Cookieless Tracking: What It Is and How It Will Optimize Your Marketing Analytics Strategy

In today’s increasingly privacy-centric marketing environment, first-party data has never been more important nor more difficult to collect. A recent research study on consumer preferences around online privacy, conducted by Google and Ipsos, reports that “91% of internet users aged 16-74 say they are more likely to shop with brands that provide offers and recommendations that are relevant to them.” 

The Current Data Privacy Landscape 

Generating insights about consumer preferences and delivering an experience based upon those insights is critical for competitive advantage. At the same time, data privacy regulations such as General Data Protection Regulation (GDPR), ePrivacy, and California Consumer Privacy Act (CCPA) require consumer choice with respect to data collection and processing activities. 

Users are increasingly exercising those rights, with between 30-40% of users declining cookie consent on EU retail sites. Technical restrictions such as Safari’s ITP and consent imposed on iOS devices further restrict access to any identifier used to associate interactions and deliver personalized experiences. 

How is a marketer supposed to meet consumer expectations while unable to collect the first-party data required to understand user preferences and how they are interacting with digital properties? 

Getting Started With Data Privacy 

At InfoTrust we recommend a privacy-centric measurement methodology to collect necessary data while fully respecting the privacy preferences of users. This measurement methodology relies on three core principles:

1. Optimize the proportion of “named” users
2. Collect and analyze anonymous data without identifiers (and cookies) with cookieless tracking
3. Develop cookieless measurement capabilities

Let’s examine each:

Optimize the proportion of “named” users with User-ID

In this context, “named” users are those who have indicated their preference to allow the use of their data for measurement use cases and have voluntarily provided some sort of persistent identifier (usually an email). These could be users who have subscribed to your newsletter, registered for a loyalty program, or submitted for an offer/coupon. For these users, there is the ability to assign a user ID, associate interactions over time, and collect qualitative preference/profile information. 

In internal analysis conducted with a number of partners in the D2C ecommerce space, we have found that on average 6% of total users fall into the “named” user segment. This 6% of “named” users are 5x more likely to convert than those who have not registered with the site in some way. It doesn’t take a complicated LTV analysis to understand that these users are highly valuable, and you want to find more consumers like them. Analyzing interaction and campaign data through this lens leads to the insights necessary to refine targeting and audience strategies in the privacy-centric environment.

Beyond leaning on the named user cohort to derive insights about high-value users, you can also take advantage of opportunities to analyze user behavior across domains via the provided user identifiers (namely email). User lists and the associated identifiers are the prerequisite to combine first-party data sets with partner first-party data via clean rooms. This can help you further analyze user behavior across the entire acquisition journey to optimize media campaigns. 

These insights and activations are not possible without this “named” user cohort. Maximize the proportion of users in this cohort and maximize the value which can be generated for your organization.

Collect and analyze anonymous data without identifiers (and cookies) with cookieless tracking

This is a new concept as it relates to digital analytics. Traditionally, web analytics has revolved around the concept of an anonymous user identifier with which all interactions are associated. This allows for the calculation of core metrics such as “sessions” and “users”. The shift to an event-based data model opens the door to measurement-based purely on interactions and does not require a persistent identifier (stored in cookies) to deliver meaningful reporting. 

The leading solution to accomplish this cookieless tracking use case currently is Google’s Consent Mode with Google Analytics 4. Using an event-based measurement architecture, pure interaction data such as page loads, button clicks, conversion events, and transactions are able to be observed and collected while no cookies are set nor accessed for users who have not given consent. For consenting users, data is collected and an identifier is set and collected in a first-party cookie as normal. This combined measurement approach allows for full visibility into event/interaction counts (number of conversions, number of transactions) while establishing the dataset necessary to model insights from the consenting user dataset to be extrapolated to the full dataset. 

Core use cases such as understanding which products or content is most popular, which campaigns are driving the most conversions, calculating the proportion of consenting users, and reporting on ecommerce metrics are still able to be accomplished in a way that fully respects the privacy preferences of users.

Develop cookieless measurement capabilities

Media measurement on the web relies on the attribution of conversions and conversion value throughout the funnel of user interactions. Traditionally, user interactions have been able to be associated with a single user and connected to conversions via the use of cookies (specifically third-party cookies due to their ability to identify a single user across multiple domains). This ability goes away with the deprecation of third-party cookie support across all browsers by the end of 2023. In response, it is imperative to develop measurement capabilities that are not reliant on cookies. Two of the approaches to focus on are regression-based attribution and the measurement of incrementality via the use of geo-testing.

Regression-based attribution is an expansion of the methodology used for media mixed modeling. It relies on aggregate ad and conversion data across channels to understand the impact of past marketing strategies and provide insights into how campaigns may behave in the future. This method can help in evaluating media effectiveness and to enable future planning and optimization in the absence of persistent user identifiers.

Incrementality measurement via the use of geo-testing is an approach where you can test different advertising strategies within a defined geographic area. Measuring the incremental change in macro-conversion outcomes (transactions, revenue, margin, etc.) in relation to those same outcomes in the control geos enables evaluation of different marketing mixes and optimization based upon those insights. Again, this approach is not reliant on any personal identifiers associated with individual users. 

Additional cookieless approaches to measure and optimize lean on the “named” user cohort discussed before. These strategies include the full-funnel evaluation of campaigns via the use of clean rooms for in-channel insights and A/B testing campaign strategies specifically on the named user cohort. Both offer the ability to measure and optimize in a privacy-centric manner.

To meet and exceed user expectations, it is critical to measure the effectiveness of current advertising activities as well as to test new strategies. These three core principles of the InfoTrust privacy-centric measurement methodology will allow you to measure user behavior and campaign effectiveness while respecting the privacy preferences of users. Competitive advantage in the new privacy-focused environment will be realized by those that execute the best within this framework.