From data sharing via ‘clean rooms’ and publisher cooperatives, to contextual advertising or AI solutions, Tim Conley explores the post-cookie future
Google’s announcement that it will phase out support for third-party cookies in its popular Chrome browser created quite a stir. For anyone watching the space closely, well aware of the cookie’s unsuitability for long-term user tracking and targeting, it came as no surprise. But there is still much discussion as to what happens now.
So what could a future without cookies look like?
Contextual advertising solutions
In digital’s early days, content was king – along with context. Advertisers bought media based on a publisher’s audience, which was determined by the site’s content; ads were relevant to the user’s immediate interests rather than based on the historical behaviour that cookies tracked.
Eventually the rise of cookie-based retargeting and its capacity to drive greater performance saw audience succeed to the throne. It mattered less where someone was reached, so long as it was the exact user the advertiser wanted to target.
With the cookie’s death now certain, the targeting pendulum appears to be swinging back toward contextual; marketers seem ready to (re)acknowledge that relevance matters, and delivering a message in the right place has a certain magic. Trying to sell someone running shoes while they are researching their genealogy doesn’t have the same impact as showing them an ad for the same shoes while they’re researching running trails.
Moreover, contextual targeting has not been stagnant; advancements in this space include faster and more granular content classification, allowing advertisers to make use of the news cycle and take advantage of trending content, for example.
More work lies ahead to augment context with sentiment to enable a more nuanced approach. This will ensure advertisers wanting to be associated with basketball content (as an example), but avoid appearing next to news about Kobe Bryant’s recent and unfortunate death, can do so in an automated and scaled way.
Publisher and advertiser initiatives
With no third-party cookies, first-party publisher data will increase in value and importance. Currently, advertisers define which of a publisher’s users fit their audiences. In the future, publishers will hold the reins, using their own data to match their users with an advertiser’s audience.
This move will see a few trends emerge:
– Redefinition of audience segments: In order to scale, publishers will need to align on labelling the audiences they now define. Beefing up the current IAB industry category lists is one option, but it’s also likely that third parties will work with publishers and advertisers to expose audiences in other ways.
– First-party data consolidation: Publishers will work with advertisers and technology partners to provide consistent cross-network audiences tied to their (currently under-utilised) first-party cookie, as well as any other data the user has offered.
– Cookie-less identity validation: The cookie-less identities that currently exist are often only reliable for a very short time period. But combining them with publishers’ robust first-party data reinforces their validity to create a targetable profile – and provide a proverbial nicotine patch for attribution-obsessed advertisers.
Data sharing via ‘clean rooms’
The concept of data ‘clean rooms’ is not new; clean rooms allow advertisers’ first-party data to be combined with other data sets, typically from a walled garden, in a privacy-safe environment to derive insights and measure performance. Neither side is allowed to use the other’s raw data outside that room.
Likewise, publishers (or publisher cooperatives) and advertisers could utilise private data clean rooms to derive new insights and value from their user data. This could potentially empower one-to-one targeting, attribution between brands and publishers, as well as cross-device insights – all without data security being compromised.
Future development will tackle the current challenges such as privacy risks preventing advertisers sharing detailed transactional data. Clean room projects therefore need to prioritise data governance, designing in restrictions that determine the type of data that can be accessed in the clean room and by whom.
AI and machine learning
This topic is open to speculation, but it is widely-accepted that machines can learn a great deal about users by analysing anonymous data signals, such as frequency of website visits or the device used to access a website, with the model continually fine-tuning through self-learning. Processing that data in huge volumes allows cohorts of users to be created, against which it is possible to make generalised predictions.
For example, people living in a particular postcode who access the web from an Apple device using iOS 13 or higher and visit Forbes.com might be five times more likely to apply and be approved for a credit card. Buying models that target and optimise spend to reach audiences identified in this way can be built instead of targeting specific users via cookies.
Obvious concerns arise around attribution, particularly how publishers get their fair share of the credit, but a cookie-less future makes this a viable way of audience targeting so expect measurement to catch up to ensure last-click attribution doesn’t become the norm.
An identity-based future
The writing has been on the wall for third-party cookies for some time, but that doesn’t mean finding alternatives will be quick or easy. What is clear however is that advertising’s appetite for digital media trading will fuel the drive to find non-personalised, identity-based solutions.
These will potentially evolve from several corners of the ad tech ecosystem in order to meet the equally important demands for advertiser returns on investment, publisher revenue and consumer privacy.
I look forward to catching up with colleagues a few years from now and asking “Remember when we thought we needed cookies?”
Tim Conley is client services director Europe at IPONWEB