drugs will affect the body) also benefit from the probabilistic approach.

A race to the bottom

Consumers’ devices are connected via these various observations and signals, and the resulting understanding of consumers and devices can not only be over 97% accurate, but available for marketers at large scale – in many cases larger than those walled-garden environments I referred to above. Probabilistic approaches mean relevant ads can be served to consumers across the internet and across devices, without relying on login data.

With deterministic practices, corporations take our email addresses, phone numbers, birthdays and other details, and give all this to marketers for use in building data-driven campaigns. While this is the trade-off made for using so-called “free” services online, unfortunately we’re seeing less and less of the value delivered back to consumers, as publishers are in a race to the bottom to monetise their inventory. 

Digital marketing is an obvious application for cross-device solutions, allowing sectors such as automotive, retail, online travel, and others to benefit from better targeting, measurement, and true path-to-purchase reporting. The industry has its focus firmly on a data-driven future, something that is clearly here to stay with programmatic ad spend up 233% in the past year across Europe, and cross-device technology has a key role to play. 

There is also a role that probabilistic technology solutions have beyond this – we see great opportunity in helping publishers and e-tailers as they seek to deliver more relevant experiences for their customers. While the dream of being able to reach consumers across devices has mostly resulted in a retargeting-fest across all of our devices with little regard for frequency-capping, or whether we’ve really engaged with an ad, I believe this is the year for change. It’s time marketers took consumers into account, rather than devices. After all, it’s us as consumers that buy things and engage with brands – not our devices.


Nimeshh Patel

Nimeshh Patel

Nimeshh leads all strategic efforts for Drawbridge in the EMEA market. He is a former Chief Operating Officer at AOL Europe, where he played an instrumental role in restructuring the business. Nimeshh holds an MBA (distinction) and M-Eng in Aeronautical Engineering from Imperial College, London.

Read more from Nimeshh

Join over 10,000 performance marketers for the ultimate weekly update on industry news