AI Will Power the Next Gen of 2nd-Party Data Exchanges (and Deliver Personalization, at Scale)
Many brands rely on second-party data exchanges to drive marketing campaigns — a useful strategy in the age of GDPR.
For those unfamiliar with second-party data exchanges, these events are where two parties — say an airline brand and a hotel chain — exchange first-party data in order to reap mutual benefits. In other words, people who fly frequently are likely to stay in hotels regularly. By sharing first-party data, the hotel is likely to find potential new customers in the airline’s customer roster, and vice versa.
Affinity data can help marketers scale their second-party data exchange strategy by looking for non-endemic publishers and partners. While an airline and a hotel brand seem like an obvious fit, there are others, such as a high-end fashion publications or DIY sites, with audiences that may overlap with the hotel audience. Once a marketer decides to do a second-party data exchange with a particular partner, the next question in my mind is: How can that exchange deliver maximum benefit for both parties?
If we are honest, we need to confront the fact that second-party data exchanges, as they exist today, are extremely limited. The insights focus on the percent of overlap in customers, e.g., 60% of our email list matches yours, or 55% of your customers are also in our CRM system. But what does that tell you about the motivations or behaviors of prospects now available to you to target? How can that help you create a campaign with a messaging strategy that breaks through all of the noise?
Where the Real Personalization Happens
In order for second-party data exchanges to succeed in delivering meaningful campaign results and scale, they need to look beyond the percentage overlap between the two audiences, and find the connections the members may have. The truth is, every person has multiple interests, pursuits and passions, and tapping into them can be game-changing for marketers.
Just consider how many attributes can be unearthed in a combined airline and hotel dataset. The hotel will have records of mini-bar purchases, WiFi usage, meal preferences, visits to the business office, or meeting rooms booked. In addition to destination, the airline will know frequency of travel, time of day and day of week for flights, as well as in-flight purchases. These insights can yield nuanced audience segments for both messaging and targeting purposes. Marketers can group customers who fly first thing in the morning and always stay at hotels located close to the business district as “business travelers.” Or they can create a group of customers who go to the hotel gym and purchase from in-hotel juice bars as “health conscious.”
The truth is, without insights about the overlapping customer base, any campaign that stems from a second-party data exchange, as they exist today, need to be inherently generic. Simply knowing that Consumer A stays in a particular hotel and flies a particular airline isn’t very nuanced. Is Consumer A a business or casual traveler? What are his or her tastes and predilections? Without insights at the beginning of the campaign, generic messaging will be the best a marketer can do.
What if marketers could leverage all the combined consumer data from both partners to gain deeper insights in advance of creating the campaign creative? Technologies like AI can identify all of the attributes contained in each company’s CRM system, and see how they’re connected (e.g. visits to the gym and healthy food choices, stays in a particular city, and athletic pursuits). Specifically, graph technology can be used to automatically cluster customers based on common attributes, so that the marketer can, at a glance, identify commonalities that make for interesting and unique marketing campaigns. Clusters, which form unique segments, are the building blocks for highly nuanced personas. If 100,000 guests who stay in a hotel gravitate toward sites about long-distance cycling, the marketer has an opportunity to build a highly effective campaign centered on those common attributes.
Moreover, these clusters would be fully actionable, meaning marketers can create a list of, say, health-minded travelers for targeting as part of the second-party campaign. Rather than build a campaign based on two attributes — this customer flies this airline and stays in this hotel — campaign strategy can leverage the interests and behaviors of the customers, and then target them with highly-personalized messaging.
Second-party data exchanges are predicated on the assumption that each party fully complies with consumer privacy protections, including proper consumer consent and fully anonymized PII data, all of which are necessary in the age of GDPR and California Consumer Privacy Act in the U.S.
Assuming that both parties have the proper rights to share these attributes, imagine the possibilities that open up? More and more, we hear marketers complaining that programmatic and digital ad-tech focuses too much on the execution and doesn’t provide enough insight into who their customers are or how to predict their behavior over the course of their lifetimes. And therein lies the real opportunity of second-party data exchanges; they allow marketers to view customers and prospects holistically and create custom campaigns that resonate.