Online advertising uses first-party cookies and third-party cookies. First-party cookies are stored on your computer by the website you are visiting while third-party cookies, as the name suggests, are placed by their partners. Third-party cookies can be read by not just the original website you visited, but also all sites using the third-party partner’s code. Third-party cookies have become the main mechanism to track users across the web and are the workhorses of the multi-billion-dollar global ad-tech business. Third-party cookies are the key connective tissue to maintain your entire activity history and not just drive targeted advertisements but also used in measuring the effectiveness of digital ads.
In the last few years, rising consumer concerns about privacy have led to governments introducing stronger data protection policies that restrict how consumer data can be collected, stored, and used. For example, the GDPR law in Europe includes ‘cookies’ in the definition of personal data and imposes restrictions on how they can be used. Internet browsers are now blocking the use of third-party cookies. Apple Safari and Firefox have already restricted them, and Google Chrome will be phasing them out by 2022.
To sum up, the ground is shifting for digital advertisers on the open web. They still have to deliver the “right ads to the right audiences” and find cookie-less solutions. The answer (or a large part of it) lies in AI.
Note that third-party cookies won’t fully go away but companies have to get clear consent from users before using them. Instead of each website having its own user ID, some companies are taking a “Unified ID” approach — using the same ID across different sites — so that users can be tracked across their network. But this does not really feel like moving away from cookies, when consumer sentiment seems to be clearly turning against them.
AI enables omni-channel
An alternative approach, called “ID resolution”, is based on AI. User identity can be determined by putting together different slivers of data about the user, and machine learning can help create the user identity graph by piecing together all the user data on hand.
Note that some channels such as in-app mobile ads, audio, TV have not used third-party cookies. They won’t be affected much going forward, but they too have been silo-ed. Omni-channel identity resolution involves matching offline to online channels and across devices. AI can help integrate these channels to deliver true omni-channel targeting by using anonymous identifiers.
Return of contextual advertising
Advertisements can also be contextually targeted — not relying as much on personal data but based on what and where they are viewing the content. AI can be used to analyse the site/page and generate relevant ads.
The idea of contextual advertising — serving content that is relevant to your brand and in a place where users are consuming similar content — is not new but has been overlooked in favour of cookie-based targeting. AI enables the next generation of contextual targeting in a variety of ways:
Analyse the website content and determine the sentiment; analyse the image and video content and determine the tone; use natural language processing to analyse any user-generated content; and, finally, dynamically generate different creative elements based on context thus determined.
Further, AI can help integrate real-world data (such as weather and current events) and increase contextual relevance and richness without using cookies.
AI for better measurement
Another looming important change is in the advertisement measurement, which again is currently dependent on cookies. Conversion modelling and marketing attribution are domains that make heavy use of machine learning to give you a better view of the entire omni-channel customer journey to better focus your marketing and advertising efforts.
In a sense, third-party cookies have become the “sugar addiction” for the ad-tech world. With their phasing out, and the maturing of AI technologies, now is an opportune moment for the industry to build better approaches and techniques that respect the privacy preferences of consumers.