From a cookie to a layer cake: Six alternatives to the third-party cookie to create a sustainable strategy

Every story comes to an end. Even the most beautiful.

For many decades, digital marketing had been built on a technology with unrivalled success: the cookie. More specifically, the third-party cookie, that little data file placed on websites to collect the behavioural data of web users.

But now that this story is coming to an end, digital marketing is about to enter a completely new era.

What will fill the gap left by the loss of cookies?

Who will benefit from the new paradigm shift about to take over the world of digital marketing?

What will the impact(s) be on marketing decision makers?

Let’s look at current alternatives to third-party cookies for building high-performance audiences.

Yes I said “alternatives” and the plural is important. There isn’t just one solution to replace third-party cookies: We’re heading towards multiple levels of overlapping technologies to make up for the contribution of this one little cookie.

We’re going to need an entire multi-layer cake.

Third-party cookies: A solution that’s still popular 

Google’s announcement that it would be gradually withdrawing the use of third-party cookies in Chrome by the third quarter of 2024 sounded the death knell for this technology. The loss has been slowed down by only a few years, given the initial withdrawal planned for 2020 was pushed back in 2022.

Many people wouldn’t mind if it was pushed back even later.

Operating for many years under restrictions that were as much technological (ad blockers, browsers that limit access) as legal (Law 25 in Quebec, the RGPD in Europe and the CCPA in California), third-party cookies were already under threat.

The loss of this digital signal directly affects the ability of advertisers to effectively target their users.

Despite this, 75% of marketing and user experience decision makers were still using the technology in 2023.  (Source : Adobe, 2023)

Which is either proof that change is hard to come by… or that advertisers lack reliable alternatives to integrate into their strategies.

Two opposing worlds for identity resolution  

Third-party cookies had formidable potential for identity resolution.

They made it possible, using someone’s browsing history (or more specifically the history of a device), to reconcile the browsing habits and interests of a profile with a marketing target.

Now, alternatives to cookies make use of two different approaches, one probabilistic and the other deterministic.

With the probabilistic approach, the probability that an individual corresponds to a certain number of criteria is inferred. The idea is to use models to deduce potential connections between certain data points such as IP addresses, device types and browsing behaviour. This means there is a certain margin of error in this approach.

The deterministic approach is based on knowledge about users that is deemed to be accurate and reliable. The level of certainty is therefore high. This approach is based on exploiting primary data: customer identifiers, customer relationship platforms or customer data, email addresses or anonymized phone numbers.

To mitigate the effect of the disappearance of cookies, the acquisition of persistent identifiers (such as email addresses) is essential to create and enrich sustainable audiences. Why isn’t this just an alternative? Because it’s actually an imperative for every organization. Don’t hesitate to read our guide to developing your primary data acquisition strategy to learn more.

Nevertheless, the acquisition of this data can be enhanced by complementary solutions. This is why there is a certain balance to be attained between precision (deterministic) and flexibility (probabilistic) when it comes to targeting.

Alternatives to third-party cookies for targeting and the creation of media audiences (

Solution #1: Google Privacy Sandbox or the “walled garden” route  

This is Google’s answer to the disappearance of third-party cookies. Composed of multiple APIs developed to replicate what used to be provided by cookies, Google’s Privacy Sandbox aims to place the browser at the centre of its strategy.

The Topics API infers users’ subjects of interest based on their browser’s website visits. The Protected Audience API facilitates the creation of remarketing audiences by adding the browser to audience lists.

Other than the fact that this solution is specific to Google, questions arise about the legality of this offering by the American web giant. It definitely strengthens their dominant position, since Google would control the entire digital advertising chain: the ecosystem (the Google Chrome browser), the bidding system and audience creation through Privacy Sandbox.

According to initial feedback, advertisers that tested Privacy Sandbox were unable to replicate the targeting performance offered by third-party cookies.

Might this be reason enough to once again delay the loss of third-party cookies in Chrome? Nothing could be less certain.

Solution #2: Universal IDs or the “identifier” route

In this case, users are still identifiable. This solution offers a variety of technologies, such as LiveRamp (RampID) and even The Trade Desk (Unified ID 2.0).

The Trade Desk example:

Targeting and measurement are highly precise, with granularity in terms of the user and initial consent that guarantees respect for privacy and maximum relevance of content offered.

Last point: for universal identifiers to work well, many companies within the ecosystem need to adopt them (advertisers, ad agencies, ad technology companies). The scalability of this solution therefore remains an open question.

Solution #3: Seller-defined audiences (SDA) or the “outsourcing”

IAB Tech Lab recently published exhaustive documentation on their proposal: seller-defined audiences or SDA. With this model, segmentation and sharing of audiences are delegated to publishers.

By using their visitors’ primary data, publishers can classify and categorize their visitors according to 1,600 predefined audiences based on three main elements: demographics, purchase intentions and interests.

This cohort-based system resembles Google’s solution, but with one major difference: IAB’s protocol is open source. Which is obviously not the case with Privacy Sandbox.

One drawback: Since audiences defined by the seller are created by publishers, advertisers have no control over the way in which they are created—they can only indicate which audiences they would like to target.

 Solution #4: Data clean rooms or the “anonymous sharing” route 

The data clean room is a technology that allows reconciliation between a publisher’s and an advertiser’s data sets. The key to making the connection? As is often the case, it’s an email address.

The good news is that the publisher can’t access the advertiser’s data and vice versa. The privacy of personal data is guaranteed. This solution is therefore a good option when it comes to respecting privacy.

The connection rate between user profiles may be a hindrance to using these technologies, since a high quantity of primary data is needed to obtain actionable results.

Solution #5: Dedicated ad platforms or the “self-serve” route 

La Presse recently made a self-serve ad platform available to advertisers called Atelier Direct. With this type of solution, publishers regain control through platforms that are directly accessible, with no intermediaries.

Publishers have fully understood the importance of primary data from their own audiences. So here we see the importance of collecting primary data and enhancing these data for publishers.

The publisher creates audiences and makes them available within a platform they control and manage, all without having to share their users’ personal data with advertisers.

The downside is that the multiplication in purchase platforms may lead to constraints in terms of management for advertisers.

Solution #6: Contextual targeting or the “back to the future” route

If, like me, you remember the web when ads were intimately related to the context of the page and site, you’ll be happy to hear that contextual targeting has made a comeback.

Except the foundations are really different!

Artificial intelligence is now used to create automated learning models specific to campaigns using proprietary data and contextual signals.

These models analyze URLs and make note of them based on their semantic relevance to the campaign brief. The result is a refined selection of content that closely corresponds to campaign objectives, surpassing the precision of standard segments.

This is way beyond your standard early-2000s display banner.

Anticipating the arrival of the layer cake 

Digital marketing is at a crossroads.

As demonstrated by our recent eMarketer survey, industry professionals are divided on which route to take in a cookie-less future world.

This is great news, since it means we are presented with an opportunity to reinvent our practice on a solid and sustainable foundation.

To navigate a world without third-party cookies, advertisers are going to need strategic advice to choose the right mix of solutions that will best respond to their specific needs, all while maintaining optimal performance and guaranteeing users’ privacy.

For everyone’s benefit.