Quality, volume, automation & privacy: too good to be true?

Quality inventory is not determined by semantics, keywords, or sentiment. For all the value these may have when it comes to avoiding brand risk or ensuring relevance, they can hardly replace manually-verified taxonomies or editorial guidelines in the context of quality media. It follows that, in isolation, in-page context analysis (and context-based activation) is better suited to maintain the sub-optimal balance currently found in the open ecosystem: volume and automation at the expense of quality.

Could we possibly aim for a perfect alignment of all three factors? Automation and speed are now available to premium outlets or niche publishers through publisher-defined cohorts. These are informed by valuable first-party data about preferred topics or formats, and are likely to provide a better match between a given brand’s aspirational values and the inventory on offer. They may also allow advertisers to define more precise look-alike audiences on the basis of reliable, interoperable content taxonomies. 

How about scale? Pooling a myriad of walled “micro-gardens” together would never satisfy the volume requirements of media agencies and advertisers spoiled by Google, TikTok and Meta. On the other hand, integrating the long tail of display and video inventory could easily bring us back to the same dilemma: a few outlets doing all the heavy lifting, seeing their efforts diluted in the vast wilderness of the open market.

Which probably means that future-proof advertisers will have to embrace a richer, more complex, and more diverse media landscape that will sound all too familiar to the industry’s old guard.

What about privacy? Is it not the primary factor underpinning this new scenario?

What you really, really want

Contextual is just another avenue to ensure relevance, and relevance boils down to the consumer’s interests and needs. There is little departure, in this regard, from behavioural targeting. After all, the latter was never about chasing people around for the sake of some dark social engineering motives, but rather about approaching them at a particular point in their own customer journeys. The privacy challenges associated with it are just an unintended - if foreseeable - consequence of granular deduplication across multiple touch points and actors.

When building a meaningful brand or satisfying a particular emotional or practical need, understanding what we read and watch goes hand in hand with mapping our values and preferences, and this calls for varying degrees of automation. Whether powered by in-page semantic analysis or not, first-party cohorts provide a privacy-friendly means of doing just that. The same could be said of Chrome’s Privacy Sandbox’s attempt to transform content affinities into a new sort of addressable interest graph (under its Topics API initiative). 

In other words: In the latest sign of growing customer centricity, an individual’s shared browsing history is simply giving way to advanced statistics and context-powered targeting. It could also one day lead to a truly demand-driven, zero-party data approach in which said interest graph never abandons the individual’s sphere of control.

All together now

Insofar as it avoids individual profiling, contextual advertising is paving the way for a new kind of marketing that does not need to rely on ongoing requests for individual consent (to something that people cannot possibly understand). 

At planning and measurement level, the absence of granular data shouldn’t result in a major drama, as aggregate data sets and a new generation of cross-device audience panels feed into data clean rooms and renewed media mix models. 

But bold advertisers could take the opportunity to address a bigger elephant in the room: Most ad formats will be just as insufferable or invisible once one-to-one targeting is replaced with anonymous cohorts. Ad blockers will still hide them. People will still hate them.

Customer centricity and individual empowerment go hand in hand with privacy safeguards. We should rethink media and advertising experiences accordingly. Along with the removal of annoying cookie banners (which will naturally flow from these new means of targeting), asynchronous ads and conversational features should produce much better results at all levels.