Can combining web and mobile communication channels reveal concealed customer value?

Many firms have implemented a customer-value based segmentation to improve the efficiency of MARCOM campaigns as part of their long term customer relationship strategies (Kumar 2010, Thomas et al. 2005). If distribution channel addition may increase the intrinsic customer value (Kumar 2005, Rangaswamy 2005) in a US context as well as in a French context (Vanheems 2009), few studies have been conducted about the impact of adding a new communication channel during the same buying journey. As the number of devices enabling to communicate with consumers increases, the goal of this research is to understand whether combining web and mobile communication channels for a same journey is more efficient than using a single web channel, monitoring the promotional pressure. Due to consumer reactance at higher levels of communication, Godfrey et al. (2011) have shown that channel interaction effects could lead to diminishing returns. The issue of this paper is to evaluate if the same result can be observed in case of omni-channel strategy involving a mobile channel (Verhoef et al. 2015). More precisely the major contribution of our paper is to extend current knowledge by measuring the impact of combining a web channel (email) with a mobile one (SMS). The measure will be made on a large population of 304,410 individuals from a French click & mortar retailer's customer database and with an unprecedented richness of data thanks to individual variables regarding customer purchases, browsing behavior, customer historical reaction to brand communications and customer distance to closest store. Different questions can be raised: (1) Can combining web and mobile channels within a communication strategy result in an increase in customer count and spend both offline and online? (2) How can we explain this impact in terms of segment specificity such as: i) FRAT segmentation, ii) Browsing data: mobile versus desktop readers, iii) Customer historical reaction to brand emails, iv) Customer distance to closest store? The aim of this research is then: (1) to evaluate, at equal effective pressure, the impact of adding a mobile communication channel on customer buying behavior. (2) to find out whether given customer segments reveal higher value thanks to multichannel targeting. (3) to explore and advance factors explaining the differential impact of multichannel versus single channel communication on purchase behavior. To comply with our focus on customer value unlocking, we measure increments provided by multichannel targeting compared to single channel targeting. As recommended by Dinner, van Heerde, Neslin (2014), the four following outcomes of interest have been retained: i) Customer count offline, ii) Customer spend offline, iii) Customer count online and iv) Customer spend online A field experiment has been carried out to measure and understand the differential impact of push multichannel communication sequences. Five populations, split randomly, have been exposed to a specific channel strategy. The first one was assimilated to a control group as exposed to single channel targeting by email only. Four multichannel test groups have been implemented; three of them were targeted with three messages but with different sequences mixing email and SMS. The fourth group was intentionally slightly over-pressured with four messages. The analysis methodology consists in calculating, for each population these four dependent variables and measure a "multichannel lift". This lift analysis will be done with the customer-level data exposed above. Very first results advance significant insights. No impact has been observed on brands' heaviest customers, which is counterintuitive in a FRAT-based targeting framework. Combining mobile and web channels results in a 40% lift in customer count on non-reactive to email customers, highlighting that inactive customers on one channel can still be active on a second channel. Finally, desktop email readers are more likely to purchase when targeted via mobile than mobile readers are. This research enables a new understanding for academia in the field of consumer response to multimedia communications as well as guidelines for practitioners about differentiated segments sensitivity, to refine their multichannel targeting strategies. Due to space limitations, underlying explanations are not presented here but theoretical propositions will be discussed in a full paper.