Third-party cookies, the long-standing targeting tool, are on their way out, although Google's recent announcement of a delay in phasing them out from Chrome offers a temporary reprieve [1]. This delay underscores the ongoing industry challenges in transitioning to a cookieless world.
Estimates suggest a significant portion of marketers (around 75% according to Adobe’s research) still heavily rely on cookies, highlighting the need for robust alternatives. This blog dives into the world of universal identifiers, potential replacements for cookies in building audience profiles. We'll decode the two main categories – probabilistic and deterministic – exploring how they differ in their approach to tying user engagements across devices and the types of data they leverage. Buckle up, as we unveil the alphabet soup of identifiers that will power targeted advertising in the coming years.
Deterministic identity resolution, as its name suggests, uses deterministic data, which is information known to be true and accurate because it is provided by users directly or is personally identifiable. Common examples of deterministic identifiers include names, email addresses, and phone numbers.
Deterministic matching links user profiles based on this deterministic data. For instance, if a user logs in with the same email address on two different devices, the deterministic model would recognise that these devices belong to the same user.
The primary advantage of deterministic matching lies in its accuracy. It provides a high degree of confidence in the user profiles created and is particularly useful for direct, one-on-one communications, such as personalised emails or targeted advertising.
However, deterministic matching also has its limitations. The most significant one is its dependence on users providing their personal information, limiting its scalability. Its effectiveness also decreases when users are not logged in or when a piece of deterministic data is not available.
On the other end of the spectrum lies probabilistic identity resolution. This model does not rely on personally identifiable information. Instead, it uses a range of data points, such as IP address, device type, browser type, or operating system, and employs predictive algorithms to link these data points to an individual user profile.
The probabilistic model is like a detective, piecing together clues to make an educated guess about a user's identity. For example, if two different devices are frequently used on the same Wi-Fi network and exhibit similar browsing behaviour, the probabilistic model might conclude that these devices belong to the same user.
Probabilistic matching offers the advantage of scalability, as it does not require users to provide personal information. It allows advertisers to build customer profiles even when first-party data is limited, making it useful for reaching top-of-funnel prospective customers.
However, the probabilistic model is not without its shortcomings. The primary one is that it involves a certain degree of uncertainty. Its accuracy depends largely on the quality and diversity of the data it uses, and it can be less effective in connecting people across different devices. Furthermore, some of the data points it relies on, such as device IDs and third-party cookies, are becoming increasingly difficult to collect due to changing privacy regulations and user preferences.
Determining which model to use—deterministic or probabilistic—largely depends on the specific use cases and data governance policies of a business.
Deterministic matching, with its emphasis on accuracy, is ideal for one-on-one communications and personalisation. It is particularly effective when a business has an extensive first-party database and a strong relationship with its customers, characterised by trust and consent to use their personal data.
On the other hand, probabilistic matching, with its scalability, is useful when first-party data is limited or when reach is a priority. It can be an effective way to build customer profiles for top-of-funnel prospective customers, or when the consequences of inaccurate matches are minimal.
Importantly, deterministic and probabilistic models are not mutually exclusive. They can be used in combination, leveraging the strengths of each to achieve a more comprehensive view of customers. For instance, a business might use deterministic matching as the foundation of its identity strategy, ensuring high-quality customer profiles, and supplement this with probabilistic matching to extend its reach and fill in gaps in its customer data.
There are several methods to store these alternative IDs other than in cookies, such as HTML5 storage, Etags, and Evercookies. All these methods have their pros and cons - but let’s deep dive about these methods and technicalities in another blog post. For now, it is important to know that HTML5 local storages and Etags are the main methods followed by major ID providers.
There are more than several Cookieless ID providers in the industry and each plays a different role in the entire advertising ecosystem. Some of the ID solutions are solely for publishers, while some others are only accessible through using a specific AdTech provider - hence they’re not independent. Additionally, geographical coverage of these IDs are also quite important. Therefore, navigating around your options might be overwhelming as a CMO, Brand Manager. Below, we will just mention a few of those solutions that your brand can work with.
ID5 is one of the native Universal ID solutions, this means that you can onboard ID5 ID and use it across your tech-stack easily. ID5 combines its history of expertise in ID matching and strong EU coverage and security, with a privacy-first focus in user recognition plus measurement capabilities. First steps involve integrating brand data with ID5 at no cost. Then, a range of solutions are available to address the needs of publishers, advertisers, supply- and demand-side platforms, while also ensuring data protection and results.
NetID is spearheaded by the European netID Foundation, which was established with the aim to provide a European alternative to US-based single sign-on solutions. The foundation was founded by three major companies: Mediengruppe RTL Deutschland, ProSiebenSat.1, and United Internet. The netID solution allows for user-consent management and offers a single sign-on solution for users, which in turn helps brands and marketers in targeting and personalising their advertising campaigns while adhering to data protection regulations. NetID’s geographical coverage is mostly Germany, Austria and Switzerland.
Utiq, is a joint venture by four of Europe’s largest telecom operators: Deutsche Telekom, Orange, Telefonica, and Vodafone.It positions itself as a solution to improve addressability and help brands create sustainable audiences by organising the digital ad-ecosystem which is often found in silos, making it challenging for advertisers to optimise their campaigns across domains. Utiq is a European AdTech company that provides a Telco-powered Authentic Consent Service aiming to promote a trusted and responsible digital world. At its launch, Utiq partnered with Adform, making the Authentic Consent Service exclusively available through Adform's universal identity solution, ID Fusion.
LiveRamp's identity solution, RampID, is dedicated to accurate and comprehensive people-based identity resolution. By prioritising privacy and Personally Identifiable Information (PII) authentication, RampID offers an encrypted, persistent identifier that enhances programmatic addressability across the open web. With its platform, LiveRamp assists businesses in managing, unifying, and activating data across various channels, excelling in integrating offline customer data into online marketing platforms.
Zeotap's ID solution, ID+, is designed to overcome identity resolution challenges caused by third-party cookie restrictions and global regulatory changes. It serves as a shared, persistent identifier that connects users across the digital marketing ecosystem, ensuring the future of identity and addressability as third-party cookies phase out. ID+ was launched as a universal identity initiative to empower the marketing ecosystem in overcoming challenges related to the demise of third-party cookies and other platform identifiers.
Parrable ID
Parrable ID is a Full Device Identifier capable of identifying a device across different browsers and webviews on a single device, even with third-party cookie restrictions. It is known for its consumer privacy and consent-forward approach, encrypting every ID to prevent misuse and addressing the issue of uncontrollable tracking associated with cookies. The Parrable ID system manages a specific cookie that contains the user's ID and opt-out states, which is also utilised by standalone Parrable integrations outside of Prebid.
In order to understand the global adoption of IDs on a weekly basis, you can use Sincera. Sincera is an engine that crawls thousands of the top websites running Prebid to provide you with helpful data on the adoption rate of different solutions globally. Below is a snapshot of the most adopted IDs on the sell-side today.
If you prefer to track only European universal ID adoption you can have a look at Roq.ad’s However, it is important to note that Roq.ad currently doesn’t list ID5 in its tracker as ID5 encrypts its IDs in the bidstream data.
When it comes to navigating the world of Universal ID, brands often find themselves in need of support and guidance. This is where Pierian and our team of experts come in. With so many factors to consider, it's crucial to have a clear understanding of the parameters before making any decisions.
One of the key factors to consider is your geographical coverage. Depending on the regions you operate in, the approach to implementing a Universal ID strategy may vary because each provider may have different scales in your geography. Additionally, the size of your first-party data plays a significant role in determining the best course of action. Understanding the scale and quality of your data will help in selecting the right ID partner or looking for alternatives.
Another important consideration is your existing team setup and tech-stack. Do you have an in-house team with their own technology infrastructure, or do you rely on an external agency with their own tech stack? The answer to this question can greatly influence your choice of ID partner and the way you activate it.
Fortunately, Pierian is here to assist you every step of the way. Our expertise extends beyond simply choosing the best ID provider for your specific needs. We can also help you seamlessly integrate the chosen ID solution with your existing tech stack, such as your Customer Data Platform (CDP) and Customer Relationship Management (CRM) system.
[1] Google delays of third-party cookies
[2] Adobe Survey
At Pierian, our Consult, Create and Operate teams have worked with multiple global advertisers to incorporate first-party data, CDPs, and Cookieless Identifiers on media roadmaps. Let our expert teams and agency partners help you define your Cookieless roadmap and recommend suitable Universal ID partners. Get in touch with us to turbocharge your journey to Cookieless advertising with confidence.
