One of the most pressing issues facing today’s marketers can be summed up, albeit glibly, by the following question: What will brands do when the cookie crumbles?
The cookie, of course, refers to those small text files embedded in internet addresses that have become a flashpoint in the ongoing debate over consumer data privacy, thanks to their ubiquitous use in targeted digital advertising.
Google is set to complete a gradual phasing out of the third-party tracking device from its Chrome browser in the second half of 2023. Last June, the company announced that it would delay the move for one year in the midst of backlash from influential players in the digital media ecosystem, including ad tech firms that depend on cookies to collect information about a web user’s browsing history, build audience segments and deliver precision-targeted ads.
Other browsers, such as Firefox and Safari, have been blocking cookies for years, but Chrome is by far the most widely used desktop web browser, with about a 60% U.S. market share, according to Statista.
How deeply Google’s decision will impact any given marketer depends on the extent to which the brand relies on programmatic ads fueled by third-party cookies (including ads that retarget or “chase” users around the internet with an offer for an item they’ve browsed or purchased before) to drive engagement, conversions and sales.
All told, U.S. marketers are projected to spend $191 billion, or two-thirds of their overall media budgets, on digital ads in 2021, according to eMarketer. Within the digital display ad category, the researcher notes, the overwhelming majority of spending (90%) is devoted to programmatic channels.
“The death of cookies could be a nightmare for a brand that builds its entire marketing engine on the back of behavioral, targeted digital advertising,” says Shubham Mehrish, vice president of analytics, data management and integration at Mars Inc. “Companies that haven’t invested in data strategies will struggle a lot more, and so will those that do not have strong direct-to-consumer (DTC) relationships. If you combine those two, a more traditional CPG company [than Mars] will absolutely struggle.”
A couple of factors will help Mars mitigate any negative consequences of the loss of cookies, explains Mehrish. One is the fact that 25% of its business comes from the broader retail confectionary and pet owner/hospital categories, both of which foster more direct communication and transactions with customers than most CPG brands. “We are a heavier first-party data business than other CPGs, since we have direct relationships with our pet parents and hospitals,” he says. “Access to this first-party data allows us to gain insights into the broader pet parent cohorts, thus helping to optimize marketing spends in our more traditional CPG pet nutrition business.”
Additionally, Mehrish says that Mars has spent the last four years building a sophisticated ROI measurement model that can compensate for the resulting gaps in campaign reporting data. “Where we will lose visibility is on the end-to-end consumer journey,” he notes. “And we’ll have to build up that visibility in a different way through the ROI engine.”
Even DTC brands with much deeper access to first-party data are making significant adjustments to adapt to the cookie-less world. “We are already working on having alternative ways of evaluating media that do not rely on cookies and are engaging in more meaningful projects and infrastructure to look at customer journey insights,” says Courtney Fadjo-Biro, vice president of growth marketing at Zenni Optical, a DTC eyewear brand. “We are testing new marketing tactics, developing new ways to understand media performance, investing into our first-party data and leaning into the data that we do have available to us. We also keep a close eye on developments within the ad marketplace that help with programmatic spending adjustments.”
The New Privacy-First Terrain
The push to eliminate cookies is part of a broader effort by privacy advocates to increase transparency in data sharing on major ad platforms like Facebook and Google, and it is being supported by government privacy laws including GDPR (General Data Protection Regulation) in Europe and CPR (Consumer Privacy Act) in California. Unscrupulous parties will still be able to track web user behavior with lesser-known technologies, such as LSOs (local storage objects or “flash cookies”), which essentially do the same thing as cookies.
Google has said it will not introduce an alternative user-level tracking system for the cookie. The company is banking on its new Federated Learning of Cohorts (FLoC) technology — which creates a lookalike audience that protects data privacy by using machine learning algorithms to process web browsing behavior on individual devices — to drive conversions at a similar rate. Digital media experts say the level of targeting precision will almost certainly diminish under FLoC. “It’s definitely a step back in the marketer’s ability to personalize,” says Tim Glomb, vice president of content and data at Cheetah Digital. “Google is basically saying, ‘Trust us. We know people and what hobbies they’re into. You don’t need all that filtering and segmenting data.’”
In the mobile marketing space, Apple recently replaced the common tracking device known as IDFA (Identifier for Advertisers) with an opt-in data sharing system on its new iOS 14, a move that industry observers say has dramatically reduced participation rates. Many experts believe the confluence of these developments is likely to hurt digital/mobile campaign performance in the short term. “The near-term impact will be a significant hit on marketing efficiency, increase in CPMs and loss in marketing-driven revenue,” says TJ Reilly, executive vice president of performance media at VMLY&R Commerce.
Attribution also will become more challenging. Increasingly, so-called deterministic attribution models that require access to user-level data are being replaced by probabilistic models that rely on aggregated sets of privacy-safe, anonymized data. As Mehrish explains, “I will not be able to deterministically say that ‘John Doe’ saw an ad on Hulu then went to The New York Times, and then I showed him my display ad and then he went to my website. But because my ROI engine is so sophisticated, I’ll still be able to determine if The New York Times is good enough for me to keep investing. I’m solving for the [ROI question] through probabilistic means, not through deterministic journey maps that I used to do earlier with third-party cookies.”