As marketers, we strive to “serve the right creative, to the right user, at the right time” but that’s about to get a little tougher in 2022.
As users become increasingly aware of the volume of data and the startling accuracy of their online “anonymous” profiles, it can feel like anything but anonymous and consumers are pushing back through the help of big tech. On the global scale, initiatives like GDPR (General Data Protection Regulation) in the EU and regionally in California, CCPA (California Consumer Privacy Act) and CPRA (California Privacy Rights Act) demonstrate the traction this movement has generated through legislation. Now the tech industry is being tasked with fixing a problem they ultimately developed, albeit I’d argue, in response to pressures from many different sources including those same consumers pushing back with data privacy concerns now.
Cookies, small pieces of code placed on your web browser that track your activity on a brand’s website, were initially created in the early 90’s as web developers wanted to differentiate between new site visitors and returning users. These cookies are referred to as 1st party cookies and were used to provide users with the optimal experience of the website. As the internet was driven by convenience and efficiency, it made sense to provide the user with every opportunity to allow them to pick up where they left off in their last visit versus having to start all over again. Imagine now if you filled your shopping cart or were 2/3 through your child’s FAFSA application and your computer died and you’d have to start all over again, not fun. As with all technology, upon its creation it isn’t good or evil, but how it’s used that will determine the perception.
Soon this technology was applied to 3rd party cookies or tracking cookies which were specifically used for data collection of a user’s behaviors and interests online, typically used in marketing and advertising. Initially this served as a mutually beneficial relationship, for publishers, brands and users. Publishers were able to monetize their audience data by allowing 3rd party cookies and passing back valuable insights on their users to brands who were willing to pay for it thus helping to keep the majority of the internet free to users via ad revenue. Brands were looking for ways to get smarter with their targeting and as users spent more and more time online, data accuracy and opportunity to serve ad inventory grew, delivering a more efficient consumer audience for the brand. Users, while not everyone will agree, were starting to get ads more in line with their interests and wants, a welcome change at the time from seeing completely irrelevant ads that felt so off it left an immediate bad perception of the brand and didn’t provide value to either the user or the brand.
As the model evolved, brands partnered with sophisticated media partners who could make sense of all this data. They employed machine learning algorithms mining user data to tell what users had done and who they predicted was the next in-market consumer. They got so good in fact, it became a little “too” predictive and the general audience started to take note. The main driver of the digital marketer’s success? Those dozens, sometimes hundreds of cookies on user’s browsers watching and reporting back their owner’s actions, anonymously but still a bit creepy. It’s important to remember that the removal of cookies does not mean you won’t see ads anymore. Ads will still be there, they just won’t be for you (specifically) anymore.
Fast forward to the present and change is upon us with 2022 being declared the end of cookie (as we know it). As we move forward to sunsetting the cookie a few interesting solutions are being offered in their place. No one knows just how this will affect publishers, brands and users as the pendulum swings back away from personalization to generalization. It should also be noted that this change won’t be as drastic as you might think with Apple’s browser Safari already blocking cookies by default as does the smaller browser Mozilla/Firefox. However, it’s the largest player, Google’s Chrome, that has set the deadline of 2022 which has marketer’s scrambling considering that makes up over 60% of web browser sessions. Interestingly Google has already informed advertisers they are adjusting their tracking to include a new “1st party Google tracking cookie” which is meant to improve accuracy and attribution for advertisers. So it seems like Google considers cookies bad if they aren’t Google cookies.
This theme plays out across the big tech players, commonly referred to as “walled gardens”. These are tech giants that have created an ecosystem where they don’t actually need a collective agreed upon framework to make their offering work. In Google’s case, they have Google Search, Google Display Network, Google Shopping and YouTube. When running a campaign with Google, they have access to all of this data about how users interact across their own properties- they just aren’t happy to share it with anyone not running an advertising campaign through Google. Same with Facebook and Amazon and to a lesser extent Apple when looking at the mobile App ecosystem.
Google’s actually gone a step further and is introducing a new targeting methodology to replace cookie tracking called “Federated Learning of Cohorts” (FLoC). While I love puns as much as the next guy, this take on “birds of a feather flock together” seems like a poor substitute for the level of accuracy marketers currently enjoy now. Google claims it will monitor users’ behavior and associate them with a cohort based on their behaviors along with others who also share those same sets of behaviors. What the classification will be based on, how many data factors are considered, or the target size of a cohort are all yet to be determined. We know users don’t all act the same, even if they look the same. There is a “next best” customer and it seems like that clarification is going away. It’s also not lost on me that those making the rules are likely to benefit from them. Brands have become so dependent on advertising that they just can’t stop because it becomes less effective. With it being less effective, will advertising costs go down? Doubtful. It will likely require more spend to achieve previous conversion volume levels but that makes sense as we are moving away from a more efficient targeting methodology to one that blurs users together. This also poses new challenges to stereotyping and discrimination that are outside the scope of this writing and also yet to be proven, however seemingly very possible as we move away from personalized marketing and require you to be homogenous enough to your closest peers before you can see ad content relevant to you.
Another solution being presented is contextual targeting which focuses on the content of the website that the ad inventory is being served on. For example if the website is reviewing tech products, through contextual targeting it would make sense for tech-centric brands (laptops, routers, etc) to advertise there. The problem with this approach is that it doesn’t tie in additional elements of user data so it can’t tell the difference between a database administrator who purchases 20+ tech products a year versus the high school student building his own computer for the first time. Contextual targeting misses out on the in-market data classification that occurs through stored data of behaviors and online interests that flesh out individuals from each other.
Another solution is registration, email data. Google has already come out against this methodology believing that this level of PII (personally identifiable information) will eventually be withdrawn from user’s willingness to share with brands and marketers. They may be right, but a lot of these websites that require you to sign in also have custom alerts and dashboards specifically for their users tailored to them and that convenience may be difficult for users to move away from. These publishers, and their rich 1st party data are often times at the core of advertising platforms that use the registration data as part of an “identity graph” they use along with other data points to form a holistic profile of users. Additional data identification is provided through the ad inventory itself in what’s called the “bid stream”, which passes through user location, device ID and IP address, the latter being something that is used in cross device campaigns to target the same user across multiple devices in the home. These other user identification techniques exist and more will likely be created as pressure from advertisers to find renewed efficiencies drive ad tech vendors to work harder and smarter or risk being deprecated themselves by the likes of Google and other walled gardens.