Can semantic targeting impose order on a nervous GDPR marketplace?
With all the talk of data-fueled innovation at CES 2019, there’s an excitement among marketers of what AI can bring to their toolkit. Yet, with the General Data Protection Regulation (GDPR) still looming large, there’s also an overriding feeling of uncertainty. Will Google finally adopt the IAB Consent framework and instill a baseline of consistency and order? Even if that goes well, there is no guarantee that we are on the verge of embracing the new normal that data regulations have sparked.
In fact, the growing pains are already underway. As evidenced by recent headlines, privacy watchdogs in Europe have filed a series of complaints and warnings against a number of companies including Vectaury, Criteo, Quantcast, and Experian for alleged GDPR violations. Here in the U.S., the specter of the 2020 California privacy legislation has added an additional dimension of confusion.
How will the programmatic ecosystem settle into this new environment?
One place to look for transformative impact is at the intersection of semantic targeting and audience targeting. Aided by the rapid growth of computing power (thanks, Moore’s Law), it is now viable to perform semantic targeting with greater cost efficiency than ever before.
This efficiency is having a democratizing effect in making it easier for all players to do page analysis in a more sophisticated way. Traditionally, semantic targeting was applied to help avoid brand safety issues, and in the intervening years, AI has now made the technique even more sophisticated and flexible (and less labor-intensive) whereby it could even be combined with audience targeting in the same impression for next-level targeting.
Presumably, when Oracle bought Grapeshot, one of ad tech’s early contextual targeting players, in 2018, they were looking beyond the brand safety benefits in its offering. Oracle likely realized what the value proposition of a leading-edge semantic targeting technology could deliver in combination with their existing third-party data assets. Part of the company’s calculation could presumably have been an example of the broader programmatic marketplace girding for the chilling effect that GDPR might create in audience targeting.
The improvement in semantic targeting efficacy over the past several years couldn’t have come at a better time. It has coincided with tech stacks becoming better integrated with multiple data sources and that data being blended in a way that should be GDPR-compliant.
A greater cross-section of consumer signals can now coalesce to give the ecosystem a more sophisticated view of the audience. CRM data is more readily organized and available. Advances in shopper look-a-like modeling now make it easier to discern individual shopper path-to-purchase in a more sophisticated manner by matching them with their contextual digital footprints. For example, it’s possible to arrive at a much richer profile of C-Suite types who buy Mercedes by cross-matching their visitation to particular websites.
Let's be honest. The programmatic community worldwide has been scarred by the multitude of issues in the past several years beyond privacy, such as fraud, viewability challenges, and brand safety. As a result, contextual semantic targeting combined with first-party data sets — including purchase and location data — offers the industry a safer alternative to the shortcomings of third-party data.
Is third-party data dead? No, but much third-party data will likely be collected in a way that will not stand the test of GDPR compliance, at least for the foreseeable future.
Most interestingly, the renaissance of semantic targeting is also a huge opportunity for publishers. The buy side has traditionally used the technique more often, but as publishers fight hard to reassert themselves against the blustery headwinds created by the duopoly and Super Publishers such as AT&T and Amazon, it’s incumbent upon independent premium publishers to beef up their semantic targeting capabilities.
The pendulum is swinging back to publisher brand value — and when publishers can offer more relevant, dynamic, and safe content, it is clearly to their benefit. It will give them one more powerful arrow in their quiver as they aim to hit the bullseye of knowing their audiences. Premium publisher data has been underleveraged over the years, but now that dynamic is starting to shift in the other direction.
Publishers have historically allowed DSPs to cherry pick audiences, thereby relinquishing responsibility for curation their inventory. Now because of GDPR and diminished cookie use, publishers would be well served by taking that mantle back. “We know our audiences and content better than anyone else,” should be the rallying cry.
I certainly don’t want to belabor the uncertainty that GDPR has cast over the marketplace, but it will continue to have a sizable effect on how business is conducted. Regardless of how it shakes out, I think we can all agree that the programmatic space needs to pivot away from pure audience targeting and create a more balanced and blended process for targeting. That means that context, data, and great creative should all play prominent roles in engaging consumers in this age of GDPR.
Michael Nevins is chief marketing officer of Smart