Express has also turned its transaction data into a revenue
stream through its Business Insights consulting division which
has aimed direct mail and online offers to card holders on behalf
of advertisers for years, though on an aggregate level. More
recently, AmEx has modeled audience segments for use in online ad
targeting. The company declined to name any partners in the
endeavor, but stressed the AmEx data models don't allow for direct
targeting of its card holders.
Mastercard recently aligned with Maxpoint, a digital-ad firm
that combines lots of publicly available information, such as data
from Secretaries of State and health data from the Centers for
Disease Control, to define audiences within specific ZIP code
regions. The Mastercard information is "the first credit-card data
set that we've incorporated into our system," said Maxpoint CEO Joe
Maxpoint sells display, mobile and video ads featuring targeted
coupons or store promotions on behalf of CPG brands and other types
of advertisers such as restaurants. The Mastercard data shows how
high or low people index within a specific ZIP code for certain
types of purchases. Using the Mastercard data, a burger or pizza
chain might use the system to push promotions to neighborhoods in
which people spend more than the average at fast-food joints. The
firms have worked together for around six months to get the system
Mastercard also makes its data available through Exelate, a
data-management platform that feeds information from Mastercard and
other partners including Acxiom and Nielsen into digital-ad
exchanges for targeting. The credit-card firm compiles audience
segments based on its cardholders' transactions, pegging people as
likely to shop at a sporting goods store or specialty retailer, for
example. Exelate uses another company to transfer the Mastercard
segments online to target large, modeled audiences geographically
according to ZIP+4 codes, rather than targeting individuals based
on their personal transactions.
Mastercard stresses the transaction data is anonymized and
provided in aggregate. "It's really more of a broad database," said
Susan Grossman, group head of media solutions for MasterCard
Advisors Information Services.
Advertisers can target ads to large pools of Mastercard
customers categorized into segments such as "Top Tier Auto
Spenders," or "Frequent Transactors," said a source at a
media-agency trading desk using Mastercard data who asked to remain
anonymous. The information is "not very granular," said the source,
who suggested privacy concerns have prevented Mastercard from
offering more detailed data on an individual basis.
Still, the source indicated he thought he was targeting actual
Mastercard cardholders through online matching and cookie dropping,
rather than merely targeting geographic-level segments. While he
said the Mastercard data performs relatively better than other
information available for digital targeting, the source called it
Privacy concerns certainly play a role in how firms like
Mastercard make their data available, but those qualms aren't
preventing purchase data from seeping online more frequently.
Acxiom has partnered with Facebook
to test ad targeting on the social site using customer loyalty
shopping data. The company also recently unveiled plans to connect
the data it stores for its clients including purchase history and
loyalty card information to publisher sites.
Venture Development Center serves as an intermediary between
data brokers and brands and data providers. The company has
discussed data monetization with most if not all credit card firms,
said VDC President and COO Matthew Staudt, noting they have "a high
level of interest, but also a very strong sense of caution."
The information supplied by credit card firms and other data
companies like Acxiom is arguably more reliable because it reflects
real consumer purchases as opposed to inferences about what people
appear to be interested in. The majority of behavioral data used to
target digital ads has its share of problems, suggested the trading
desk source who said people might be added to audience segments
(think in-market shoe buyers) based only on a one-time site visit.
In addition, some computers in homes are used by multiple people,
muddying behavioral profiles.