••• customer satisfaction research Transcending brand and loyalty Rethinking the customer experience

| By Danica R. Allen

snapshot Over the past two decades there have been few paradigmatic changes in the world of customer satisfaction research. We seek to remedy this situation through the introduction of a new, broader construct that subsumes many elements of both customer satisfaction In part one of a two- and brand measurement research. Historically, customer satisfaction research has focused part article, Danica on measuring service and product quality and relating these to attitudinal outcomes such as overall satisfaction and loyalty or more recently, economic outcomes such as market Allen presents a share and profitability. Similarly, brand research has generally been oriented toward new approach to Ocapturing consumer perceptions of brand imagery or performance and relating these to psychometric or econometric measures of brand equity and, ultimately, purchase behavior. viewing the consumer Efforts to innovate around these two constructs have generally focused on new pscyho- experience and metric outcome measures. In particular, the last major advancement in the customer satisfaction arena involved the introduction of loyalty. This was initially formulated as a its relationship to three-item index but later was afforded greater complexity as a consensus grew around the customer satisfaction. notion that loyalty subsumed both rational and emotional components. Underlying the shift toward loyalty measurement was the intuitively palatable notion that even satisfied customers depart. Similarly, in the brand research arena much effort has been expended on refining or enhancing dependent measures under the guise of brand equity, brand at- tachment or brand vitality. A robust economic measure of brand equity remains elusive. We turned to consumer experience as a construct to bridge the chasm between brand and loyalty research. It was preceded, however, by an industry-wide obsession with differentiating loyalty from customer satisfaction. Reichheld and Sasser (1990) and later Reichheld (1996) produced compelling work that essentially set the stage for a paradigmatic shift toward consumer experience. Interest in this field became especially keen in the late 1990s with the publication of Pine and Gilmore’s 1998 article “Welcome to the Experience Economy.” Pine and Gilmore (1998) argued that from a macroeconomic perspective, the economic offering of firms has progressed from commodities to goods and from goods to services. They argued that from this progression has emerged the experience economy wherein the economic function is not to extract fungibles, make tangibles or deliver service but instead to stage experiences. The nature of the offering is memorable. The seller is the stager and the buyer, the guest. The demand fac- tor in the experience economy is the memorable sensation. Despite the compelling nature of this framework and more recent epiphanies such as that reflected in the Adweek excerpt below, there were no substantive efforts to produce a commercially viable, comprehensive system for measuring, tracking and ultimately manipulating the charac- teristics of a consumer experience. The modal treatment of consumer experience during the last 10 years focused on its sensorial aspects (see for example Lindstrom [2005] and Michelli [2007]). “Whatever the methodology, it’s increasingly clear that customers desper- ioral, cognitive and emotional. sion ads, billboards, magazines, Web ately want goods and services, commu- Each dimension clearly measures sites and so on. In contrast, direct XPs nications and marketing campaigns that a different nuance and each is also involve substantive involvement with dazzle their senses, touch their hearts presumed to be important depending a product or service. These XPs tend and stimulate their minds – delivering a on the idiosyncratic appeal of a given to be the domain of loyalty research. positive experience they will remember. experience. For example, brands such The distinction between vicarious Businesses will live or die not by the as Starbucks would certainly find the and direct XPs can be leveraged to attributes they promise but by the brand sensorial dimension critical to the change the focus of the measurement experiences and value they offer custom- experience it offers. The extent to system described in this article. It is ers at every touchpoint.” which an experience should emphasize important to note that distinguishing Adweek, September 29, 2008 behavioral characteristics may also between them is unnecessary and is differ. Harley-Davidson’s core experi- done so here to illustrate the ability of Mind modularity ential value proposition is clearly more this research to transcend both brand A unique aspect of GfK’s ConEx system oriented to the behavioral, sensorial and loyalty research. is its integration of a concept devel- and emotional dimensions. Our research was driven by the no- oped by evolutionary psychologists: A more in-depth treatment of the tion that any experience can be charac- mind modularity. Although contro- Brakus et al research is beyond the terized in terms of the five dimensions versial in the early 1920s when sug- scope of this article. Suffice it to say introduced by Pinker and later adopted gested by Dewey (1922, 1925) and even that in addition to Pinker’s five- by Brakus et al. In order to make the di- more so in 1997 when championed by dimension architecture, these authors’ mensions more consumer-friendly, we Harvard University’s Steven Pinker programmatic research represents the translated the original labels into talk, in How the Mind Works, the notion theoretical and empirical foundation think, sense, feel and act and paired that the human mind has evolved in upon which we have built a highly them with graphic icons as represented a Darwinian fashion, yielding five flexible, modular system for measur- in Figure 1. distinct modules to facilitate humans’ ing and tracking consumer experience. interactions with their environment, Large sample sizes has gained wide acceptance. Transcend the boundaries This study spanned five business sec- Our interest in the application This study is based on a series of R&D tors. Within each sector we measured of mind modularity and brands was projects spanning two years. The five brands and within each brand, piqued by treatments of the subject objective was to develop a system that eight XPs. This structure demands spanning about 10 years: Brakus (2001); could transcend the somewhat artifi- large sample sizes. Indeed, in total Schmitt (2008, 2003); and Brakus, cial boundaries that delineate the do- the design subsumes 200 unique Schmitt and Zarantonello (2009). The mains of brand and loyalty research. sector/brand/XP cells. In order to five experiential dimensions that The common currency involves the achieve n = 200 within each cell, a represent the foundation upon which five experiential dimensions described total of 40,000 observations would be these authors developed a brand-orient- above. If the five dimensions are the necessary. Time and economic limita- ed architecture were based largely upon common currency for transcending tions precluded such a large-scale Pinker’s (1997) work. Their program- brand and loyalty, then experience effort and instead, we opted to select matic efforts, which spanned essen- touchpoints (XPs) represent a com- the top five brands within each sector tially the first decade of this century, mon language. In this research, we based on market share. continually confirmed Pinker’s five differentiate between vicarious and Experience touchpoints play an dimensions: sensorial, social, behav- direct XPs. The former include televi- important part in this architecture; Table 1: Five Sectors, Each with Eight XPs respondents were shown an illustra- tion of how a person might place the Sector Direct XPs Vicarious XPs five experience icons and an overall Banking Visit the Web site TV advertising touchpoint icon with respect to wait- Visit branch office Print advertising ing in line to place an order in an Visit ATM Internet advertising upscale coffee shop. In this exercise, Interact with loan advisor Friends’ recommendation respondents were shown how to drag Automotive Visit showroom/dealer TV advertising each icon into the quadrant space Test drive Print advertising and place it based on both impression Driven before Internet advertising and memorability. After all six icons Seen on street Friends’ recommendation were placed in this training exercise, a respondent’s efforts would yield Laptop Tried at store TV advertising the final illustrative result presented Used at friend’s house Print advertising in Figure 3. By placing the six icons Use at job Internet advertising into the quadrant space a respondent Interacted with salesperson Friends’ recommendation creates 12 scores since each icon has a QSR Visit the Web site TV advertising unique x/y position. Restaurant Eat in the restaurant Print advertising As described earlier, this research Placed an order at counter Internet advertising project subsumed five sectors. Within Used rest room Friends’ recommendation each sector five brands and eight Shampoo Use it at home TV advertising XPs were of interest. Respondents Handling the package Print advertising were routed into two sectors depend- Receive free sample Internet advertising ing upon screening and within each, Purchase at store Friends’ recommendation asked to rate two XPs in terms of the five experiential dimensions de- scribed earlier. Thus, any respondent they represent a common language for rant similar to that depicted in Figure could complete at most four quadrant both loyalty and brand research. As 2. Clearly, in order to complete this exercises in which the five experi- noted earlier, we selected eight XPs for task, respondents needed substantive ence icons and a single dependent each sector. Vicarious XPs are the same briefing with respect to the five expe- touchpoint metric were dragged and for each sector. The four direct XPs dif- rience icons and the two axes. The five dropped into the impression–memo- fered based on the idiosyncratic nature icons were illustrated with a series of rability quadrant. of each sector. The XPs for each sector photographs and descriptions. The two In total, we collected 3,000 com- are presented in Table 1. The system quadrant axes were described using a pleted interviews using GfK’s online described in this research is highly series of examples that illustrated how consumer panel. On average, respon- flexible; the XP set may be focused on an experience could be highly memo- dents completed 2.7 quadrant exercises more direct XPs and therefore be of rable due to its positive or negative each. This yielded a total of 8,145 com- greater utility to managers interested nature. They were also shown how, re- pleted quadrants. The resultant dataset in enhancing the consumer experience. gardless of quality, experiences could provided rich depth and breadth for In contrast, if more emphasis is placed be highly forgettable. analysis both within and across sectors. on vicarious XPs, the results will likely Following the introductions to The two 85-point quadrant axes have more utility to brand managers the icons and quadrant dimensions, yielded distributions with more and advertising researchers. Data relating to each of the five experiential dimensions were col- lected using an innovative technique. We were hesitant to utilize Likert scales for each of the five experien- tial dimensions since a memorability and impression (valence) rating was needed. Lengthy batteries of Likert scale items become tedious very quickly and our fear was that straightlining and disengagement would ameliorate the relationships we sought to reveal. In order to capture impression and memorability ratings for each of the five experiential dimensions (Talk, Think, Feel, Sense, Act) we embraced a novel approach. This involved a quad- axis were strong/memorable and weak/forgettable. Again, a row-wise approach to examining the table will help us understand cross-sector differences. With respect to the Talk icon distributions we see that only the automotive sector data were as- sociated with a significant number of positive (i.e., positioned high on the y-axis) icon placements. The second row of Figure 5 shows the Think icon distributions. As shown, the automotive and computer sectors had the most memorable icon placements and as expected, the shampoo category Think icon was typically placed at the bottom of the y-axis, indicating its rather desirable characteristics than similarly aimed at familiarizing the reader with forgettable nature. The placement administered seven-point Likert scales1. the common currency and language of of the Sense icon illustrates several Figures 4 and 5 present the x-axis and consumer experience research. intuitively appealing patterns. For y-axis distributions for the five sectors The data presented in Figure 4 example, both the food and shampoo aggregated over all brands and touch- exhibit some interesting patterns sectors enjoyed many icon placements points. Clearly, the data presented in despite the aggregation by brand and at the top of the quadrant, indicating the figures cannot reveal the nuanced XP. If we examine the data in a row- strong memorability. In contrast, the differences that more granular brand wise fashion and begin with the top bank sector was associated with the or XP-level treatment would have al- tier (Talk) there are some noteworthy most forgettable sensorial ratings. lowed. Unfortunately, space constraints differences. The automotive data, for The Feel icon, which represents precluded treatment of the brands example, yielded many more positive the emotional impact of an experi- within each sector and this will be icon placements. To a certain ex- ence, elicited the most positive place- described in part two of this article, tent, the computer data distribution ments within the automotive sector which will appear in next month’s appears to resemble the automotive and fewest in the bank category. Quirk’s Marketing Research Review. data. In the next row are the x-axis Interestingly, there was a significant distributions of the Think icon. As group of respondents who placed Very appealing distributions shown, the automotive and banking the Feel icon in the highest vertical The quadrant data collection mecha- sectors resulted in the most positive position for the computer segment. nism yielded very appealing distri- icon placements and the shampoo Finally, the Act icon y-axis distribu- butions. Clearly, the x-axis distribu- category the fewest. The shampoo tions are presented in the fifth row tions differed substantively from category emerged as essentially the of the figure. These clearly show that the y-axis distributions. Recall that inverse of these with a high propor- the behavioral dimension was very the former reflected impression and tion of respondents placing the Think memorable in the automotive sector; the y-axis represented memorabil- icon on the far left of the quadrant. this left skew is also reflected in ity (ranging from weak/forgettable The third row of Figure 4 pres- the computer sector and is notewor- to strong/memorable). One aspect ents the x-axis placements of the thy. The remaining sectors yielded of the quadrant data collection Sense icon. In this case we see the bi-modal distributions since many mechanism that remains untested automotive, food and shampoo respondents placed the icon very high is the possibility that respondents category distributions yielding the or very low in the quadrant space. utilize vertical and horizontal spaces most positive placements. The Feel The distributions in Figures 4 and differentially. Thus, we concede at icon distributions are presented in 5 demonstrate interesting differences this point that had the axes been the fourth row of Figure 4 and sug- across the sectors and across the five reversed, it is possible the distribu- gest a stronger emotional component experiential dimensions. For exam- tions would reflect this. associated with the automotive and ple, a column-wise review of Figure Figure 4 presents the x-axis (im- especially computer sectors. Finally, 4 reveals substantive stability with pression) distributions for the five the Act icon distributions suggest respect to the x-axis placement of the sectors. Within each distribution, the that the automotive and computer five icons. However, a parallel exami- data are aggregated across five brands categories left the most positive (va- nation of Figure 5 suggests consider- and eight XPs. A comprehensive lence) impressions. ably more differentiation across the treatment of the data would neces- Figure 5 presents the y-axis five experiential dimensions (Talk, sitate 40 distributions for each of the (Memorability) experience distribu- Think, Sense, Feel, Act). That is, with five sectors. Part one of this article is tions. Recall that the anchors of this respect to vertical axis placement Danica R. Allen is global director customer satisfaction and experience at GfK Custom Research North America, New York. She can be reached at 212-240-5503 or at [email protected].

Footnote 1 Note that a parallel study was undertaken wherein the two axes were administered as seven-point Likert scales and that these data were determined to be inferior to the quadrants with respect to factor structure, dependence models and distributional criteria including skewness and kurtosis. A comprehensive psychometric comparison of the two approaches will be the subject of a forthcoming paper.

References Brakus, J., Schmitt, B., and Zarantonello, L. (2009). “Brand Experience: What Is It? How Is It measured? Does It Affect Loyalty?” Journal of Marketing, 73:52-68. Brakus, J. (2001). A Theory of Consumer Experience. Doctoral dissertation, Columbia University. Dewey, J. (1925). Experience and Nature. New York: Dewey. Dewey, J. (1922). Human Nature and Conduct. New York: The Modern Library. Lindstrom, M. (2005). Brand Sense: How To Build Powerful Brands Through Touch, Taste, Smell, Sight and Sound. New York: The Free Press. Michelli, J. (2007). The Starbucks Experience. New York: McGraw-Hill. Pine, B. and Gilmore, J. (1998). “Welcome to the Experience Economy.” Harvard Business Review. July-August: pp. 97-105. Pinker, S. (1997). How the Mind Works. New York: Norton. Reichheld, F. (1996). The Loyalty Effect. Boston: Harvard Business School Press. Reichheld, F. and Sasser, W. (1990). “Zero Defections: Quality Comes To Services.” Harvard Business Review 68:105-111. Schmitt, B. (2008). “A Framework For within each sector there were strik- relate the five experience icons to Managing Customer Experiences.” In Handbook ingly different distributions for each the single XP outcome measure icon. on Brand and Experience Management, Bernd H. Schmitt and David L. Rogers, Eds. icon, suggesting that the memorabil- In next month’s issue we will pro- Northampton, Massachusetts: Edward Elgar ity of touchpoint exposure may be a vide a deep dive into the unique as- Publishing. critical dimension that until now, has pects of the quadrant data and some Schmitt, B. (2003). Customer Experience not received much attention. approaches to driver analysis within Management. New Jersey: John Wiley and Sons. Our analytic efforts have un- a two-dimensional space. Our focus Schmitt, B. (1999). Experiential Marketing: equivocally confirmed that tra- will be on one sector and illustrate How to Get Customers to Sense, Feel, Act, and ditional linear techniques yield the richness of these data by examin- Relate to Your Company and Brands. New York: anemic models when we attempt to ing both brand and XP dynamics. The Free Press.

Reprinted with permission from Quirk’s Marketing Research Review, January 2012. On the Web at www.quirks.com. © Quirk Enterprises, Inc. All Rights Reserved. Foster Printing Service: 866-879-9144, www.marketingreprints.com. Gfk Custom Research 75 Ninth Avenue, 5th floor New York, NY 10011 Phone: 212-240-5300 [email protected] • www.gfkamerica.com