• by Holly Pavlika , Columnist, June 15, 2017

Research by Colloquy found U.S. consumers were enrolled in 3.5 billion loyalty programs with the average person having as many as 29 memberships. But massive enrollment in programs doesn’t always guarantee loyalty with the advent of digital and social media. Today’s consumers do their research before making a purchase and can be easily swayed to another brand upon discovery of educational content, a product review or upon finding a valuable coupon. 

There are many different methods and proxies to use in measuring ROI. The best metric up until now has been Total Media Value or Earned Media Value. TMV is typically researched across multiple sources, and a value for each social metric is assigned according to industry standards.
With so many companies stepping into the influencer marketing arena, there is fervor to measure its value. We decided to take a look at loyalty card data to see if we could tie it to the efficacy of influencer campaigns.

We started by pairing our audience data with frequent shopper and loyalty card data for a major confection brand that was aimed at multiple retailers during the Halloween timeframe last year. 

To measure the Return on Ad Spend (ROAS) of the program, we paired retailer frequent shopper card (loyalty card) data with our first-party audience pixel data. Note: all household (HH) Level Design of Experiments (DOEs) require pixel-based/cookie tracking to marry shoppers with their Frequent Shopper Card (FSC) or mobile devices, except for coupon studies. Promotional tie-ins, on the other hand, require an observational study versus the average benchmark, instead of Test and Control. Coupon studies on most platforms require users to have pre-existing accounts. 

Our research partners paired exposed individuals with look-a-like controls for HHs based on a list of covariates from demographics to lifestyle and actual historical purchases. Once Test HHs and controls receive vectors of their “Shopper DNA,” both groups are then paired up using distance routines such as genetic score matching or k-Nearest Neighbors (k-NN). Finally, we measured the average lift on the exposed readers. 

Not only was the study able to optimize the target audience, but it also provided data-backed sales lift data to the confections advertiser. This sales lift model provides concrete revenue data related to influencer content that exceeds any Total Media Value increments, creating a brand new perspective for the influencer marketing industry pairing the loyalty card data with Collective Bias pixel data created the ultimate measurement opportunity for the advertiser. By serving targeted influencer content to the audience, the advertiser saw an increase in incremental spending per household. 

Audiences exposed to influencer content had an incremental spending per household of $0.75 more than audiences not exposed to influencer content, an increase of 5.4%. 

This sales lift model provides revenue data related to influencer content that exceeds any Total Media Value (TMV) in traditional influencer marketing metrics, such as engagement, impressions or time spent, again creating a brand new perspective for the influencer marketing industry. This gives marketers another tool in the arsenal to prove influencer relationships and their content are a valuable way to not just reach audiences but underscore the value it plays on the path to purchase. Or should we say, the path to influence?