Most financial institutions and issuers of debit and credit cards, as well as merchant branded private label payment schemes, miss three key areas of due diligence in their return on investment (ROI) analysis of card-linked offers (CLO) programs: Incremental merchant results (Performance), Consumer offer targeting (Relevancy) and Merchant offers (Content). 

Though all three components are required to facilitate a CLO ecosystem, the first two are the most important. It is in these two core area, performance and relevancy, that a CLO business needs to create competitive advantage. And herein lies the flaw of most CLO deployments:  they have focused resources on securing vast quantities of content (merchant offers) at the expense of building sustainable, core technology with a high degree of analytical rigor that provides increased Performance and Relevancy. 

In our observation, many of the early entrants have not adequately invested in the data science required to achieve Performance and Relevancy. As a result, they suffer from two critical challenges that limit their long-term success.

The first challenge is poor customer targeting, or the inability to predict relevancy and excessively long redemption windows. The critical success factor in CLO is the sophistication of the predictive analytics used to render the right offer to the right person at the right time. The early entrants have yet to graduate from a simple combination of historical spend data and look-alike modeling to target their offers. As a result, their ability to predict future consumer behavior is limited. This keeps them from being able to deliver truly personalized offers and thus limits uptake by consumers and the ultimate value to merchants and advertisers. If a CLO can’t predict future spend behavior, when an offer will be relevant to a specific individual or the size of the offer (discount percentage) that will drive incremental behavior, a CLO program is forced to utilize an “Offer Mall” as their product user interface.

This product experience creates a number of challenges. For many consumers, the “Offer Mall” creates stress and frustration as they scroll through a long list of random offers. Subsequently, this approach mostly attracts a subset of extreme “coupon clippers” who scour the Internet for deals and discounts. Since the presented offers are based largely off of historical spend data and are typically active for months at a time, the majority of these redemptions are predisposed or “accidental” from existing baseline customers, and would have taken place regardless of the Card Linked Offer.  This group would be better served through a loyalty reward program, not a program designed to bring in first-time customers.

The second challenge is poor performance and lack of incremental sales lift, audit trail and valid reporting. The predisposed/accidental problem, which stems from poor targeting, creates the most frequent merchant objection to the entire card-linked category:  lack of incremental revenue or paying extra for customers who were already planning a purchase. The scale of predisposed redemptions further exacerbates this problem. In many cases, merchants who utilize CLOs with inferior targeting often receive negative ROIs from CLO programs—they lose money since they are providing an unnecessary financial incentive/discount to those already planning to patronize a particular merchant.

Sponsoring issuers, financial institutions, and merchants themselves, need to take a closer look at the source and formulas of any statistics used to report the results of a CLO program. In many cases, the group receiving the offer is compared to another group to show effectiveness of the program, yet the two groups have no commonality that would make the results statistically meaningful. An example of this is the utilization of a “Hold Back Group.” This flawed methodology attempts to show a “before and after” impact analysis. In the Hold Back Group Test, a portion of the population is removed from receiving offers. This group is then compared to the group that receives offers and the difference in performance of the two groups is credited to the CLO program.

The problem here lies in the legitimacy of the Hold Back Group. For example, if the CLO program was targeted at selling diapers to cardholders with infants, and you hold back a random group of individuals, most of those individuals in the Hold Back Group won’t actually be in the market to buy diapers. The “incrementality” shown through this test is inherently skewed in favor of the group receiving the incentive since they were selected to receive the offer precisely because of their historical purchase of diapers.

In contrast, the scientific community relies on the more rigorous Test vs. Control method. This Scientific Control Test requires matching statistically identical groups with respect to the offer, category etc. To measure the true effectiveness of the offer, the Control group (not receiving the offer) must share all relevant characteristics and profile as the Test group receiving the offer, in order to achieve a statistically valid comparison. As marketing budgets face increasing scrutiny and transparency, advertisers are demanding measurable, verifiable, and quantifiable results from Card Linked Offer programs. CLOs not supported by a sophisticated data matching engine will be unable to drive true incremental performance, and will struggle to maintain a strong merchant pool without real, auditable results of increased traffic and consummated sales.

As the CLO market continues to mature, the quantity, creativity and quality of merchant content and offers will steadily increase. In this advancing ecosystem, marketers will steadily shift their advertising dollars to CLO programs that deliver the highest impact, to the most targeted audience with the greatest ROI. Issuers, financial institutions and other funded account processors (e.g., private label closed looped payment processors, Internet and mobile wallets schemes, etc.) will need to judiciously render highly targeted content, amping the relevancy of advertisements and offers to the customer segments most likely to increase net new sales for merchants, and avoid discounting to loyal customers that have already solidified their commitment to a retailer. This will require continuous improvement in the platforms that drive the predictive analytics to move beyond the inefficient and annoying “spray and pray” legacy promotion models of the past.

The number of input sources and volume of big data will challenge many of the early entrants to keep up with the massive scale and analytical science required to digest and render relevant ads and offers. This will grow in importance as card issuers, financial institutions and other payment account providers move to mobile wallets as their primary user interface (UI) for reviewing, activating and redeeming offers. Mobile will challenge most CLO platforms to move beyond the static, after-the-sale batch data feeds of transaction and statement processing systems principally used today. The critical success factor for CLO programs will be the ability of their predictive analytics engine to digest numerous data feeds in a continuous and real-time fashion to present relevant ads and offers not only for retailers, but also for the consumer packaged goods (CPGs), product manufacturers and brands across the continuum of the mobile sales cycle before, during and after every payment, not just at the static statement level. 

The big win will not be from linking offers to cards, but to mobile wallets. This means the CLO platform must provide an open application programming interface (API) that not only issuers and financial institutions can use inside their wired Internet sites, but can also be integrated easily with mobile banking and retailer apps. Mobile-enabled CLO platforms will also be positioned to power the mobile wallets of retailers, retailer-lead payment consortiums such as the Merchant Customer Exchange (MCX) and other new entrants such as Google Wallet, ISIS, PayPal, Amazon, Apple and the like.

CPGs and product brands account for the majority of today’s advertising spend, up to 95% for some categories. CLO programs with the power to ingest and analyze multiple big data sources real-time and render relevant CPG and manufacturer sponsored ads and offers dynamically, will enable an enormous new source of revenue as the major product brands seek to reach the mobile consumer at the right time, in the right place, with the right message.

These mobile functional requirements will create a chasm in the market between those tech-centered CLO programs that can provide the sophisticated data analysis engine to enable mobile relevancy at massive scale in real-time versus old-fashioned, simple offer sourcing, expecting viewership through wired desktop connections to legacy, after-the-fact transaction listings in online account management systems.

Richard Crone and Heidi Liebenguth lead Crone Consulting LLC, an independent advisory firm specializing in mobile strategy and payments.