It was inevitable: in 2020 Google announced that it will no longer allow third-party cookies on its Chrome web browser as of 2022. In doing so, Chrome joins other major browsers like Safari and Firefox that have already done so, with Safari being the first to do so with its Intelligent Tracking Prevention (ITP) in 2017, and to which it made continuous tweaks since. As of 2020, Safari does so by default, having joined the likes of small, more privacy-obsessed browsers such as Brave and Tor.
These cookies (read: trackers) have become a central target of online privacy advocates and even ordinary people who find them invasive and even somewhat creepy, which is why browsers have felt compelled to introduce these measures. It is well-known that they track and capitalize on people’s online habits and footprints, and even their personal data.
Nevertheless, the stakes are immense: as of 2021, third-party cookies remain the leading driving force for advertising seen by those surfing across a vast number of websites on the web. This development by Google could possibly be akin to an ‘ad-tech Armageddon,’ especially for those smaller businesses that have been heavily reliant on 3rd-party cookies to date. This article will briefly explore how artificial intelligence (AI) should help bridge that gap.
Google’s Cookie Domination
Google has a resounding share of the online advertising market, at nearly 70%, even with recent gains by the likes of Amazon. Google’s approach to banning these cookies has been somewhat more conservative compared to Safari and Firefox, and the tech giant is at pains to stress that it wants to protect the adtech ecosystem. It iterates this stance by stating, “Advertising is essential to keeping the web open for everyone, but the web ecosystem is at risk if privacy practices do not keep up with changing expectations.”
Even so, Google’s blocking initiative does seem somewhat risky. After all, as of late 2020, 55.2% of all websites on the internet were using Google Analytics, which is predicated on third-party cookies, and which is a huge revenue generator for the company.
AI and Audience Modeling
Audience modeling, also known as look-alike modeling, uses AI, usually in the form of machine learning (ML), to identify people who behave online in a way that is aligned to a specific target audience or consumer group. The focus is on slotting a person into a certain group of similar users based on their browsing history and other online factors. That should be different to the more precise, individual (and invasive) ‘profiling’ that occurs with third-party cookies. It is segmented and persona-based profiling that should result in high-value online marketing, as Adobe Africa points out. It’s considered the “new gold standard” in behavior-based analytics, aka Segmentation, Targeting and Positioning (STP) marketing.
Worth noting is that Google will itself be using audience modeling with its ML-based system, the Federated Learning of Cohorts (FLoC), which it refers to as a ‘Privacy Sandbox’. Google claims that adtech benefits will still exist with the new scheme, since specific groups of users will be directly targeted for advertising purposes, whilst ensuring that people’s individual data and privacy is better protected.
AI and Content Scoring
Content marketing can be very tricky. How can one evaluate the performance of given content compared to that of competitors in the same industry? The bottom line: online marketers spend up to 70% of their time creating content that only has a 30% hit rate. Even the ‘70”20:10 rule’ as a marketing model is problematic in terms of accurately gauging the success rate of content being generated.
AI can greatly assist marketers in better understanding the nuanced persona of their audience and, therefore, attain a better idea of what content to craft so that content performance and ROI improves. Natural language processing (NLP) is invariably at the core of this form of AI-assisted content scoring. Content can be tagged and categorized hyper-specifically using the most salient keywords, to then be auto-clustered into appropriate subsets of topics. Social engagement with content, for example, can then be attributed a topic-based content score, which can then further inform and shape new content creation.
We at Genus AI like to say that, thanks to this form of AI content scoring, digital marketers today have no excuse to continue running “be the loudest in the crowd”-style social campaigns that are frankly outdated and very often highly inefficient. We know that this technology in the form of our Customer Modeling Platform can help online marketing campaigns be more emotionally intelligent, personalized and smarter vis-à-vis their targeted customers. Furthermore, this achetyping or persona-building can be achieved without infringing the privacy if customers.
AI and Content Generation
How to generate content can be as vexing for digital marketing teams as what content it should be. Experts acknowledge that AI-generated content cannot yet fully replace content crafted by human copywriters. However, what AI can do (and, increasingly, very well) is help provide direction and strategic context regarding what content to post how, where and when. It achieves this by providing invaluable, instantaneous customer profiles based on their interests and purchasing journeys. AI is also invaluable for its insight regarding marketing trends based on predictive intelligence modeling. The result is nuanced customer personae that can help direct needed content, including directed promotion thereof.
There is a lot to be said for technology that relieves marketing teams of essential time-sapping and exhausting content development tasks, such as keyword-based research, content optimization, the building of content briefs and strategies, and the like. For example, for given content AI can be used to analyze the topic, contextualize it relative to allied topics and existing content and then generate a usable draft for a writer or editor to then further hone as needed.
The 2022 ban of third-party cookies on Google Chrome will not mean the end of tracking that is so important to the adtech industry. Remember: the internet cannot survive if all advertising on it were to be decimated. Instead, the nature of that tracking will segue into a more nuanced form of persona-building based on customer traits aligned to similar trait-based groups.
As shown, the potential of AI in audience modeling, content scoring and content generation, to name but a few derived benefits, is already immense. It will no doubt only continue to do so exponentially. The belief is that this AI-enhanced form of tracking will allow online advertising to continue to flourish, albeit differently and more intelligently, whilst better protecting the online data privacy of users. If so, that should be considered a win-win scenario.