The Evolution of Identity - Part 1: Identity Solutions for addressable: The five attributes you should care about

The industry has undergone an identity transformation in recent years, welcoming a new era of advertising where user privacy is at the center. Google announced today another delay in the deprecation of third-party cookies on Chrome, which is now scheduled for the beginning of 2025. Many may be breathing a sigh of relief that the deprecation of third-party cookies in Chrome is delayed again. However, Google is only extending the timeline and the industry must prepare and adjust to the changes regardless. Device IDs and IP addresses are also at risk and privacy laws keep emerging around the globe, changing and challenging the long-held norms of identity resolution.

The evolution of identity has already split the advertising world into two environments: 

• Addressable: when a user can be identified with privacy-preserving methods
• Non-addressable: when a user can’t be identified

An example of an addressable user is one who logs in on the web or opts into sharing their data within app environments. Addressability is incredibly valuable since it not only allows advertisers to persevere targeting and measurement but enables seamless and relative ad experiences for the consumer. However, not every user will login everywhere. Looking at my own online behavior, I don’t log in to every news website or blog that I visit. In fact, I only log in to my top 10% favorite websites, ones that I visit very often. Thus, most of the time my behavior on the open web would fall into the non-addressable category. That is why it is critical for advertisers to embrace solutions for both addressable and non-addressable to maximize audience reach. 

While both addressable and non-addressable solutions are critical to the advertiser success, this blog post breaks down the attributes of identity solutions for addressable inventory and their benefits for you and your consumer.

1. Identity resolution - What an ID really is?

An ID solution helps advertisers identify a person and/or household across digital environments for a multitude of advertising use cases, including frequency capping, targeting and measurement. The variety of signals considered and the ability to tie them together to create user profiles differs for each ID solution. It’s essential to understand what ID really represents for each solution and how it affects your campaigns and the consumer.

Some solutions resolve on a single signal. For example, if a user has three different email addresses, three different IDs are created, and each email is treated as a unique user. What does this mean for advertisers and users? Forget about frequency capping controls. The user will receive the same ad three times, resulting in unpleasant experiences. Also, say goodbye to true attribution. If the user received an impression being logged in with email A, but they converted using email B, then there is no way to attribute the conversion to the initial impression.

ID solutions that resolve on a people-level offer more sophisticated approaches for smart budget spending and meaningful experiences for the consumer. In this case, multiple signals get tied together into a single ID representing an individual. This means that an ID can be tied to multiple email addresses, phone numbers and other signals, so advertisers eliminate the frequency capping and attribution challenges of a solution relying on a single signal. An even better scenario is to accompany people-level identity with a household-level where a single ID represents all individuals and their devices within a household. Combining these two identity-resolution levels allows you to achieve effective omnichannel targeting and attribution, including Connected TV.

2. Data:  Probabilistic or Deterministic?

The identity solutions in market today take different approaches regarding the type of data they collect to create user profiles. Data can be deterministic or probabilistic. Deterministic data is information known to be true (e.g. name, age etc.). Probabilistic data relies on inference-based signals indirectly tied to a user. Fingerprinting is a type of probabilistic identification in the advertising industry, though it is increasingly scrutinized. 

Both approaches have pros and cons when used for identity solutions, so let’s look at them through the lens of the consumers to understand their impact on them.

ID solutions that rely on deterministic data have high accuracy. Accuracy means that the advertiser can be confident they are reaching the right user in the right context to drive the best possible ROI on every advertising dollar spent. Consumers, in the meantime are served a more relevant ad experience. Deterministic IDs avoid situations where for example a person identifying as male is served ads for female clothing. Even more importantly, deterministic IDs offer privacy controls so that users are able to control their privacy choices at any time and have confidence that their privacy choices will be honored. 

Probabilistic IDs leverage signals such as IP address, user agents, etc. to identify the user, but can’t reliably retain the privacy choices of users. While advertisers do get more scale with a probabilistic ID, consumers lose control over their data, which contradicts the industry's goal to protect consumer privacy better. On top of that, the longevity of probabilistic solutions is questionable as the signals they rely on are at risk with the introduction of browser and OS changes such as Apple’s iCloud Private Relay and Privacy Sandbox IP protection.

3. Safety-first: Privacy controls, governance and security 

Safety in the context of identity is all about understanding if data is safely held with identity partners and how easily consumers can manage their privacy preferences.

As Shiv Gutpa, Founder of U of Digital mentioned in our blog post “Convergence Disrupts Our Comfort Zones” the changes in the advertising industry have forced the convergence of connecting first-party data and customer management tools inside the marketing stack with the audience segments and targeting tools of the advertising stack. Leveraging first-party data for effective advertising is critical, and will only become more important as the landscape continues to evolve.

At the same time, we understand that consumer data is one of our industry’s most valuable assets, and advertisers must ensure that identity partners treat it with the utmost concern for its security and protection. Ensure you have a clear view of your identity partner's privacy and data collection policies, such as where their data is coming from, who has access to it, and how it is stored. 

Also, consumer privacy should be at the core of every identity solution. Given that the transformation in the industry started in the first place to better protect consumer privacy, this should be a non-negotiable characteristic. ID solutions should provide transparency to consumers regarding how they collect and use their data and always respect user preferences. Does your partner have controls in place for users to manage their privacy and data collection preferences? Can users easily opt-out of these solutions? Finally, does your identity partner have a well-defined governance and security framework that ensures smooth operation and longevity through industry changes?

4. Interoperability - Does your ID partner play well with others?

Interoperability is the ability of ID solutions to connect and exchange data with each other for effective targeting and measurement.

Interoperability is key in the identity-constrained world. If your identity partner is not interoperable with other solutions (and more than one is key), it’s like speaking a foreign language that no one else understands. For example, your DSP leverages identity solution A, but your measurement vendor leverages identity solution B. If these two identity solutions don’t communicate, they aren’t interoperable. This means there’s no way to understand the impact of this campaign on your business outcome and how to improve the experiences for your high-value audiences. A highly interoperable ID solution means increased scale and effective measurement for marketers and cohesive and seamless experiences for the consumer.

5. Proof points and testing: Does my ID solution work?

Finally, none of the above matters if the ID solution doesn’t have the stats and metrics to back up its performance and demonstrate that it actually works and works well. 

Metrics around the adoption of ID solutions are a good indication of success when combined with performance metrics. For example, get insights into the number of publishers who have adopted and the average coverage percentage on publisher supply. Can advertisers actually spend their budget using an ID solution? Testing availability, especially one that allows you to run advertising campaigns using the ID solution from your identity partner, is necessary at this stage. Don’t hesitate to ask for it; it’s a valuable tool that can help you understand the impact of the industry changes on your campaigns and gain valuable insights for the future.

Stay tuned for the second blog post to learn about solutions you should be leveraging for non-addressable inventory. 

Yahoo Identity Solutions takes an integrated, privacy-centric approach to identity to help advertisers future-proof their business in the cookieless and ID-less world. Our approach enables advertisers to reach their audiences across both addressable and non-addressable environments maximizing campaign effectiveness. Learn more about Yahoo Identity Solutions here.