Feeds:
Posts
Comments

Posts Tagged ‘Consumer behaviour’

One of the more difficult and troublesome decisions in brand management arises when entering a product category that is new to the company: Whether to up-start a new brand for the product or to endow it with the identity of an existing brand — that is, extending a company’s established brand from an original product category to a product category of a different type. The first question that would probably pop-up is “how different is the new product?”, acting as a prime criterion to judge whether the parent-brand fits the new product.

Notwithstanding, the choice is not completely ‘black or white’ since intermediate solutions are possible through the intricate hierarchy of brand (naming) architecture. But focusing on the two more distinct strategic branding options above helps to see more clearly the different risk and cost implications of launching a new product brand versus using the name of an existing brand from an original product category. Notably, the manufacturers, retailers and consumers, all perceive risks, albeit from the different perspective of each party given its role.

  • Note: Brand extensions represent the transfer of a brand from one type of product to a different type, to be distinguished from line extensions that pertain to the introduction of variants within the same product category (e.g., flavours, colours).

This is a puzzling marketing and branding problem also from an academic perspective. Multiple studies have attempted in different ways to identify the factors that best explain or account for successful brand extensions. While the stream of research on this topic helpfully points out to major factors, some more commonly agreed upon, a gap remains between the sorts of extensions predicted to succeed according to the studies and the extensions performed by companies that happen to succeed or fail in the markets in reality. A plausible reason for missing the outcomes of actual extensions, as argued by the researchers Milberg, Sinn, and Goodstein (2010), is neglecting the competitive settings in categories that are the target of brand extension (1).

Perhaps one of the most famous examples of a presumptuous brand extension has been the case of Virgin (UK), from music to cola (drink), airline, train transport, and mobile communication (ironically, the origin of the brand as Virgin Music has since been abolished). The success of Virgin’s distant extensions is commonly attributed to the personal character of Richard Branson, the entrepreneur behind the brand: his boldness, initiative, willingness to take risks, and adventurism. These traits seem to have transferred to his business activities and helped to make the extensions more credible and acceptable to consumers.

Another good example relates to Philips (originated in The Netherlands). Starting from lighting (bulbs, now more in LED), the brand extended over the years to personal care (e.g., face shavers for men, hair removal for women), sound and vision (e.g., televisions, DVD and Blue-Ray players, originally in radio sets), PC products, tablets and phones, and more. Still, when looking overall at the different products, systems and devices sharing the Philips brand, they can mostly be linked as members in a broad category of ‘electrics and electronics’, a primary competence of the company. As the company grew with time, launched more types of products whilst advancing with technology, and its Philips brand was perceived as having greater experience and good record in brand extensions, this could facilitate the market acceptance of further extensions to additional products.

  • In the early days of the 1930s to 1950s radio and TV sets relied for operation on vacuum tubes, later moving to electronic circuits with transistors or digital components. Hence, historically there was an apparent physical-technological connection between those products and the brand’s origin in light bulbs, a connection much harder to find now between category extensions, except for the broad category linkage suggested above.

Academic research has examined a range of ‘success factors’ of brand extensions, such as: perceived quality of the parent-brand; fit between the parent-brand and the extension category; degree of difficulty in making an extension (challenge undertaken); parent-brand conviction; parent-brand experience; marketing support; retailer acceptance; perceived risk (for consumers) in adopting the brand extension; consumer innovativeness; consumer knowledge of the parent-brand and category extension; the stage of entry into another category (i.e., as an early or a late entrant). The degree of fit of the parent-brand (and original product) with the extension category is revealed as the most prominent factor contributing to better acceptance and evaluation (e.g., favourability) of the extension in consumer studies.

Aaker and Keller specified in a pioneer article (1990) two requirements for fit: (a) the extension product category is a direct complement or a substitute of the original category; (b) the company, with its people and facilities, is perceived as having the knowledge and capability of manufacturing the product in the extension category. These requirements reflect a similarity between the original and extension product categories that is necessary for successful transfer of a favourable attitude towards the brand to the extension product type (2). A successful transfer of attitude may occur, however, also if the parent-brand has values, purpose or image that seem relevant to the extension product category, even when the technological linkage is less tight or apparent (as the case of Virgin suggests).

  • Aaker and Keller found that fit, based especially on competence, stands out as a contributing factor to higher consumer evaluation (level of difficulty is a secondary factor while perceived quality plays more of a ‘mediating’ role).

Volckner and Sattler (2006) worked to sort out the contributions of ten factors, as retrieved from academic literature, to the success of brand extensions; relations were refined with the aid of expert advice from brand managers and researchers (3). Contribution was assessed in their model in terms of (statistical) significance and relative importance. The researchers found  fit to be the most important factor driving (perceived) brand extension success in their study, followed by marketing support, parent-brand conviction, retail acceptance, and parent-brand experience. The complete model tested for more complex structural relationships represented through mediating and moderating (interacting) factors (e.g., the effect of marketing support on extension success ‘passes’ through fit and retailer acceptance).

For brand extensions to be accepted by consumers and garner a positive attitude, consumers should recognise a connectedness or linkage between the parent-brand and the category extension. The fit between them can be based on attributes of the original and extension types of product or a symbolic association. Keller and Lehmann (2006) conclude in this respect that “consumers need to see the proposed extension as making sense” (emphasis added). They identify product development, applied via brand (and line) extensions, as a primary driver of brand growth, and thereby adding to parent-brand equity. Parent-brands do not tend to be damaged by unsuccessful brand extensions, yet the authors point to circumstances where greater fit may result in a negative effect on the parent-brand, and inversely where joining a new brand name with the parent-brand (as its endorser) may protect the parent-brand from adverse outcomes of extension failure (4).

When assessing the chances of success of a brand extension, it is nevertheless important to consider what brands are already present in the extension category that a company is about to enter. Milberg, Sinn, and Goodstein claim that this factor has not received enough attention in research on brand extensions. In particular, one has to take into account the strength of the parent-brand relative to competing brands incumbent in the target category. As a starting point for entering the extension category, they chose to focus on how well consumers are familiar with the competitor brands vis-à-vis the extending brand.  Milberg and her colleagues proposed that a brand extension can succeed despite a worse fit with the category extension due to an advantage in brand familiarity, and vice versa. Consumer response to brand extensions was tested on two aspects: evaluation (attitude) and perceived risk (5).

First, it should be noted, the researchers confirm the positive effect of better fit on consumer evaluation of the brand extension when no competitors are considered. The better fitting extension is also perceived as significantly less risky than a worse fitting extension. However, Milberg et al. obtain supportive evidence that in a competitive setting, facing less familiar brands can improve the fortune of a worse fitting extension, compared with being introduced in a noncompetitive setting: When the incumbent brands are less familiar relative to the parent-brand, the evaluation of the brand extension is significantly higher (more favourable) and purchasing its product is perceived less risky than if no competition is referred to.

  • A reverse outcome is found in the case of better fit where the competitor brands are more highly familiar: A disadvantage in brand familiarity can dampen the brand extension evaluation and increase the sense of risk in purchasing from the extended brand, compared with a noncompetitive setting.

Two studies performed show how considering differences in brand familiarity can change the picture about the effect of brand extension fit from that often found without accounting for competing brands in the extension category.

When comparing different competitive settings, the research findings provide a more constrained support, but in the direction expected by Milberg and colleagues. The conditions tested entailed a trade-off between (a) a worse fitting brand extension competing with less familiar brands; and (b) a better fitting brand extension competing with more familiar brands. In regard to competitive settings:

The first study showed that the evaluation of a worse fitting extension competing with relatively unfamiliar brands is significantly more favourable than a better fitting extension facing more familiar brands. Furthermore, the product of a worse fitting brand extension is preferred more frequently over its competition than the better fitting extension product is (chosen by 72% vs. 6%, respectively). Also, purchasing a product from the worse fitting brand extension is perceived significantly less risky compared with the better fitting brand. These results indicate that the relative familiarity of the incumbent brands that an extension faces would be more detrimental to its odds of success than how well its fit is.

The second study aimed to generalise the findings to different parent-brands and product extensions. It challenged the brand extensions with somewhat more difficult conditions: it included categories that are all relevant to respondents (students), and so competitor brands in extension categories are also relatively more familiar to them than in the first study. The researchers acknowledge that the findings are less robust with respect to comparisons of the contrasting competitive settings. Evaluation and perceived risk related to the worse fitting brand competing with less familiar brands are equivalent to the better fitting brand extension facing more familiar brands. The gap in choice shares is reduced though in this case it is still statistically significant (45% vs. 15%, respectively). Facing less familiar brands may not improve the response of consumers to the worse fitting brand extension (i.e., not overcoming the effect of fit) but at least it is in a position as good as of the better fitting brand extension competing in a more demanding setting.

  • Perceived risk intervenes in a more complicated relationship as a mediator of the effect of fit on brand extension evaluation, and also in mediating the effect of relative familiarity in competitive settings. Mediation implies, for example, that a worse fitting extension evokes greater risk which is responsible for lowering the brand extension evaluation; consumers may seek more familiar brands to alleviate that risk.

A parent-brand can assume an advantage in an extension category even though it encounters brands that are familiar within that category, and may even be considered experts in the field: if the extending brand is leading within its original category and is better known beyond it, this can give it a leverage on the incumbents if those brands are more ‘local’ or specific to the extension category. For example, it would be easier for Nikon leading brand of cameras to extend to binoculars (better fit) where it meets brands like Bushnell and Tasco than extending to scanners (also better fit) where it has to face brands like HP and Epson. In the case of worse fitting extensions, it could be significant for Nikon whether it extends to CD players and competes with Sony and Pioneer or extends to laser pointers and faces Acme and Apollo — in the latter case it may enjoy the kind of leverage that can overcome a worse fit. (Product and brand examples are borrowed from Study 1). Further research may enquire if this would work better for novice consumers than experts. Milberg, Sinn and Goodstein recommend to consider additional characteristics that brands may differ on (e.g., attitude, image, country of origin), suggesting more potential bases of strength.

Entering a new product category for a company is often a difficult challenge, and choosing the more appropriate branding strategy for launching the product can be furthermore delicate and consequential. If the management chooses to make a brand extension, it should consider aspects of relative strength of its parent-brand, such as familiarity, against the incumbent brands of the category it plans to enter in addition to a variety of other characteristics of product types and its brand identity. However, the managers can take advantage as well of intermediate solutions in brand architecture to combine a new brand name with an endorsement of an established brand (e.g., higher-level brand for a product range). Choosing the better branding strategy may be helped by better understanding of the differences and relations (e.g., hierarchy) between product categories as perceived by consumers.

Ron Ventura, Ph.D. (Marketing)

Notes:

1. Consumer Reactions to Brand Extensions in a Competitive Context: Does Fit Still Matter?; Sandra J. Milberg, Francisca Sinn, & Ronald C. Goodstein, 2010; Journal of Consumer Research, 37 (October), pp. 543-553.

2.  Consumer Evaluations of Brand Extensions; David A. Aaker and Kevin L. Keller, 1990; Journal of Marketing, 54 (January), pp. 27-41.

3.  Drivers of Brand Extension Success; Franziska Volckner and Henrik Sattler, 2006; Journal of Marketing, 70 (April), pp. 18-34.

4. Brands and Branding: Research Finding and Future Priorities; Kevin L. Keller and Donald R. Lehmann, 2006; Marketing Science, 25 (6), pp. 740-759.

5. Ibid. 1.

Read Full Post »

It is usually not a pleasant feeling to be alone in a scary place or event — think of being stuck in a dark elevator or being involved in a car accident. People commonly seek to be with someone for comfort and company. But the companion does not always have to be another person. A research by Dunn and Hoegg (2014) provides corroboration that the need to share fear matters to humans while the identity of the companion, whether a person or an object, is less critical.  More specifically, sharing fear with a product from an unfamiliar brand may facilitate a quick emotional attachment with that brand without requiring to build a relationship over a lengthy period of time (1).

Fear is evoked by the presence or anticipation of a danger or threat. Feeling fear may be triggered by an unfamiliar event to which one is unsure how to respond (uncertainty) or an unexpected event at a specific moment (surprise); experiencing fear is furthermore likely when the event encountered is both unfamiliar and unexpected. It is important to note, nonetheless, that not every encounter with an unfamiliar or unexpected event necessarily leads to  fear. The Amygdala in the temporal lobe of the brain is the “centre” where fear arises. However, the amygdala like other brain structures is responsible for multiple functions. The amygdala is activated in response to unfamiliarity, unpredictability or ambiguity, but not every instance necessarily means the evocation of fear. For example, tension from facing an unfamiliar problem that one is at loss how to solve may not result in fear. Additionally, fear as well as other states of emotion are the outcome of appraisal of physical feelings (e.g., faster heartbeats, startle, warmth), considering the conditions in which they were triggered; it is a cognitive interpretation of their meaning (“why do I feel that way?”). Activation of other brain structures together with the amygdala may influence whether similar feelings triggered by an unexpected event are interpreted, for instance, as fear, anger, or surprise. The context in which an event occurs can matter a lot for the appraisal of emotions (2).

Dunn and Hoegg emphasise the emotional charge of consumer attachment with a brand versus cognitive underpinnings. Brand attachment has often been conceptualised as the product of a relationship between consumers and the target brand built over time. It should take a longer time to achieve a more solid brand attachment because of cognitive processes for establishing brand connections in memory and stronger favourable brand attitudes. However, this explanation is subject to criticism of missing the important role of emotions in bonding between consumers and a brand which does not necessarily require a long time. By focusing their studies on unfamiliar brands, Dunn and Hoegg intended to show that emotional attachment can emerge much more quickly when the consumers are distressed and are looking for a partner to share their fear with, and that partner or companion can be a brand of a given product.

On the same grounds, the researchers chose a scale of emotional attachment (Thomson, MacInnis and Park, 2005 [3]) as more appropriate over a scale that combines emotional and cognitive aspects of attachment and gives greater weight to cognitive constructs (Park, MacInnis et al., 2010 [4]). The emotional scale comprises three dimensions: (a) Affection (affectionate, friendly, loved, peaceful); (b) Passion (passionate, delighted, captivated); (c) Connection (connected, bonded, attached). Nevertheless, in the later research Park and MacInnis with colleagues offer a broader perspective that accounts for two bases of brand attachment: (i) a connection between self-concept and a brand; and (ii) brand prominence in memory.

While ‘brand prominence’ can be regarded as more cognitive-oriented (accessibility of thoughts and feelings in memory), a ‘brand-self connection’ entails the expansion of one’s concept of self to incorporate others, such as brands, within it — and that involves an emotional element. Park and MacInnis et al. emphasise the brand-self connection as the emotional core of their definition of brand attachment, while brand prominence is a facilitator in actualizing the attachment (analyses substantiate that brand attachment is a better predictor than attitudes of intentions to perform more difficult types of behavior reflecting commitment, and the brand-self connection is more essential for driving this behaviour). The three-dimension scale of emotional brand attachment seems very relevant for the research goals of Dunn and Hoegg, even though it is more restricted from a stand-point of the theoretical roots of brand attachment.

The desire to affiliate with others in scaring and upsetting situations is recognised as a mechanism for coping with negative emotions in those situations. Episodes of armed conflict, terrorist attacks, and natural disasters make people get closer to each other, unite and show solidarity. However, the researchers note that the act of affiliation is essential for coping rather than the affiliation target. That is, the literature on affiliation or attachment relates to interpersonal connections as well as attachment to objects (although objects are viewed as substitutes in absence of other persons [pet animals should also be considered]). We can find support for possible attachment to products and their brands in the human tendency to animate or anthropomorphise objects by assigning them traits of living beings, whether animals or humans. Brands may be animated in order to help consumers relate with them more comfortably, making them appear more vivid to them. It is one of the processes that facilitates the development of consumers’ relationships with their brands in use; consumers connect with brands also through the role brands fulfilled in their personal history, heritage and family traditions, and how brands integrate in their preferred lifestyles (5).

Dunn and Hoegg investigate how consumers connect with a brand on occasions of incidental fear. They make a clear distinction between events that may trigger fear (or other emotions) and fear appeals strategically planned in advertising (e.g., in order to induce a particular desired behaviour). Events that incidentally cause fear would be independent and uncontrolled. Additionally, the intensity and range of emotions felt is expected to differ when consumers actively participate in an event and hence experience it directly in contrast to watching TV ads — in direct consumer experiences, emotional feelings are likely to be more intensive and specific.  In a model for measuring consumption emotions developed and tested by Richins, fear is characterised as a negative and more active (as opposed to receptive) emotion, next to other emotions such as anger, worry, discontent, sadness and shame (6).

  • In their experiments, the researchers try to emulate incidental fear by displaying to participants clips from cinema films or TV series’ episodes, and present evidence that manipulations successfully elicited the intended emotions as dominant in response to each video clip. Yet, it remains somewhat ambiguous how real and direct the experience of watching scenes in a film or a TV programme is perceived and felt with regard to the emotions evoked.

The following are more concrete findings from the studies and their insights:

Emotional brand attachment is generated through perception that the brand shares the fear with the consumer — Study 1 confirms that emotional attachment with an unfamiliar brand is generated when a product (juice) by that brand is present and can be consumed during the fear-inducing experience (more than for emotions of sadness, excitement and happiness). But moreover, it is shown that the emotional attachment is mediated (conditioned) by perception of the consumer that the brand shared the fear with him or her.

Humans precede product brands —  Sharing fear with a brand contributes to stronger emotional brand attachment, but only if they still have a desire generated by fear to affiliate with others. If conversely that desire is satiated by a perception of the consumers that they are already socially affiliated with other people, the effect on brand attachment is muted.

  • Note: Participants in Study 2 were asked to perform a search with words related to feelings of affiliation and social connectedness (e.g., included, accepted, involved) to prime affiliation. Given the statements used to measure (non-)affiliation (e.g., “I feel disconnected from the world around me”), it is a little questionable how effective such a priming condition could be (though the authors show it was sufficient). It might have been more tangible to ask participants to think of people dear to them, family and close friends, and write about them.

Balancing negative and positive emotional effects on attitudes — Based on analyses in Study 2 the researchers also suggest that increased positive effect of emotional brand attachment may counterbalance and override a negative influence of ‘affect transfer’ on attitudes due to fear.

Presence of the brand and attention to it are required yet sufficient — Study 3 demonstrates that neither consumption of the product (juice) nor even touching it (the bottle), both forms of physical interaction, are really needed for feeling affiliated and forming emotional attachment — forced consumption in particular does not contribute to stronger perceived sharing or emotional attachment than merely seeing the product when feeling fear, that is making an eye contact and visually attending to the product in search for a companion. (Unexpectedly, in the case of action and excitement, consuming the drink increases emotional attachment.) Study 4 stresses, nevertheless, that the brand must be present during the emotional event for generating increased emotional attachment — having the brand nearby while experiencing the fear is essential for consumers to feel connected with the brand as their sharing partner (tested with a different product, potato chips).

The research paper suffers from a deficit in practice. That is, marketing managers and professionals might be disappointed to discover that it could be most difficult to have any control of those situations of incidental fear and to act on them to their advantage. In order to have any influence on the consumer a company would be required to anticipate an individual event in advance and to find a way to intervene (i.e., make their product present) without being perceived too intrusive or self-interested — two non-negligible challenges. An additional restriction is posed by the relation of the ‘fear effect’ to brands not previously familiar to the consumers.

Let us consider some potential scenarios where brands might benefit and the difficulties that are likely to arise in implementing it:

Undertaking medical treatments or tests — Some treatments can be alarming and frightening on occasion to different patients. A sense of fear is likely to enter already, and perhaps especially, while waiting. It is a opportunity for introducing the brand-companion in the waiting hall; even more so given that patients are usually not allowed to or prevented from using artifacts during the treatment (mostly no food and drinks). First, a company may have a difficulty to obtain access to places where patients wait for treatment. Second, consumers-patients are likely to bring products with them from home to entertain them (of brands they know). Third, patients often arrive with a family or friend companion, thus satisfying their need for affiliation with another person which dominates affiliation with an object. Still, there is room for ingenuity how to locate the brand close enough to the treatment episode (e.g., shops offering books or toys, especially for children, in the premises of a clinic or hospital).

Trekking or hiking in nature — Some routes, particularly in mountainous areas, can be quite adventurous, not to say dangerous. If a brand could find a way to introduce its product just before the consumer starts the hiking trip, it may benefit from being with him or her if fear arises. One problem is that hikers are advised and even required not to embark alone on more dangerous routes. Another problem is that those trekking or hiking sites often offer local brands, that while not being familiar to the consumers they also are not likely to be available to them at home, and thus the opportunity to develop a relationship based on the early emotional attachment is lost.

Offering legal, financial, insurance, and technical services in events of crisis — In various occasions of accidents, malfunctions, and disasters, people need help to cope with the crisis and the negative emotions it may evoke, particularly fear. A service provider would be expected to counsel the customer in his or her distress, and of course propose a solution (e.g. how to fix one’s home after a fire or an earthquake). Unfortunately,  one cannot make an eye contact with an intangible service. The company has to find creative and practical ways to make itself readily visible and accessible to the consumer when needed by offering instruments and cues for making contact (e.g., an alarm and communication device for the elderly and people with more risky medical conditions).

  • Dunn and Hoegg are aware of the limitation of the findings to unfamiliar brands. They reasonably propose that “because fear leads to a general motivation to affiliate, emotional brand attachment would be enhanced regardless of the familiarity with the brand” (p. 165). It should take further research, however, to substantiate this proposition.

Despite the possible difficulties companies will likely need to deal with, the doors are not completely shut to them to benefit from this phenomenon. But they must come up with creative and non-intursive solutions to make their brands and products present in the right place at the right time. At the very least, marketers should be aware of the potential effect of sharing fear with the consumer and understand how it can work in the brand’s benefit. It is worth remembering, after all, the saying “a friend in need is a friend indeed” whereby in some incidents the friend can be a brand.

Ron Ventura, Ph.D. (Marketing)

Notes:

(1) “The Effect of Fear on Emotional Brand Attachment”; Lea Dunn and JoAndrea Hoegg, 2014; Journal of Consumer Research, 41 (June), pp. 152-168.

(2) “What Is Emotion?: History, Measures and Meanings”; Jerome Kagan, 2007; New Haven and London: Yale University Press. Also see: “The Experience of Emotion”; Lisa Feldman Barrett, Bejta Mesquita, Kevin N. Ochsner, & James J. Gross, 2007; Annual Review of Psychology, 58, pp. 373-403.

(3) “The Ties That Bind: Measuring the Strength of Consumers’ Emotional Attachments to Brands”; Mathew Thomson, Deborah J. MacInnis, & C. Whan Park, 2005; Journal of Consumer Psychology, 15 (1), pp. 77-91.

(4) “Brand Attachment and Brand Attitude Strength: Conceptual and Empirical Differentiation of Two Critical Brand Equity Drivers”; C. Whan Park, Deborah J. MacInnis, Joseph Priester, Andreas B. Eisengerich, & Dawn Iacobucci, 2010; Journal of Marketing, 74 (November), 1-17.

(5) “Consumers and Their Brands: Developing Relationship Theory in Consumer Research”; Susan Fournier, 1998; Journal of Consumer Research, 24 (March), pp. 343-373.

(6) “Measuring Emotions in the Consumption Experience”; Marsha L. Richins, 1997; Journal of Consumer Research, 24 (September), pp. 127-146.

Read Full Post »

The location-based technology of beacons is a relatively recent newcomer in the retail scene (since 2013). Beacons provide an additional route for interacting with shoppers in real-time via their smartphones as they move around in stores and malls. Foremost, this technology is about marrying between the physical and the digital (virtual) spaces to create better integrated and encompassing shopping experiences.

It is already widely acknowledged that in-store and online shopping are not independent and do not happen completely separate from each other; instead, experience and information from one scene can feed and drive a shopping experience, and purchase, in the other scene. In particular, mobile devices enable shoppers to apply digital resources while shopping in a physical shop or store.  Beacons may advance retailers and shoppers another step forward in that direction, with the expectation to generate more purchases in-store. The beacon technology was received at first with enthusiasm and promising willingness-to-accept by retailers, but these subdued in the past year and adoption has stalled. A salient obstacle appears as consumers remain hesitant and cautious about letting retailers communicate through beacons with their smartphones and the implications it may have on their privacy.

In essence, beacons are small, battery-powered, low-energy Bluetooth devices that function as transmitters of information — primarily unique location signals — to nearby smartphones with an app authorised to receive the information. The availability of an authorised app (e.g., retailer’s, mall operator’s) installed on the consumer’s smartphone (or tablet) is critical for the communication technology to function properly. Upon receiving a location signal, the app is thereby triggered to display location-relevant content for the shopper in-store (e.g., product information, digital coupons, as well as store activities and services).

Additional requirements may be in force such as the retailer’s app being open during the shopping trip or that the shopper consents (opts-in) to allow the app receive information from beacons, but these do not seem to be necessary or mandatory conditions for the technology to work (e.g., an app may be set with ‘approval’ as default). Ambiguity that seemingly prevails about the extra requirements could be one of the sour points in the technology’s implementation. On one hand, the application of beacons is more ethical when setting up at least one of these requirements, and should endow it with greater credibility among consumers. On the other hand, any additional criterion for access of beacons to smartphones — assuming the app is already installed — could limit further the number of participating shoppers and reduce its marketing impact.

  • Only smartphones (and tablets) support apps, not any mobile phone. It should not be taken for granted that everyone has supporting smartphones, hence raising another possible limiting requirement on access for beacons (though in decline in developed countries). Another problem, yet, concerns the distinction between Apple iPhones operated with iOS and smartphones of other brands operated with Google’s Android — beacons have to work with either type of operating system and compatible apps but they do not necessarily do so (e.g., iBeacons are exclusive for Apple’s own mobile devices).

There are some more variations in the application of beacon technology in retail. Beacon devices may be attached to shelves next to specific product displays or to fixtures and building columns in positions aimed at capturing smartphones of shoppers moving in a close area (e.g., an aisle). If the beacon is associated with a particular product, the shopper may engage using the app by actively approaching the phone to the beacon. Otherwise, the app communicates with the beacons without  shoppers taking any voluntary action. Furthermore, some applications of beacon technology suggest sending information other than location signals from the beacon, such as product-related information, and receiving customer-related information by the beacon from the smartphone.

Reasonably, retailers would be interested first in applications of the technology for practical marketing purposes in their stores. However, beacon technology may also be utilised in research on shopper behaviour, a purpose now appreciated by many large retailers.

Marketing Practice in Retail

The instant sales-driven idea of application of beacon technology evoked by retailers is to introduce special offers, discount deals and digital coupons for selected products as shoppers get near to their displays. Notwithstanding this type of application, location-based features and services enabled via beacons can be even more creative and useful for shoppers, and beneficial for the retailers.

Relevance is key in achieving an effective application of the technology. Any message or content must be relevant in time and place to the shopper. That is, the content must be related to available products when the shopper is getting close enough to them. The content should not be too general in reference to any product in the store but to products in a section of the store where the shopper passes-by. Triggering an offer for a product just after the shopper entered a store is less likely to be effective, unless, for example, there is a special promotional activity for it in a main area of the floor. The retailer should not err in introducing an offer for a product item that is not available in the specific store at that time. Furthermore, if the app can link product information with customer information, it may be able to generate better content that is both location-relevant and personalised. The app could make use of accessible information on personal purchase history, interests and demographic characteristics. This higher-level application surely requires greater resources and effort of the retailer to implement.

The beacons’ greatest enemy could be their use for bombardment of shoppers with push or pop-up messages of offers, deals, discounts etc. This practice is suspected as a major fault in the early days of the technology that may be responsible for the slowdown in adoption lately. There could be nothing more irritating for a shopper if every few meters walked in the store he or she is interrupted by a buzz and message of “just today offer on X” that appears on the smartphone’s screen. Retailers have to be selective lest customers will avoid using their apps. It is much more important to produce adaptive, relevant and customer-specific messages and content overall (Adobe, Digital Marketing Blog, 4 February 2016).

  • The grocery retail chain Target, that launched a trial with beacons in 50 US stores in the second half of 2015, committed, for instance, to show no more than two promotional (push) messages during a store visit (TechCrunch.com, 5 Aug. ’15).

More intelligent and helpful ways exist to apply the beacon technology in interaction with the app than promotional push messages. First, content of the “front page” of the app can change as the shopper progresses in the store to reflect information that would be of interest to the shopper in that area of the store (e.g., show hyper-linked ’tiles’ for nearby product types). Second, beyond ‘technical’ information on product characteristics and price, a retailer can facilitate shopper-user access to reviews and recommendations for location-relevant products via the app. Third, if the shopper fills-in a shopping list on a retailer’s app (e.g., a supermarket), and the app has a built-in plan of the store, it can help the shopper navigate through the store to find the requested products, and it may even re-order the list and propose to the shopper a more ‘efficient’ path.

Beacons are associated mostly with stores (e.g., department stores, chain stores, supermarkets). However, beacons may also be utilised by mall operators where the ‘targets’ are stores rather than specific products. An application programme in a mall may command collaboration with the retailers (e.g., store profile and notifications, special promotional messages [for extra pay], content contributions).

In another interesting form of collaboration, the fashion magazine Elle initiated a programme with ShopAdvisor, a mobile app and facilitator that assists retailers in connecting with their shoppers through beacons. As an enhancement to its special 30th anniversary issue, Elle launched a trial project in partnership with some of its advertisers (e.g., Guess, Levi’s, Vince Camuto) to introduce their customers to location-based content with the help of ShopAdvisor (focused on promotional alerts)(1).

Consumers are concerned about tactics of location-based technologies like beacons that get intrusive and even creepy; they become adverse towards the way such apps sometimes surprise them (e.g., in dressing rooms). Indeed, only shoppers who installed an authorised app can be affected, but for customers who installed such a retailer’s app, with other benefits in mind, it can be disturbing at times. The hard issue at stake is how the app alerts or approaches its shoppers-users with location-based messages. Shoppers do not like to feel that someone is watching where they go.

The shopper may believe that if the app remains closed on the smartphone he or she cannot be approached. But if, as reported in CNBC News, a dormant app can be awaken by a beacon signal, this measure is not enough. This may happen because the shopper previously allowed the app to receive the Bluetooth signal or the app “assumed” so as default.  The shopper must take an extra step to disable the function at the app-level or device-level (Bluetooth connectivity). Retailers should let their customers opt-out and be careful in any attempt to remotely open their apps on smartphones (so-called “welcome reminders”), because imposing and interfering with customer choices may get the opposite outcome of removing the app.

The app may display ‘digital’ coupons for the shopper to “pick-up” and show later at the cashier (or self-service check-out). It is reasoned that if coupons are shown at the right time shoppers will welcome the offer, no resentment. The manner shoppers are alerted can also matter, by not being too obtrusive (e.g., “Click here for coupons for products in this aisle”). Shoppers told CNBC News that if digital coupons were offered to them by the app just when relevant, they would be glad to use this option, being more convenient than going around with paper coupons, but they would want the ability to opt-out.

Shopper Behaviour Research

The beacon technology may further contribute to research on shopper behaviour in stores or malls. Specifically, it may be suitable for collecting data of shopper traffic to be used in path analysis of the shopping journeys. The information may cover what areas of the store shoppers visit more frequently, how long one stays in a given area, and sequences of passes between areas.

Nonetheless, there are methodological, technological and ethical factors retailers and researchers have to consider. At this time, there are distinct limitations to be recognized that may inflict on the validity and reliability of the research application of beacons. Ethical issues discussed above regarding the provision of access of beacons to mobile apps furthermore apply in the research context.

This methodology involves tracking the movements of shoppers. Beacon technology may record frequency of visits in each area of the store separately or it may track the presence of a particular shopper by different beacons across the store. A beacon may also be able to send repeated signals at fixed intervals to a smartphone to measure how long a shopper remains in a given area. However, this type of research is not informative about what a shopper does in a specific location as in front of product shelves, and thus it cannot provide valuable details on her decision processes. Hence, retailers cannot rely on this methodology as a substitute for other methods capable of studying shopper behaviour more deeply, especially with respect to decision-making. A range of methods may be used to supplement path analysis such as interviewer’s walk-along with a shopper, passive observations, video filming, and possibly also in-store eye-tracking.

An implementation of the technology for research would require a comprehensive coverage of the premises with beacons, perhaps greater than needed for marketing practice. It should be compared with alternative location-based technologies (e.g., Radio Frequency Identification [RFID], Wi-Fi)  on criteria of access, range and accuracy, and of course cost-effectiveness. For example, the RFID technology employs tags ( transmitters) regularly attached to shopping carts — if a shopper leaves the cart at the end-of-aisle and goes in to pick-up a couple of products, the system will miss that; smartphones, however, are carried on shoppers all the time. Beacon technology may have an important advantage over RFID if location data is linked with customer characteristics, but this is a sensitive ethical issue and at least it is imperative to ensure no personal IDs are included in the dataset. All alternative technologies may also have to deal with different types of environmental interferences with their signals. Access would have both technical and ethical aspects.

A mixture of problems emerges as responsible for impairing the utilisation of beacon technology, according to RetailDive (online news and trends magazine), mainly consumers who do not perceive beacon-triggered features as useful enough to them and retailers troubled by technical or operational difficulties. Among the suggestions made: encourage pull of helpful information from beacons by shoppers rather than push messages, and speed-up calling staff for assistance via beacons (RetailDive, 17 December 2015). A recent research report by Adobe and Econsultancy on Digital Trends for 2016 indicates that retailers are becoming more reluctant to implement a geo-targeting technology like beacons this year compared with 2015 (a decrease in proportion of retailers who have this technology in plan or exploring it, against an increase in proportion of those who are not exploring or do not know). Conspicuously, there seems to be much more optimism about high effectiveness of geo-targeting technology at technology and consultancy agencies than among retailers, who seem to be much more in the opinion that it is too early (2). Agencies could have better understanding of the field, yet it signals an alarm of disconnect between agencies and their clients.

There is potential to beacon technology with clearly identifiable benefits it can deliver to retailers and consumers. It is still a young technology and requires more development and progress on various technical, applied and ethical aspects.  Promotional messages are  important tools but must be used in a good and sensible measure. A retailer cannot settle for a small set of fixed messages. It has to develop a dynamic ‘bank’ of messages, large enough to be versatile over products, (chain) stores, and consumer groups, and maintain regular updates. However, retailers have to develop and provide a more rich suite of clever content and practical tools based on location. Consumers will have to be convinced of the benefits enabled by beacons, yet feel free to decide when and how to enjoy them.

Ron Ventura, Ph.D. (Marketing)

Notes:

(1) “App Helps Target Shoppers’ Location and Spontaneity”, Glenn Rifkin, International New-York Times, 31 December 2015 – 1 January 2016.

(2) “Quarterly Digital Intelligence Briefing: 2016 Digital Trends”, Adobe and Econsultancy, January 2016 (pp. 24-25). The findings are considered with caution because of relatively small sub-samples of respondents on this topic (N < 200).

Read Full Post »

Psycographic-oriented research of consumer lifestyles based on surveys for collecting the data is losing favour among marketers and researchers. Descriptors of consumer lifestyles are applied especially for segmentation by means of statistical clustering methods and other approaches (e.g., latent class modelling). Identifying lifestyle segments has been recognized as a strategic instrument for marketing planning because this kind of segmentation has been helpful and insightful in explaining variation in consumer behaviour where “dry” demographic descriptors cannot reach the deeper intricacies. But with the drop in response rates to surveys over the years, even on the Internet, and further problematic issues in consumer responses to survey questionnaires (by interview or self-administered), lifestyle research using psychographic measures is becoming less amenable, and that is regrettable.

The questionnaires required for building lifestyle segmentation models are typically long, using multi-item “batteries” of statements (e.g., response on a disagree-agree scale) and other types of questions. Initially (1970s) psychographics were represented mainly by Activities, Interests and Opinions (AIO). The measures cover a wide span of topics or aspects from home and work, shopping and leisure, to politics, religion and social affairs. But this approach was criticised for lacking a sound theoretical ground to direct the selection of aspects characterising lifestyles that are more important, relevant to and explanatory of consumer behaviour. Researchers have been seeking ever since the 1980s better-founded psychology-driven bases for lifestyle segmentation, particularly social relations among people and sets of values people hold to. The Values and Lifestyles (VALS) model released by the Stanford Research Institute (SRI) in 1992 incorporated motivation and additional areas of psychological traits (VALS is now licensed to Strategic Business Insights). The current version of the American model is based on that same eight-segment typography with some updated modifications necessary to keep with the times (e.g., the rise of advanced-digital technology) — the conceptual model is structured around two prime axes, (a) resources (economic, mental) and (b) motivation or orientation. Scale items corresponding to the AIOs continue to be used but they would be chosen to represent constructs in broader or better-specified contexts.

Yet the challenge holds even for the stronger-established models, how to choose the most essential aspects and obtain a sufficient set of question items consumers are likely to complete answering. Techniques are available for constructing a reduced set of items (e.g., a couple of dozens) for subsequent segmentation studies relying on a common base model, but a relatively large set (e.g., several dozens to a few hundreds of items) would still be needed for building the original model of lifestyle segments. It is a hard challenge considering in particular the functions and limitations of more popular modes of surveys nowadays, online and mobile.

Lifestyles reflect in general the patterns or ways in which people run their ordinary lives while uncovering something of the underlying motives or goals. However,  ‘lifestyles’ have been given various meanings, and researchers follow different interpretations in constructing their questionnaires. The problem may lie in the difficulty to construct a coherent and consensual theory of ‘lifestyles’ that would conform to almost any area (i.e., product and service domain) where consumer behaviour is studied.  This may well explain why lifestyle segmentation research is concentrated more frequently on answering marketing questions with respect to a particular type of product or service (e.g., banking, mobile telecom, fashion, food). It can help to select more effectively the aspects the model should focus on and thereby also reduce the length of the questionnaire. The following are some of the concepts lifestyle models may incorporate and emphasise:

  • Values that are guiding and driving consumers (e.g., collectivism vs. individualism, modernism vs. traditionalism, liberalism vs. conservatism);
  • In the age of Internet and social media consumers develop new customs of handling social relations in the virtual world versus the physical world;
  • In view of the proliferation of digital, Internet and mobile communication technologies and products it is necessary to address differences in consumer orientation and propensity to adopt and use those products (e.g, ‘smart’ products of various sorts);
  • How consumers balance differently between work and home or family and career is a prevailing issue at all times;
  • Lifestyles may be approached through the allocation of time between duties and other activities — for example, how consumers allocate their leisure time between spending it with family, friends or alone (e.g., hobbies, sports, in front of a screen);
  • Explore possible avenues for developing consumer relationships with brands as they integrate them into their everyday way of living (e.g., in reference to a seminal paper by Susan Fournier, 1998)(1);
  • Taking account of aspects of decision-making processes as they may reflect overall on the styles of shopping and purchasing behaviour of consumers (e.g., need for cognition, tendency to process information analytically or holistically, the extent to which consumers search for information before their decision).

Two more issues deserve special attention: 

  1. Lifestyle is often discussed adjacent with personality. On one hand, a personality trait induces a consistent form of response to some shared stimulating conditions in a variety of situations or occasions (e.g., responding logically or angrily in any situation that creates stress or conflict, offering help whenever seeing someone else in distress). Therefore, personality traits can contribute to the model by adding generalisation and stability to segment profiles. On the other hand, since lifestyle aspects describe particular situations and contexts whereas personality traits generalize across them, it is argued that these should not be mixed as clustering variables but may be applied in complementary modules of a segmentation model.
  2. Products that consumers own and use or services they utilize can illustrate  figuratively their type of lifestyle. But including a specific product in the model framework may hamper the researcher’s ability to make later inferences and predictions on consumer behaviour for the same product or a similar one. Therefore, it is advisable to refer carefully to more general types of products distinctively for the purpose of implying or reflecting on a pattern of lifestyle (e.g., smartphones and technology-literacy). Likewise, particular brand names should be mentioned only for an important symbolic meaning (e.g., luxury fashion brands, luxury cars).

Alternative approaches pertain to portray lifestyles yet do not rely on information elicited from consumers wherein they describe themselves; information is collated mostly from secondary databases. Geodemographic models segment and profile neighbourhoods and their households (e.g., PRIZM by Claritas-Nielsen and MOSAIC by Experian). In addition to demographics they also include information on housing, products owned (e.g., home appliances), media used, as well as activities in which consumers may participate. However, marketers are expected, by insinuation, to infer the lifestyle of a household, based, for instance, on appliances or digital products in the house, on newspaper or magazine subscriptions, on clubs (e.g., sports), and on associations that members of the household belong to. Or consider another behavioural approach that is based on clustering and “basket” (associative) analyses of the sets of products purchased by consumers. These models were not originally developed to measure lifestyles. Their descriptors may vicariously indicate a lifestyle of a household (usually not of an individual). They lack any depth in describing and classifying how consumers are managing their lives nor enquiring why they live them that way.

The evolving difficulties in carrying-out surveys are undeniable. Recruiting consumers as respondents and keeping them interested throughout the questionnaire is becoming more effortful, demanding more financial and operational resources and greater ingenuity. Data from surveys may be complemented by data originated from internal and external databases available to marketing researchers to resolve at least part of the problem. A lifestyle questionnaire is usually extended beyond the items related to segmentation variables by further questions for model validation, and for studying how consumers’ attitudes and behaviour in a product domain of interest are linked with their lifestyles. Some of the information collected thus far through the survey from respondents may be obtained from databases, sometimes even more reliably than that based on respondents’ self-reports. One of the applications of geodemographic segmentation models more welcome in this regard is using information on segment membership as a sampling variable for a survey, whereof characteristics from the former model can also be combined with psychographic characteristics from the survey questionnaire in subsequent analyses. There are furthermore better opportunities now to integrate survey-based data with behavioural data from internal customer databases of companies (e.g., CRM) for constructing lifestyle segments of their customers.

Long lifestyle questionnaires are particularly subject to concerns about the risk of respondent drop-out and decreased quality of response data as respondents progress in the questionnaire. The research firm SSI (Survey Sampling International) presented recently in a webinar (February 2015 via Quirk’s) their findings and insights from a continued study on the effects of questionnaire length and fatigue on response quality (see a POV brief here). A main concern, according to the researchers, is that respondents, rather than dropping-out in the middle of an online questionnaire, actually continue but pay less attention to questions and devote less effort answering them, hence decreasing the quality of response data.

Interestingly, SSI finds that respondents who lose interest drop-out mostly by half-way of a questionnaire irrespective of its length, whether it should take ten minutes or thirty minutes to complete. For those who stay, problems may yet arise if fatigue kicks-in and the respondent goes on to answer questions anyway. As explained by SSI, many respondents like to answer online questionnaires; they get into the realm but they may not notice when they become tired or do not feel comfortable to leave before completing the mission, so they simply go on. They may become less accurate, succumb to automatic routines, and give shorter answers to open-end questions. A questionnaire may take forty minutes to answer but in the estimation of SSI’s researchers respondents are likely to become less attentive after twenty minutes. The researchers refer to both online and mobile modes of survey. They also show, for example, the effect of presenting a particular group of questions in different stages of the questionnaire.

SSI suggests in its presentation some techniques for mitigating those data-quality problems. Two of the suggestions are highlighted here: (1) Dividing the full questionnaire into a few modules (e.g., 2-3) so that respondents will be invited to answer each module in a separate session (e.g., a weekly module-session); (2) Insert break-ups in the questionnaire that let respondents loosen attention from the task and rest their minds for a few moments — an intermezzo may serve for a message of appreciation and encouragement to respondents or a short gaming activity.

A different approach, mentioned earlier, aims to facilitate the conduct of many more lifestyle-application studies by (a) building once a core segmentation model in a comprehensive study; (b) performing future application studies for particular products or services using a reduced set of question items for segmentation according to the original core model. This approach is indeed not new. It allows to lower the burden on the core modelling study from questions on product categories and release space for such questions in future studies dedicated to specific products and brands. One type of technique is to derive a fixed subset of questions from the original study that are statistically the best predictors of segment membership. However, a more sophisticated technique that implements tailored (adaptive) interviewing was developed back in the 1990s by the researchers Kamakura and Wedel (2).

  • The original model was built as a latent class model; the tailored “real-time” process selected items for each respondent given his or her previous responses. In a simulated test, the majority of respondents were “presented” with less than ten items; the average was 16 items (22% of the original calibration set).

Lifestyle segmentation studies are likely to require paying greater rewards to participants. But that may not be enough to keep them in the survey. Computer-based “gamification” tools and techniques (e.g., conditioning rewards on progress in the questionnaire, embedding animation on response scales) may help to some extent but they may also raise greater concerns for quality of responses (e.g., answering less seriously, rushing through to collect “prizes”).

The contemporary challenges of conducting lifestyle segmentation research are clear. Nonetheless so should be the advantages and benefits of applying information on consumer lifestyle patterns in marketing and retailing. Lifestyle segmentation is a strategic tool and effort should persist to resolve the methodological problems that surface, combining where necessary and possible psychographic measures with information from other sources.

Ron Ventura, Ph.D. (Marketing)

Notes:

(1) Consumers and Their Brands: Developing Relationship Theory in Consumer Research; Susan Fournier, 1998; Journal of Consumer Research, 34 (March), pp. 343-373

(2) Lifestyle Segmentation With Tailored Interviewing; Wagner A. Kamakura and Michel Wedel, 1995; Journal of Marketing Research, 32 (Aug.), pp. 308-317.

Read Full Post »

During a shopping journey in a store where a consumer intends to buy multiple products, he or she is required to make a sequeqnce of choice decisions. Each decision is about to be made in a category with different product attributes, but beyond that there could also be differences in the settings of the choice situations, such as the size of the choice set, the structure of information display for product items, and information format. The transition between choice problems that differ in their characteristics should require shoppers to make some adjustments in preparation to reach a decision, each time in somewhat different settings. This is in fact true when filling a basket either in a physical store or on a website of an online store — shoppers have to shift between decision problems, and on the way they may need to replace or correct their choice strategy.

Researchers have been studying the paths that shoppers frequently follow, moving between sections of a store during their shopping trip. This type of research usually focuses on identifying and depicting the sequence in which store sections and product categories are visited, and the frequency in which category displays are stopped-by. However, the transitions from a choice decision in one category to another may also have  consequences for the decision process in any single category visited (e.g., as in adjusting for every new choice problem). Moreover, the sequence or order in which choice problems are resolved may have an effect on particular decisions.

  • Different techniques are applied for tracking the pathways of shoppers in brick-and-mortar stores (e.g., RFID, mobile-based GPS, video recording through surveillance cameras). Studies in supermarkets have shown what areas of a store shoppers approach first, and how they start by walking to the back of the store and then make incursions into each aisle (without leaving the aisle on the other end but returning to their point-of-entry). Hui, Bradlow and Fader reveal that as shoppers spend more time at the store, the checkout looms more attractive — the shoppers who feel a stronger time pressure become more likely to go through an aisle and approach a checkout counter. As perceived time pressure increases shoppers also tend to cut-off exploration and concentrate on visiting product displays from which they are most likely to purchase. (1)

Consumers have been described as adaptive decision-makers who adjust their decision strategies according to characteristics of the problem structure or context — for example, the amount of information available (given the number of alternatives or attributes), the type of information (e.g., scale, units), or the order in which information elements are displayed. In the outset, consumers may be guided by top-down goals — maximizing accuracy (relative to a maximum-utility ‘rational’ rule) and minimizing cognitive effort; a decision strategy (i.e., a rule like Equal-Weights or Lexicographic) can be selected in advance with respect to the accuracy-effort trade-off assessment of rules in a given choice situation, this according to Payne, Bettman and Johnson. However, they argue that this approach may not be sufficient on various occasions. When the characteristics of a choice problem are not familiar to the consumer, he or she will construct a strategy step-by-step as the structure and detail of information on alternatives is observed and learned. Even in cases the choice situation and context are familiar, the consumer may face unexpected changes or updates in information (e.g., inter-attribute relations) that may require her or him to modify the strategy. Hence, a consumer who started with a specific rule may replace it with another on-the-fly in response to data encountered, and often elements from different rules may be combined into an adaptive new choice strategy (as opposed to a ‘pure’ strategy)(2).

The construction of a decision strategy is therefore frequently the product of a delicate balance between top-down (goal driven) and bottom-up (data driven) processing. When in particular preferences also are not well-established by the consumer, preferences (e.g., importance weights of attributes) also are formed or constructed as one proceeds in the decision process. In such a case the preferences formed would be more contingent on the particular process followed and the strategy constructed thereby. Bettman, Luce and Payne extended the constructive choice model and added to the goals of maximizing accuracy and minimizing effort two more goals (directed by a perceptual framework): minimizing negative emotions (e.g., perceived losses, difficult trade-offs) and maximizing the ease of justifying decisions (to others or to oneself). (3)

However, the adaptation of consumers may not be complete, and thus a shopper may not fully “reset” or fit his decision strategy to features of the next choice problem, which may differ from features of the previous choice setting. Levav, Reinholtz and Lin investigated specifically the impact of one characteristic of decision problems on a decision process: the number of alternatives (4). They tested how many alternatives consumers would inspect more closely from each choice set, if the total number of alternatives available increases from the first to the last decision problem (e.g., 5, 10, 15 and so on until 50), versus a decrease in the number of alternatives available from the first to the last decision (e.g., 50, 45, 40 and so on until 5 — participants were allowed to sample songs to listen to before choosing a song for each track on a disc).

In one of the decision contexts tested, most relevant here, the researchers simulated an online shopping trip: participants in the experiment were asked to choose in sequence from eight different product categories (e.g., body lotions, energy bars, notebooks, shampoo). For some of the participants the number of alternatives increased between categories (i.e., 5, 8, 13, 17, 20, 23, 26, 30) whereas for the others the number of alternatives in a choice set changed in a reverse order (product categories were also presented in two opposite sequences of alphabetical order). Participants could examine more closely each option in a choice set by mouse-hovering on a thumbnail photo of the product item to see its enlarged photo image, its price, and a short product description.

  • Note: In a physical store the equivalent would be picking a product package from a shelf, inspecting it from different angles, reading the label etc. Advanced 3-D graphic simulators let a user-shopper in a like fashion to virtually “pick” a product item from a shelf display image, rotate it, “zoom-in” to read more clearly its label, etc.

Levav and his colleagues found that the direction in which the size of the choice set changes matters, and that particularly a low or high number of options in the first decision problem induces consumers to examine more or less information on options through the shopping trip. If a shopper starts with a small choice-set, he or she is more strongly inclined to inspect every option or acquire more information on each option available. This tendency endures in the next choice problems as the number of options increases, though it may level-off at some point.

In the online shopping experiment, the “shoppers” in the increasing condition examine on average the description for each option more times than “shoppers” in the decreasing condition for smaller choice sets. The former gradually adjust downward the amount of information acquired on each option but the amount of information “gathered” overall does not decrease; for relatively small choice sets (up to 13 options) they would still examine more information on options than “shoppers” who started their journey with the largest choice set. A “shopper” who starts with a large choice set constrains himself from the beginning to inspect options less closely; even as the choice set may become more “manageable” in size, the average “shopper” does not intensify the examination of information on single options considerably, clearly not to the level as “shoppers” whose first decision is from the smallest choice set.

  • For choice sets larger than 17-20 options, where the task for “shoppers” in the increasing condition may become too time-and-effort consuming and “shoppers” in the decreasing condition may still feel too pressed, the level of information acquisition is more similar.

The researchers refer to this form of behaviour as “bounded adaptivity“; they explicate: “Our results indicate that people are actually “sticky adapters” whose strategies are adapted to new contexts — such as the initial choice set — but persist to a significant degree even in the face of changes in the decision environment” (p. 596). The authors suggest, based on results from one of their experiments, that an increasing condition, where consumers’ first choice decision is made from a small choice set, may activate in  consumer a ‘maximizing’ mind-set, searching deeper into information on alternatives (as opposed to a probable ‘satisficing’ mind-set of a consumer in a condition of decreasing size of choice set). Levav et al. note that while ‘maximizing’ has often been regarded in literature as a chronic trait of personality, they see the possibility that this mind-set can be triggered by a decision situation.

If decisions during the shopping trip are not made independently, since adaptation where necessary is not complete or “sticky”, studying in isolation the decision process a shopper goes through in front of a particular product display could be misleading. For instance, the shopper’s decision strategy may be influenced by a choice strategy used previously.  An “imperfect” or “sticky” adaptivity does not have to reflect a deficiency of the consumer-shopper. It may simply designate the sensible level of adaptivity needed in a given decision situation.

(1) Shoppers may not have to hurry to modify their strategy if the perceived change in conditions of the choice problem is small enough to allow them to act similar as before. Shoppers can often adjust their decision tactic gradually and slowly until they get to a situation when a more significant modification is required. (“Shoppers” in the decreasing condition above seem to be more “in fault” of remaining “sticky”.)

(2) Shoppers-consumers look for regularities in the environment in which they have to decide and act (i.e., arrangement of products, structure and format of information) that can save them time and effort in their decision process. Regularities are exhibited in the ways many stores are organised (e.g., repetitive features in display of products) that shoppers can gain from in decision efficiencies. Regularities are likely to reduce the level of ongoing adpativity shoppers may need to exercise.

(3) On some shopping trips, ordinary or periodic (e.g., at the supermarket), shoppers frequently do not have the time, patience or motivation to prepare and deliberate on their choice in every category candidate for purchase. They tend to rely more on routine and habit. Prior knowledge of the store (e.g., one’s regular neighbourhood store) is beneficial. Shoppers would want to adapt more quickly, perhaps less carefully or diligently, and they may be more susceptible to “sticky” adaptivity.

It can be difficult to influence when and how shoppers attend to various sections or displays for performing their decision in differing choice settings. But it is possible to identify what zones shoppers are more likely to visit in early stages of their shopping trip. If a store owner or manager wants to induce shoppers thereafter to search product selections at greater depth, he or she may arrange in those locations displays with a small number of options for a product type. It should be even easier to track movements and direct shoppers to planned sections in an online store website. On the other hand, the retailer may stage a display with some surprising or unexpected information features for disrupting the ordinary search, and induce shoppers to work-out their decision strategy more diligently, thus devoting more attention to the products. However, this tactic should be used more carefully and restrictively so as not to turn-away frustrated or agitated customers.

Displays in the store (physical or virtual) and information conveyed on product packaging (including graphic design) together influence the course of consecutive decision processes shoppers apply or construct.

Ron Ventura, Ph.D. (Marketing)

Notes:

(1) Testing Behavioral Hypotheses Using an Integrated Model of Grocery Store Shopping Path and Purchase Behavior; Sam K. Hui, Eric T. Bradlow, & Peter S. Fader, 2009; Journal of Consumer Research, 36 (Oct.), pp. 478-493.

(2) The Adaptive Decision Maker; John W. Payne, James R. Bettman, & Eric J. Johnson, 1993; Cambridge University Press.

(3) Constructive Consumer Choice Processes; James R. Bettman, Mary Frances Luce, & John W. Payne, 1998; Journal of Consumer Research, 25 (Dec.), pp. 187-217.

(4) The Effect of Ordering Decisions by Choice Set Size on Consumer Search; Jonathan Levav, Nicholas Reinholtz, & Claire Lin, 2012; Journal of Consumer Research, 39 (Oct.), pp. 585-599.

Read Full Post »

The concept of brand attachment has become a frequent, almost integral component of attitudinal models of brand equity in commercial studies since the late 1990s. It has been introduced to represent an emotional bond that is expected to build between a brand and consumers to allow for their sustainable loyalty to the brand. Perceived quality and other assessments of a rational nature about branded products and services are generally not regarded as sufficient to connect a consumer with a brand. In addition, brand attachment is meant to represent a disposition towards a brand that is more solid and enduring regarding consumer-brand relations than brand attitude. However, what signifies and distinguishes that construct has not been properly and agreeably defined.  In a recent seminal article (2010), Park and MacInnis of the University of Southern California and their colleagues (*) offer an approach to fill this caveat coherently. The researchers define the construct of brand attachment, test and demonstrate how it differs from brand attitude strength.

We may find a spectrum of relationships between consumers and brands at different levels of depth and strength. What creates, for example, the deep attraction, to an extent of passion, of consumers to brands like Nike or Apple? On the contrary, the brand Nokia of mobile phones has for the past few years lost its favour with consumers. Last month it was revealed that Microsoft, which had acquired the mobile division from the Finnish mother company, intends to abolish the Nokia brand name but retain the “Lumia” name for new models of smart-mobile devices (e.g., phones, tablets, and whatever comes next); if indeed brand attachment by consumers to Nokia has diminished, no one would shed a tear. Coca Cola almost ruined its brand equity in 1985 due to its New Coke ordeal — apparently consumers were insistent on their attachment to the brand and what it represented to them to force the company to step back and save the brand. Brand attachment captures the affective linkage that is created between a brand and its customers.

A brand equity model may start from awareness of and basic familiarity with the brand of interest at its ground. On top of which should come associations of product attributes and functional benefits (tangible product assets) next to “softer” associations of feelings or personality traits assigned to the brand (intangible assets). After accounting for these building blocs, a bridge of attachment can be erected between the consumers and the brand, leading to commitment (a manifest of attitudinal loyalty). Several facets can be proposed based on pervious academic and applied research in the field to represent brand attachment: (a) respect for the brand and its leadership; (b) personal identification with the brand; (c) favourability of brand legacy and values; and possibly also (d) appreciation of how the brand treats its customers.

Park and MacInnis et al. develop a scale of brand attachment that formally specifies aspects of brand-self connection — it emphasises identification of the consumer with the brand; yet to this factor they add a second factor of brand prominence in memory. Thus, they suggest a scale constructed from two factors; they show that treating the scale as a composition of the two components has better validity than a single-unified scale. Furthermore, the authors demonstrate the effect of brand attachment on behavioural intentions as well as actual behaviour (self-reported and as registered in customer database records).  They cover a range of activities or actions that differ in their level of difficulty.  It is shown that brand attachment is able to predict the intention to perform the more difficult types of behaviour that brand attitude strength cannot.

Brand attitude strength is measured by the valence of an attitude (positive-negative) weighted by the confidence with which the consumer holds that attitude. However, research repeatedly has shown that attitudes get to impact behaviour when the valence is more extreme in either direction and confidence is strong. Attitudes do have an affective basis but it is generally sublime and concerned primarily with valence. That is, brand attitude alone does not contain a scope of emotions people may exhibit; it is very limited in its emotional capacity. Brand attachment, on the other hand, is more emotionally charged and can tell a better story about the relation of the consumer to the brand. Park and MacInnis et al. conceptually define brand attachment as “the strength of the bond connecting the brand with the self” (p. 2). This bond materializes when it is supported by a rich and accessible network of positive thoughts and feelings about the brand in the consumer’s memory.  A brand-self connection ascribes to the extent to which a consumer identifies with a brand as if they could merge together. In other words, the self (concept) of the consumer is extended so as to absorb the brand and make it part of his or her own self (image or goals). While the representation is cognitive, the researchers note, the brand-self linkage is inherently emotional. Brand prominence indicates in addition the ease and frequency with which  thoughts and feelings (underlying the connection) are brought to the consumer’s mind. The brand-self connection can “come to life” more readily when brand prominence is greater, hence the consumer experiences a stronger brand attachment.

  • The researchers first constructed a scale with five items for each of the two components. However, they sought to make the scale more parsimonious and practical to implement, and proposed a reduced scale of two items for brand-self connection and two items for brand prominence. Looking at the factor loadings suggests that it would be justified to keep four items for the first component and three items for the second. But in the researchers’  judgement parsimony should win over. For example, the item “feel emotionally bonded” could be discarded in favour of “feel personally connected”.

In their analyses, Park and MacInnis and their colleagues confirm that brand attachment and brand attitude strength are related yet empirically distinct constructs — while correlation between them is moderate-high they cannot be confounded. This supports the convergent and discriminant validity of brand attachment. The authors provide further support for the validity of attachment by showing an interesting relation to separation distress, a negative emotional state that may occur when losing a relationship with an entity people felt close to (e.g., feelings of depression, anxiety and loss of self). Brand-self connection and prominence each independently “contribute to the prediction of separation distress as indicators of brand attachment” (p. 8). The research additionally substantiates that brand attachment is distinct from attitude strength, the former being more strongly associated with separation distress.

Eventually, marketers would want to know how brand attachment is linked to behaviour. Three categories of difficulty are distinguished: (1) Among the most difficult forms of behaviour are buying always the new model of brand X, waiting to buy brand X versus an alternative brand, and spending money, time and energy to promote brand X (e.g., in pages and forums of social media and in blogs). (2) Moderately difficult forms of behaviour include paying a price premium for brand X and defending it when others speak bad of it. (3) The least difficult modes of behaviour include, for example, recommending brand X to others and buying the brand for others. Notably, recommending a brand to relatives or friends involves a certain personal risk for the endorser because one puts his or her own reputation or credibility on-line by suggesting to others to buy and use the particular brand. Yet, this alone is not considered hereby as a major cause of difficulty vis-á-vis the investment of time, money or energy to promote the brand (e.g., tell a story in a blog post, add photos).

With respect to intention to behave in ways that favour a brand (reflecting brand commitment) it is found that brand attachment predicts the intention to engage in behaviours regarded as the most difficult remarkably better than brand attitude strength. Brand attachment also better predicts intention to behave in moderately difficult ways but the difference from attitude strength, although also statistically significant, is rather small. There is no significant difference between attachment and attitude strength in predicting intention of performing the least difficult behaviours — they do equally well.  These findings bolster the importance of addressing brand attachment as a driver of brand commitment, particularly via more demanding modes of behaviour.

  • An additional test suggests that brand prominence is less essential than the brand-self connection component in predicting intentions. (Intentions were tested with respect to Nike.)
  • In a different set of analyses of actual behaviour (banking-investments), the researchers found furthermore that brand attachment is a better predictor of past purchases than brand attitude strength. In this case, however, brand attachment represented by both brand-self connection and brand prominence is predicting behaviour better than the former alone. That is, with regard to actual behaviour, brand prominence is an essential component.

Many brand owners would find utility in applying this scale of brand attachment (in a full or reduced form): from food (e.g., Nestlé) or toys (e.g., Lego) to banking (e.g., Royal Bank of Scotland) or carmakers (e.g., Peugeot). Take for instance Microsoft that now holds four brand names they may apply for marketing mobile devices: their own corporate name, Surface, Nokia or Lumia. Microsoft could use the aid of such a scale to decide which brand proves as better ground to build upon and which name is better eliminated. It may be a major factor in the contest of brand equity for mobile-smart devices of Microsoft versus Apple, Samsung Electronics, and Lenovo (Motorola).

Although the brand strength construct may capture a brand’s mind share of a consumer, attachment is uniquely positioned to capture both heart and mind share (p. 14).

The scale of brand attachment constructed by Park and MacInnis and their colleagues emphasises consumer identification with a brand, representing an emotional connection, and actualised through its prominence in memory. It does not cover other possible sources of attachment, but the approach taken is focused, concrete and well-substantiated. The researchers provide a valid scale for practitioners in brand management and research for measuring brand attachment, stand-alone or as part of a brand equity model.

Ron Ventura, Ph.D. (Marketing)

(*) Brand Attachment and Brand Attitude Strength: Conceptual and Empirical Differentiation of Two Critical Brand Equity Drivers; C. Whan Park, Deborah J. MacInnis, Joseph Priester, Andreas B. Eisingerich, & Dawn Iacobucci, 2010; Journal of Marketing, 74 (November), pp. 1-17.

Read Full Post »

Consumers often use price information as a cue to infer the quality of products — it is a familiar phenomenon based on the belief that price and quality are positively correlated. Consider for instance  laptop computers: consumers may rely on price to predict the quality of a laptop model for which there is lack of information about attributes that determine its quality, or rather because they have a difficulty to understand the technical features and try to infer the laptop’s expected quality based on its (list) price. Wine is another excellent example for a product whose quality consumers try to assess based on its price. The perceived price-quality relation is not always well-substantiated, which may lead to some costly mistakes. Reliance on price to judge quality is contingent on individual, contextual (e.g., product type) and situational factors.

Consumers may rely on price as an informational cue for different purposes: (a) to reduce the risk of buying a product of an unacceptable low quality; (b) avoid or mitigate effort of evaluating complex product information; (c) anticipate differences in quality between product brands and models (but sometimes also their symbolic meanings associated with prestige and luxury). Price-quality judgements involve two essential steps: estimating the strength of a relationship between price and quality in a focal product category, and applying this judgement to predict the quality of a particular product item (e.g., a new product model). Consumers may differ in their proficiency both to assess the relationship and applying it in various every-day situations.

The magnitude of price-quality correlations varies between product categories, and most consumers are aware of it. However, their calibration of the price-quality relationship for particular product types is often flawed and consumers over-estimate the correlations. Consumers tend to follow a general belief about price-quality relation without properly testing it as a hypothesis in the product category under consideration for purchase; alternately they bias their judgement by considering only evidence consistent with the prior belief (e.g., as the load of information to process is larger and harder to grapple with, and when information is organised in a format that highlights price-quality correlation [1]). Consumers also differ in the first place in their propensity to hold a price-quality belief (i.e., how strongly are consumers price-quality schematic). Capturing the actual reliance on price as a quality cue may also turn to be elusive because applying such a rule depends on the amount and nature of product information available.

In a research recently published (2013) Lalwani and Shavitt study how consumer propensity to perceive a price-quality relationship is governed or moderated by thinking styles and modes of self-construal exerted from consumers’ relations with others in their groups of membership. They distinguish between (1) independents (individualists) who prefer to form their opinions and set personal goals on their own, in hope those will be accepted by their in-group peers but not to be censored by the latter, and (2) interdependents (collectivists) who are inclined to form opinions and set goals that are subordinated to those of the in-group to which they belong. They refer to cultural self-construal by acknowledging that independence has been associated more closely with Western nations or Caucasian societies and interdependence with South and East Asian nations or societies. The distinction is primarily relevant to the construction of price-quality judgements by its correspondence with analytic vs. holistic styles of thinking, respectively. The authors additionally examine specific conditions that may enhance or inhibit the use of price to infer quality.

Analytic thinking orientates to process and evaluate a single piece of information at a time — for example, examine a value for a product item on a specific attribute. The ‘analytic’ consumer may compare between a few models on a specific attribute but ignore any other attributes. In a pictorial image, analytic thinking implies that the individual would look at each object in the image separately rather than inspecting a collection of elements in a scene. Holistic thinking, on the other hand, orientates to observe and evaluate relations between attributes and objects. It is much less focused on single items of information in favour of considering collections of them and how they relate to each other. In a pictorial image, holistic thinking means that an individual more easily identifies combinations of elements and conceives inter-relations between them in the whole scene. The argument put forward, and tested, by Lalwani and Shavitt posits that interdependents (collectivists) who are reliant on their social connections, and who are more considerate of the needs and goals of others in their in-groups before their own, are more predisposed to apply holistic thinking; independents (individualists) who tend to focus on their single-self’s needs and goals before others are more inclined to adopt an analytic style of thinking. Holistic thinking that endorses relational processing is clearly essential for making judgements about a price-quality relationship. The authors are particularly concerned with the boundary conditions under which the advantage of holistic thinking in making price-quality judgements has an impact.

Lalwani and Shavitt take notice that independent and interdependent modes of self-construal are not exclusive of each other, that is, they may be exhibited simultaneously in the same person or within a particular society. Therefore, following previous research, the authors apply two scales, one to measure independence and the other for interdependence as opposed to treating these modes as polar ends of the same continuum. They find that a stronger tendency to perceive a price-quality relationship (a global belief) is predicted by greater inclination for interdependent self-construal. No similar relation is found with independent self-construal. This confirms that only interdependent self-construal may support consumer tendency to rely on a price-quality relationship. [2]

Asians and Hispanic (in the US), representing interdependent self-construals, have been found to utilise price to infer the quality of a “new” target product item (alarm clock) whereas Caucasians (independents) showed no significant sensitivity to differences in price for the target product. It is emphasised that the Asians/Hispanics participants not just considered price-quality information available on “base” items but also practically used price in its evaluation of quality for the target item.

The difference in type of self-construal does not clarify sufficiently how this should lead to differences in approach to the perceived price-quality relationship. That is where the difference between holistic and analytic thinking takes its role. If we look only at the distinction between American nationals and Indian nationals, it would be relatively difficult to understand why the Indians have been found to exhibit a stronger tendency to rely on price as a quality cue. This difference is partially explained (mediated) once the researchers account for a difference in tendency to think holistically — the Indians also have a stronger tendency for that type of thinking that better supports processing of relations between price and quality.

Even more convincing are the results from a study in which an exercise with a pictorial image was conducted to encourage (prime) analytic versus holistic thinking by participants (American Asians/Hispanic vs. Caucasians). As expected, holistic thinking facilitated reliance on price when evaluating the quality of a “new” target product item (calculator) for both Asians/Hispanic and Caucasians. That is, they evaluated the higher priced target brand to be of higher quality than a lower priced brand. Nonetheless, the Asian/Hispanic who are more likely to be ‘interdependent’ differentiated even more strongly the quality between higher- and lower-priced target brands — revealing their advantage for relational processing. In contrast, when both Asians/Hispanic and Caucasians are primed to think analytically, none of them seems to use price as a quality cue. This highlights the power of holistic thinking for making price-quality judgements; vice versa, “imposing” analytic thinking on those who have a stronger tendency for holistic thinking seems to over-ride their advantage in predicting quality based on price.

Lalwani and Shavitt point-out that an advantage for relational processing in using price as a quality cue takes effect in kind of intermediate conditions: when there is a logical basis and supportive evidence (e.g., market conditions, product information available) for relying on price to infer quality, yet neither when conditions are poor/prohibitive nor when evidence of a price-quality relationship is just obvious and applying it is fairly easy. This is demonstrated in two cases: (a) an advantage for relational processing with regard to non-symbolic, functional or practical products (e.g., paper towels) vs. symbolic products that are better able to express one’s identity (e.g., watches, bicycle) — the latter product type induces a price-quality tendency in both ‘independents’ and ‘interdependents’; (b) an advantage for relational processing when information is provided on (non-price) attributes of moderate bandwidth (e.g., quality, durability, reliability), not for broad, generalised evaluations/attitudes (everybody uses price) and not narrow, specific features (nobody uses price). When conditions are sufficient but not too permissive, only those who have the advantage will discriminate products on perceived quality according to price.

The distinction between independent and interdependent self-contrual is somewhat circumstantial with respect to the utilisation of price as a quality cue. It does not immediately make sense why the two behavioural phenomena should be related. References to national and ethnic origins may also be too liberal generalisations that do not contribute enough to our understanding except for exposing the relationship. At the bottom of a distinction between modes of self-construal regarding price-quality judgement underlies the important distinction between holistic and analytic thinking. Lalwani and Shavitt effectively suggest that the extent to which people think in terms of relations between objects or their attributes corresponds with their attitude towards relations with other people, and hence the latter’s connection with the relationship between price and perceived quality. The distinction between thinking styles therefore seems to shed more light on conditions that induce or limit reliance on price as a quality cue.

Yet, establishing a connection between self-construal. particularly represented by national or ethnic (sociocultural) origins, and reliance on price as a quality cue, can be most productive and helpful for segmentation — it facilitates the identification of and access to relevant segments for marketing initiatives associated with the price-perceived quality relationship. The implications may be in devising advertising messages or premium product offering that target consumers with expected greater tendency to make price-quality inferences.  Consequently those consumers would likely be more favourable towards and receptive of higher-priced products/brands. This research further contributes to previous knowledge in the field by suggesting conditions under which most consumers or only selective segments would be evoked to make price-quality judgements. Marketers may consider the breadth of attributes described (broader dimensions vs. features) in addition to the structure of information presented to consumers [e.g., rank-order products by quality vs. random order, [3]).

Source:

You Get What You Pay For? Self-Construal Influences Price-Quality Judgements; Ashok K. Lalwani and Sharon Shavitt, 2013; Journal of Consumer Research, 40 (August), pp. 255-267, DOI:
10.1086/670034

Notes:

[1] A Selective Hypothesis Testing Perspective on Price-Quality Inference and Inference-Based Choice; Maria L. Cronley, Steven S. Posavac, Tracy Meyer, Frank R. Kardes, & James J. Kellaris, 2005; Journal of Consumer Psychology, 15 (2), pp. 159-169

[2]  Statistical Note: The validity of the results of multiple regression analysis performed is contingent on the two scales of individualism-independence and collectivism-interdependence not being negatively correlated. Such evidence is not reported. Turning to the source (Oyserman, 1993) reveals, as logically expected, that some of the statements are in contradiction between the pair of scales. In this case, the version of scales adopted by the authors suggests less conflict and the correlation between them is near zero. On the one hand, it is a little surprising that not even a low negative correlation was found to indicate the contrast between these constructs. On the other hand, a strong negative correlation between the scales could mean that only the stronger predictor, ‘interdependence’, won over the other confounded predictor and thus came out as the single significant predictor.

[3] Ibid. 1.

Read Full Post »

Older Posts »