In the first part of this blog series, I shared some background about consumer data privacy legislation, and how it got to this point.
In this blog, I will share some tactics you can begin incorporating into your programmatic paid media strategies, as well as explanations of up-and-coming solutions marketers should familiarize themselves with.
So, let's dig in. What can marketers do to begin preparing your teams for these upcoming privacy changes?
First Party Data
The key to any successful digital campaign is quality, clean, consent-given first party (1P) data. There is no replacement for good 1P data - even those 3P blackbox audience segments marketers purchase from DMPs for prospecting purposes are often of poor quality because we don’t know what’s in them, according to the Harvard Business Review
. Brands that utilize 1P data for activation will be far ahead competitors in terms of addressability and campaign efficiency and permits algorithms to create better lookalike audiences. 1P data also allows brands to better understand their best customers, who has lapsed, who have only shopped with the brand once a year, etc., and then develop different strategies against each audience profile, solving for different outcomes while driving growth.
Most DSPs offer this out of the box or rely on whitelists, so for most marketers with access to a DSP, this will be an easy tactic to apply. However, there are ways to make contextual targeting even smarter. Oracle realized this opportunity in 2018 when it purchased Grapeshot for $400M. Companies like Grapeshot are unique
in their contextual targeting capabilities, allowing media buyers to decide in real time based on the landing page’s keywords, whether the buyer wants that impression. This solution still requires cookies, but in a compliant way
, and has historically been used for brand safety purposes to negate certain content, or to increase the accuracy of contextual matches. Embracing contextual targeting will be a cornerstone of many future programmatic tactics, especially as programmatic spend is forecasted to increase by several billion.
Google has made big strides in developing its cookie-less alternative, Federated Learning of Cohorts (FloC)
. FloC takes users and groups them into cohorts based on interests, so that individuals are unidentifiable from the rest of the users in the cohort. FloC is comparable in that way to Facebook which uses a similar machine learning technique called self-supervised learning
to build its audience segments.
Google recently released some results and claim that advertisers can expect to see 95% of conversions per dollar
spent when compared to cookie-based advertising. We won’t have to wait much longer to test it - Google is making FloC-based cohorts available in March for developer testing and for public testing in Google Ads next quarter
. Marketers should test Google’s FloC solution when available and compare results to its cookie-based campaigns to test the validity of Google’s conversion rate claims. There is mixed support for Google’s FLoC tech
Unified ID 2.0
The industry leading Demand Side Platform, The Trade Desk
, has spearheaded an initiative to create a single, unified ID based on hashed emails. The idea is that once a user visits a web page, they will be asked to provide their email address for the purpose of identification across a network of publishers that have adopted Unified ID 2.0. An encrypted ID is than passed along the bidstream. Only time will tell if there is widespread publisher adoption, which would be key to its success. TTD has reiterated multiple times that it does not own the project and that it is an open source collaboration amongst multiple partners
, which is distinct from Google’s unilateral approach. Some critics have argued that having any PI in the bid stream would still constitute a violation of user privacy laws and some companies are looking to build solutions that avoid this entirely, like LiveRamp.
Data onboarding company Liveramp, for instance, has announced their answer to cookieless tracking by building Authenticated Traffic Solution (ATS), a targeting method that can be described as this
“A visitor goes to a publisher’s web site and logs in or signs up for a mailing list. If a log-in, the email address is part of the logged in profile. This ATS solution only works for users whose email address is known, not anonymous visitors. As part of the login or signup process, the user is made aware that their info will be matched to LIveRamp’s massive IdentityLink database, which connects online data with offline data to create what the company describes as profiles of people, rather than simply trails of online browsing habits. Clinger said the default choice is to accept the connection of this email address to IdentityLink, although a user can opt-out. Opting-out would also opt-out for all uses of ATS across all publishers.”
The idea is that there will be “walled-garden” like targeting on the open internet made possible
, and for free. First-party authenticated (known) users who have logged in to a publisher’s website with ATS installed, will be able to be cross referenced against LiveRamp’s IdentityLink and identified (if opted in). If a user has no history with the publisher, and ATS is installed on the website, the user remains unknown. ATS works by creating a giant identity infrastructure that all publishers can access to connect and cross reference different sets of first party data
, similar to data bunkers. Liveramp promises its audiences will be built on 100% deterministic identifiers and will maintain the gold standard of data quality set by Facebook and Google. Lev has established partnerships with Liveramp to help serve our clients data needs and help power the performance of paid user acquisition campaigns in a more private, cookieless world.
Companies like Infosum
act as data bunkers where companies can store 1P data and reference it amongst other 1P datasets from other brands while never disclosing personal information. Gurman Hundal, CEO of programmatic agency MiQ
, suggests that data bunkers
will be an essential part of cookieless targeting. Hundal explains, “brands have a neutral, privacy-first space where that data exchange can happen securely, thus allowing both parties to make more of their first-party data without needing any third-party cookies.” Data bunkers may be a critical component of the cookieless solution moving forward because of the scarcity of behavioral data.
The only thing that is certain is that more than one solution will be viable and much testing needs to be done in every area of opportunity to drive the best results and bring value to clients. Advertising will continue to thrive, and projected programmatic spend is to increase YoY despite user privacy concerns, indicating advertiser confidence in the ability of marketers to adapt. Tim Cook is likely right -- the evolution of advertising is more user and privacy-centric and ads will continue to be served with or without troves of behavioral data. At Lev, we are preparing for this future by building the right relationships amongst partners, getting client first party data in top shape, and watching AdTech developments closely.