PM360 asked analytics experts about the impact of Google, Safari, and Firefox all phasing out third-party cookies and the biggest missed opportunities for marketers regarding data. Specifically, we asked them:
- With third-party cookies crumbling, how will a marketer’s ability to target their campaign, measure attribution, and/or use programmatic buying be impacted? What other areas of marketing may be affected?
- What are the biggest opportunities in terms of data and/or analytics that not enough companies or marketers either know about or are leveraging to their full potential?
In response to the demise of cookies, we recommend advertisers take command and control of relationships by focusing on their “private” first-party data. A brand’s data is like a gardener’s soil; it’s the single most important factor—within the advertiser’s control—to grow, nurture, and establish positive, sustainable relationships with their customers. Advertisers should follow three steps to adjust to the changes.
Step 1: Build a first-party identity graph. The best way to prepare for an unpredictable future is to first focus on what you can control. By building a scalable and configurable private identity graph, brands can set themselves up for success despite constant technology, consumer privacy, and vendor changes. Identity is the cornerstone of any brand’s audience, analytics, or measurement needs.
Step 2: Connect to private partnerships with publishers. Establishing direct relationships with the right publisher partners will allow advertisers to leverage the identity coverage of the worlds’ largest media brands while building on their foundation of first-party data for insights and activation purposes.
Step 3: Leverage new referential PII-based consortiums. Brands can use emerging PII-based consortiums for expanded reach and integration within the programmatic ecosystem (DSPs, SSPs, DMPs). It also offers a cookieless alternative for redirecting site visitors across participating publishers.
Personalized Creative: The rise of programmatic buying has placed too much faith, focus, and effort on audience targeting, often at the expense of creative. But this Nielsen study proves that creative is about 5x more important than targeting when it comes to the impact on sales. The end of cookies should lead smart marketers back to a focus on creative/content/messaging. Rather than making one or two ads that are served to everyone, make dozens and dozens, each designed for a specific type of user at a specific point of interaction with the brand. To fulfill this objective, marketers will benefit from technology that efficiently produces assets at scale, such as Ad-Lib.
Powered by a CDP: There are no silver bullets when it comes to martech, but a lot can be said for the promise of CDPs. With a CDP a brand can organize, understand, and act on every interaction with its users, whether they take place digitally or in the real world. This is an asset the brand owns, and which can be used to inform the delivery of messages users are most likely to value.
Over the next two years, programmatic marketers will need to evolve beyond using targeting methods that depend on cookies. Changes will impact HCP and patient campaigns differently, with key challenges and opportunities in each.
HCP target list buys on endemic websites will flourish as these tactics often depend on NPIs instead of cookies. On the opposite side, HCP non-endemic buys are primarily cookie-based and will need to pivot to alternative approaches, such as using CRM data activation with large publishers who have login data.
Patient marketing will experience growth in keyword contextual and private marketplace tactics, as these approaches don’t depend on cookies. Also, similar to HCP programs, it’s likely CRM data activation will become another replacement. The growth of CRM targeting will signal a rising importance of large publishers, who are able to protect their content behind a login. An unintended consequence of this change may be a decline in independent content, if smaller publishers aren’t able to monetize as effectively as before.
To prepare for these changes, marketers should audit current media plans. This exercise will be helpful in demonstrating that pharma marketers already employ many non-cookie-based strategies, such as NPI targeting, CRM data targeting, and contextual approaches.
Healthcare marketers spend approximately $4 billion per year on digital advertising employing data sets to reach consumers and HCPs. Most companies syndicate data sets through third parties, which increases the risk of intermediaries combining data sets and re-identifying individuals. With this year predicted to be the first when “connected TV” viewership will surpass traditional cable television in reachable U.S. households, marketers have an unprecedented opportunity. However, they need rich, privacy-safe data sets to identify and reach HCPs and patients.
One of the biggest missed opportunities is to employ differential privacy, which is considered “the new gold standard for data privacy protection” by the U.S. Census Bureau. When combined with artificial intelligence (AI), the technologies have the potential to extract critical insights from very sensitive data sets without the risk of protected data being used to identify a patient. In essence, it allows marketers to build an AI system that operates in a data “clean room,” where personally identifiable data is inaccessible to any human. The AI can scan through anonymized patient data sets located behind a privacy firewall that protects the data from leaving the “clean room.” Together, the technologies have the ability to fully leverage sensitive health data to help build models and extract insights about certain conditions and optimal treatments.
1. Data Readiness via an On-Demand Analytics Ecosystem to allow speed to market of advanced analytics. This includes consolidation of big data, promotional activity, descriptive data, and advanced analytics beyond basic data repositories.
2. Deploying platforms and analytic engines to drive advanced analytic scale moving beyond point-in-time analyses into real-time insights that shape marketing and sales initiatives. This allows for consistency of recurring analytics while freeing time from advanced analytics resources to solve new business needs.
3. Linking analytic outputs to CRM platforms to ensure actionability. Transition from static business rules for promotional delivery, which are already dated by the time they are deployed, to analytics-driven execution. The above-mentioned data readiness as well as scaling of advanced analytics allows for real-time inputs (segmentation shifts, promotional impact and affinity, messaging impact, channel cadence/frequency, etc.). This info allows for true personalization of how we engage with customers along their journey.
Maturity of the areas above requires a cohesive team working towards the same business outcomes across analytics, marketing, sales operations, IT, and the right analytic partners.
The proliferation of AI and machine learning applied to marketing is staggering. Understandably, the easiest way dip your toe in the AI/ML water is with structured and clean data sets; however, I would suggest organizations do not ignore semi-structured and unstructured data. As a proportion of total data this makes up over 80%, with trends suggesting that this proportion is growing. This doesn’t just include social listening, but voice, photos, text files, video, and more.
Incredible advancements in AI approaches—such as image recognition, natural language processing (NLP), and speech recognition—have allowed us to increasingly apply structure and analysis to this previously inaccessible data, uncovering substantial opportunities. A combination of speech recognition, text recognition, and NLP have the opportunity to help patient care initiatives extract cross-channel insight that can influence product development, communications, and more. Image and video recognition can find conversations about your brand or relevant conditions that were previously invisible and give you the opportunity to interact and/or gain insight. Choosing the right partner and focusing on a single semi-structured or unstructured data set for a minimum viable product (MVP) is a great way to start unlocking the wealth of information and opportunity in this data.