When Google announced its plan to remove cookies from Chrome, whispers of an apocalypse echoed across the ad industry. Many feared this change would disrupt the $120 billion in ad spending that currently relies on cookies. Words like “demise” and “unprecedented” appeared in headlines.
And then Google surprised us with its latest announcement: Cookies are actually here to stay.
A Privacy Sandbox blog post cites feedback from regulators, publishers and industry bodies as the reason for this pivot. It also hints at alternative privacy proposals, most notably a “new experience in Chrome” that lets people make informed choices that apply across their web browsing.
If this vision comes to pass, it would be parallel to Apple’s AppTrackingTransparency (ATT), which effectively eliminated Apple’s mobile ad ID, the IDFA, by prompting people to say yes or no to sharing their data.
Like cookies, the IDFA enabled advertisers to target and measure ad performance across apps. When Apple introduced ATT, many mobile marketers were unprepared, and business suffered as iOS addressability fell. Facebook’s Audience Network took a $10 billion revenue hit. Today, ATT opt-in rates range from about 12%-40%, depending on the app category.
Despite these setbacks, the mobile advertising industry has not only recovered but thrived in the post-ATT landscape. eMarketer predicts in-app video ads will capture over 30% of total US mobile advertising spend for the first time this year. Games advertising alone, an in-app format, is the second most effective for driving purchase and conversion, according to IAB. Globally, the in-app market is forecasted to achieve an annual growth rate of 8.65% by 2029.
So what can overcoming the impact of ATT teach us about overcoming the latest uncertainty around the future of the cookie?
Advertising is still effective without cross-site tracking
Google’s proposal and ATT evoke a common fear: that a reduced ability to track users across touch points will lead to worse ad performance. Indeed, there was a notable drop in performance shortly after ATT was introduced, particularly for user acquisition campaigns. But publishers and advertisers adapted.
In place of the IDFA, advertisers experimented with other, then-underutilized signals such as context and historical performance to predict user interests and behavior. Many invested in machine-learning resources, focusing their efforts on optimizing for outcomes rather than targeting individuals. Others established their own IDs, persistent within apps instead of across them, to categorize users. Some switched to fingerprinting, or probabilistic attribution, which Apple has since cracked down on.
Apple’s alternative method for attribution introduced in the wake of ATT, SKAdNetwork 2.2 (SKAN), left much of the industry dismayed. Performance marketers criticized its conversion postback delays and lack of granularity, which made attribution nearly impossible.
AdExchanger Daily
Get our editors’ roundup delivered to your inbox every weekday.
Daily Roundup
But SKAN has evolved since then, giving app marketers greater granularity in post-install performance, conversion optimization and web-to-app measurement – all while keeping user privacy intact. Some are even calling 2024 “the year of SKAN,” with 94% of in-app inventory reportedly SKAN-ready.
This evolution in mobile attribution highlights an important point for the cookie-based web: Much like ATT, adapting to cookie changes won’t be a one-off project but a continuous evolution that unfolds over many years.
Google’s cookie substitute, the Privacy Sandbox, will likely also evolve past the challenges plaguing it today. The same goes for other cookieless alternatives, such as email authentication, contextual and on-device targeting.
While the current state of cookieless solutions may paint a bleak picture, technology is continuously improving, and with it our ability to balance privacy with effectiveness.
Addressability is only one piece of the revenue puzzle
We’ve all seen the data suggesting that only a small portion of the digital universe will be addressable without cookies. We’re already seeing comparisons between ATT and cookie opt-in rate projections on Chrome. However, the in-app industry has proven that addressability isn’t always needed to drive revenue.
When ATT was first introduced, developers quickly explored new ways to monetize users who opted out of targeted ads. User segmentation was a popular strategy, where non-addressable users were targeted with campaigns that prioritized context, placement and timing. Until the widespread adoption of in-app bidding, this strategy drove significant revenue for apps with diversified demand sources.
Today, a parallel strategy is emerging for web environments. Publishers are increasingly relying on their own first-party data to understand and engage users more effectively. Contextual targeting is also evolving, using machine learning to analyze page content and deliver ads that align with user interests.
New metrics such as attention are gaining traction, providing insights into user engagement and enhancing targeting precision without relying on traditional tracking mechanisms.
The inability to target consumers individually doesn’t have to diminish their value. Non-addressable users are still capable of taking an action or making a purchase. It’s our responsibility to infer these behaviors and preferences and match them with the right advertisers, which is very possible.
Yet adapting to ATT is still very much a work in progress. Each impending SKAN update leaves many wondering: “What will Apple do next?”
Nonetheless, the ATT experience has shown that, with creativity, experimentation and a willingness to embrace new methodologies, the advertising industry can adapt and grow in a more privacy-focused ecosystem. We’ve already demonstrated resilience that offers hope and guidance for the future of cookies, whatever that may be.
“Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.
Follow Jun Group and AdExchanger on LinkedIn.
