A primer on privacy-forward advertising

In North America alone, $300 billion is spent on digital advertising every single year. And all signs point to that figure increasing.

But where is this money going?

Walled gardens are an increasingly large source of this ad-spend. According to Jounce Media’s latest market outlook, walled gardens are set to capture 69% of global non-search digital advertising investments in 2021. This is up from 28% in 2017. And as the demand for walled gardens is actively growing, the demand for independent ad tech is shrinking. 

Four companies (Google, Facebook, Amazon, and The Trade Desk) control 60% of demand for all web, mobile app, and CTV spend in 2021, up from 33% in 2017.

Generally speaking, these dollars go to two main types of advertising: demographic or behavioral. 

Demographic advertising targets users based on the attributes publishers think they know about you, such as age, gender, location, etc. and results in a very broad, and often inaccurate, view of the user. 

That’s why most businesses tend to favor behavioral advertising, which relies on targeting users based on identifiers including email addresses, third-party cookies, MAIDs, etc, based on the behaviors they have performed. Think about a time when you visited a website and then saw an advertisement for that website later in the day. That retargeting ad is thanks to behavioral advertising and third-party cookies.

However, recent ecosystem changes have threatened this type of behavioral advertising as we know it.

How the cookie crumbles

Let’s take a quick step back to investigate the two types of cookies that websites employ to make this type of targeted advertising possible.

First-party cookies are created and stored by the website itself and typically enhance your experience. They help the website remember preferences, store identifiers, track analytics, and provide a more personalized experience. This information is stored by the website in the web browser’s memory so they can remember it next time you come back. Many websites store this data for future analysis.

Third-party cookies are created and stored by companies separate from the website’s owners. They use cookies for similar reasons, to remember who you are and collect data to provide ads relevant to you. These are typically set up across thousands of websites and the data is stored by ad tech vendors. Third-party cookies allow these vendors to build a picture of your actions and sell that data to brands and advertisers across the internet.

However, third-party cookies are being threatened as governments create more user privacy regulations. As GDPR and CCPA came into effect in recent years, it’s become harder (and in many cases, illegal) to collect personal data in such a fashion. Nor is this a temporary fad. In the next few years, 65% of all people in the world will be covered by some kind of privacy regulation.

In addition to governmental regulations, internet browsers are blazing their own privacy-centric trail. Apple Safari and Mozilla Firefox have greatly reduced the efficacy and use of third-party cookies on their browsers and Chrome announced plans to eliminate third-party cookies in 2022 (link). 

The mobile ecosystem quickly followed suit, and in June 2020, Apple made IDFA opt-in, asking users to explicitly consent to be tracked across mobile apps, substantially reducing the value of MAIDs like IDFA. Since then, the opt-in rates have been quite low, coming in at under 20%. Google promises to bring down the hammer next.

From what we’ve observed so far, these changes are affecting businesses in two major ways:

• Behavioral retargeting of anonymous prospects: Businesses are only able to show ads to consumers collected by their own first-party data.

• Measuring the impact of advertising: Ad attribution is becoming much more complex. Businesses have no way to measure the full customer journey from ad to conversion as it’s difficult to attribute a lead to its exact source.

But wait! There is hope on the horizon.

While there is no one silver bullet to replace third-party cookies, based on our experience of working with thousands of Segment customers, we are seeing businesses respond in the following ways:

1. Establish a first-party, consented data pipeline

Marketers have access to generic demographic information through tools like Google Analytics, but need a conversion to capture users’ data before they can begin to make data-driven decisions. Once a user makes a purchase, submits a form, etc. you can match the interactions to a specific user. However, any interaction completed before conversion remains anonymous.

Segment can track the actions of a user before they convert through an anonymous ID. Once the user data is revealed, Segment bridges the anonymous ID interactions with those of the now-known user. You can collect behavioral, web analytic data at the user level and supplement that with the transactional data you already have.

2. Provide more options for pure server-side integrations (e.g., Facebook Conversion API and Google Enhanced Conversion)

Through server-side integrations, marketers must tell platforms when a user takes an action that progresses them to their goal. While more difficult to configure than just copying and pasting a third-party pixel to your site, server-side integrations let you bring your first-party marketing data directly to your website in real-time. 

In working with your engineering team, you can configure APIs for each website that you want to send data from. You can also add data from offline sources, such as your CRM, apps, support tools, and more. This results in personalized ads being shown without compromising the user’s security.

Using Segment gives teams more control over what they wish to send to server-side integrations as engineers do not need to build new APIs for every network, just rely on the one Segment API.

3. Migrate prospecting audience targeting to cohort audience targeting

Instead of retargeting your existing users, reach new leads through lookalike audiences (called cohorts). Once you determine what criteria your highest-valued customers all have in common, employ your data/analytics team to batch export and upload all users based on that criteria. You can tell your ad network how to use the list, such as retargeting, creating a lookalike, or suppression.

This creates an algorithm that can find net new leads to targets who follow similar online behavior to your target customers. Cohorts are designed to be broad enough groups so that individuals remain anonymous, using common interests in place of individual identifiers.

Segment can automate the batching for you, which stops the need to manually export and upload your data. This frees up your data team and lets you focus on the creation of the cohorts.

4. Transition to technologies that accept first-party IDs for targeting

We’ve discussed walled gardens and how they have a large understanding of their users as they collect so much data. Segment integrates well with walled gardens because we integrate with the players who have the majority of their share (Facebook, Amazon, Google).

But what if you want to collect first-party data outside of a walled garden? Tools like The Trade Desk and UID 2.0 take a user’s email address upon submitting a form and turns it into a string of numbers. This results in a consented way to market that maintains the user’s anonymity. 

In closing

What’s the future of advertising? While nobody knows for certain, what we do know is that the third-party cookie that long powered digital advertising is gone forever. This change might provide businesses with some short-term challenges but ultimately will provide the ground for marketers to rethink and re-strategize alternatives that deliver a better experience for users and advertisers.