Apple’s decision last year to add privacy features to its IOS 14.5 browser, and force iPhone app users to opt into tracking them, continues to be felt throughout the digital advertising industry as it heads into 2022.
But even with the dust not quite settled, a few things are starting to come into focus. Walled gardens are beginning to meet their downside. You see, the problem with big walls is there is no transparency to what is going on behind them and even less interoperability between the walls themselves. It turns out brands don’t like that.
Being able to cleanly associate an audience from lead to conversion as they pass through and across multiple platforms is a pretty big deal. A consumer journey from lead to conversion can take more than one path. An identifier like the cookie, mobile ID, email and location is needed to navigate search, social, display, connected TVs, foot traffic and CRM platforms. It all counts. The idea of sampling in one walled garden and modeling a forecast across another is probably not going to fly with brands.
Facebook (now Meta) seems to have quietly given up after taking stock of the fact that Apple’s device footprint is down to 30% of what it was. Deciding that most of their base is on mobile most of the time, they have no path to execute deterministic attribution without the iPhone identifier. Without a lot of fanfare, they seem to have pulled back until will remain that way until a solution emerges.
With no persistent Apple ID, advertisers are still able to use the Apple-controlled system that is designed to keep the user in the Apple ecosystem. Apple’s SKAdNetwork is a point solution that proponents claim can enable tracking based on data points other than user-level identifiers. However, that means advertisers are stuck cobbling together the Apple attribution option with Google cohorts and Android device IDs.
Companies are rushing in to try to fill this breach, with solutions that leverage modeling augmented with post-attribution data from any and all sources. Most of these new solutions both remain unproven and rely on modeling, which means many brands are going to tread carefully. The concept of modeled probabilities that are directionally correct doesn’t necessarily mix well with board room planners and accountants. A deterministic match is proof they need.
What is missing in the attribution discussion is the direct impact it has on campaign optimization. Optimization made harder will result in spending inefficiencies and that will result in higher costs to achieve the same results.
In order to counteract this inefficiency, advertisers will slowly adjust their buys in favor of a more predictable quality of audience which they can find with premium publisher inventory.
Publishers to the rescue
Content that attracts premium users brings certainty to the advertising investment. As this shift occurs, publishers are beginning to realize they have a precious commodity they may not have exploited well. Publishers are noticing that they are now able to leverage their first-party assets to help individual brands execute narrowly on campaigns within their sites. The less obvious is in the near future, where they’ll also recognize they have data assets they need to successfully participate in buyer/seller consortiums.
Premium publisher consortiums will end up being repositories where first-party data from a group of like-minded brands are combined for better marketer outcomes. Similar to what took place during the heyday of the direct mail industry, brands that have the ability to leverage deterministic behaviors and ownership inventory that hasn’t been diluted by look-alikes and mush that comes from an endless supply of unqualified inventory. Imagine a digital advertising ecosystem that has been purged of wasted impressions.
Publishers are developing the tech stack to make the first move, but brands are starting to follow. Brands are moving quickly to digitize their customer base, enhance their profiles and monetize their base. Brands especially will become effective at correlating their customers with the right publisher co-op channels spurring more online ad spending against those customers.
Modeling and cohorts may not be a good fit for 1st party publisher and brand audience graphs. But the fatal flaw that all adtech, social and platform execs share, is that they fundamentally believe they can almost always invent their way out of any technical challenge. Advertisers forced to settle for targeting a cluster or a cohort of individuals after years of people-based targeting is going to be tested into oblivion. Evidence suggests that a technological solution based on modeling, cohorts and opaque walled gardens may not be enough. Our best shot may be the 1st party publisher and brand identity movement that is not dependent (or controlled) by any single identifier but rather exists as a constellation of identifiers distributed across the ecosystem of cooperating publishers and brands in a privacy-safe manner. This may be the silver lining to the attribution problem.
CEO & Founder, Semcasting
Ray Kingman is the CEO and co-founder of Semcasting, a data-as-a-service (DaaS) provider. He leads the company in the development and commercialization of its automated targeting and data offerings. As an experienced innovator in content management, analytics and data visualization fields, Ray directs the day-to-day operations of Semcasting. His extensive experience working in the marketing and advertising industry field allows him to speak confidently on matters of consumer data privacy, identity resolution, brand empowerment, customer acquisition and digital and online marketing.