By Praveen Balla
Vice President, Production, True Influence
Targeting Audiences, Not Individuals
With Google FLoC, display advertisers will have to target small, closely related groups of prospects, not specific individuals. This may prove problematic in some down-funnel B2B applications – which I will get to a bit later – but it may be something of a blessing in disguise at the early engagement / brand phase of campaigns.
Our COO, Craig Weiss, has often commented that programmatic display is a great, cost-effective channel for sizing market demand and finding accounts and in-market individuals who aren’t a 100 percent match for your ideal customer personas. Viewing audiences in terms of 1,000-member cohorts may actually be a useful constraint; marketers will have to think about the primary qualifiers for a prospect, and not refine targeting to an implausibly fine point.
The details are not entirely clear, but I can only imagine Google will continue to support overlays of cohorts to create custom targeting segments. This kind of audience building has become table stakes in programmatic, and again, Google is not going to compromise its main business model in such a dramatic fashion.
Re-targeting to groups is a little trickier
At this point, it seems that retargeting will be accomplished by re-assigning anonymized users to new cohorts based on activity, not by sending follow-on creative to a distinct individual. This activity could be indicative of advancement along the purchase journey. But under FLoC, ads are always targeted to a group, not an individual, at least as Google now describes its intentions.
It’s a fair assumption that at least initially, Google will be making market-based refinements based on customer demand and offering those to the market, as opposed to creating unique cohorts at the behest of customers. But there are other components of the privacy sandbox, most notably FLEDGE, that appear to be devoted to analyzing brand loyalty – an essential factor in B2C retargeting.
For B2B, not being able to re-target a time-sensitive offer (say a webinar registration) to a specific individual is less than ideal. Lifting multi-channel campaigns is a key application of display, and that will be compromised, at least to some extent, by FLoC’s insistence on anonymization to a group.
Some version of intent will be required
At the moment, Google has not said that inferred data, including purchase intent signals, will be built into the FLoC cohort assignment algorithms. In fact, Google has stated that it does not plan to engage in identity triangulation that maps purchase research activity to individuals or accounts.
However, it’s hard to imagine that Google will completely leave this powerful – and for B2B, fundamental – targeting mechanism out of the mix. Brand affinity is a kind of purchase intent, and it is working on solutions in that category, so I can see Google’s AI team announcing a version of purchase research monitoring that it uses to create “in-market” anonymized cohorts for B2B.
Google may well stick to its pledge to not target ads one-to-one (although some observers point to the TURTLEDOVE “interest groups” component of the privacy sandbox as a backdoor to this pledge). But there is zero chance anyone is going to go back to the dinosaur ages of online advertising. So some form of intent and inferred data will almost certainly be baked into Google’s plans.
I imagine that these cohorts will be less effective than what we are able to accomplish with our Identity Graph and other inferred data technologies, particularly at the account level. This level of group anonymization may well also not support Buying Group roles and other key B2B personalization criteria.
First-party data is king
One thing that’s clear is first-part data is going to become more critical than ever. As a programmatic ad service executive notes in this piece at CMS Wire, aggregated, third-party targeting data is most useful when it is “elevated” by a clear understanding of explicit first-party behavior.
Previously, the connective tissue here has been cross-platform cookies and identity triangulation. In the new world of the privacy sandbox, this may mean building out internal AIs to build your own cohort behavioral models and then mapping those to segments offered by Google and similar black boxes.
Marketers will also probably focus more on high-quality content offers and other tactics to prompt first-party opt-in early in the purchase journey. And well-established channels, such as content syndication, may become even more critical for building out a broader first-party contact database.
Trusted second-party data marketplaces may become more popular
Another tactic for those not willing to stick with just Google may be building custom audience segments through broader use of data marketplaces, where publishers and other partners offer audience segments on an a la carte basis.
Data co-ops have been the basis of plans by programmatic DSPs to build universal ID platforms for ad personalization in the looming face of third-party cookies’ demise. But competition for online ad dollars is intense, and getting publishers and other data owners to sign on to a truly cooperative structure has been a roadblock.
So offering ad hoc access to opt-in, second-party targeting data may emerge as a viable alternative.
Advanced attribution will become a little trickier
Of course, you can always tag landing pages and other conversion mechanisms to clearly show that a given ad creative resulted in a direct conversion. Quantifying the less direct impact of a specific CEO viewing a pre-roll video on her smart TV will be much more difficult without one-to-one ad targeting.
The answer here is still unclear, but will likely involve some probability analysis of the prospect being in a given cohort. It’s less precise, but it’s a concession that B2B marketers who want to allocate programmatic spend outside the Google ecosystem will likely have to make.
It’s still all about audiences and data
Google’s moves to own most of the internet’s behavioral ad targeting data is obviously going to shake up B2B marketers’ daily operations a bit. New and better solutions for protecting consumer’s privacy may emerge, or Google may itself back away from some of the more restrictive aspects of its current privacy sandbox initiative.
But in the coming world without third-party cookies, B2B marketing and sales will need to aggressively convert prospects into first-party relationships, then use the behaviors they present to model out the ideal audience segments they need for successful display ad targeting.