Variety

Consumers rank variety of assortment straight after location and price when naming reasons why they shop at their favourite stores. Consumers care about variety because they are more likely to find what they want when going to a store that offers more varied assortments (Hoch et al., 1999; Helander and Khalid, 2000). Even more important than variety is perceived variety. Optimal consumer assistance during the interaction process considerably decreases the perceived variety, while the actual variety can be very high (Blecker et al., 2006).

Variety is a term that sometimes is confused with customization. Customization and variety are strongly related, but are not the same. Variety provides choice for consumers, but not the ability to specify the product as is the case with customization. When variety is high, it can be a substitute or enabler for customization, but customization and variety are distinct (Duray et al., 2000; Pine et al., 1993). Variety is the tool for mass customization, because it can create the link between the consumer and the product (Svensson and Jensen, 2001).

Digital products were described in paragraph 2.1. The characteristics of digital products, such as versioning and transmutability, can result in an increase in product variations. The fact that digital products take no stocking space, place no limits on a variety maximum. This paragraph addresses variety, first by addressing the value of variety from the perspective of the consumer. Variety also leads to complexity, which is addressed in section 2.3.2.

The value of choice

Variety simply involves more choices from which the consumer is able to choose. Not always does a consumer know with certainty whether he or she will obtain what he or she wants when looking for a product, but the greater the number of items carried by the place the consumer is looking for the product, the greater the expectation that the consumer finds what he needs (Baumol and Ide, 1956; Desmeules, 2002).

A mass customizer should be able to offer a wide range of useful external variety, while maintaining low internal variety (Blecker et al., 2006). External variety refers to the product variations that are perceived by the consumers. The perspective which is chosen for this thesis focuses on the consumer, thus the external variety. Since mass customization aims at fulfilling individual consumer needs, it involves an extensive external variety.

Consumers do not want more choices, they want exactly what they want – when, where and how they want it (Pine et al., 1995). Important drivers that lead to more variety and more mass customization are globalization and market turbulence (Svensson and Jensen, 2001; Santonen, 2003). Markets are becoming increasingly global, and competition from low costs manufacturing countries is getting larger. Products are no longer targeted at one geographic market only, but moreover towards a global market. Product variations therefore must be made possible in order to adapt to different needs and tastes. Through the use of modularization in mass customization, the risk in product development can be reduced as a large number of variants can be launched (Pine, 1993).

Other drivers for more variety are demand, deregulation and the experience economy. Consumers demand and get more variety and options in all kinds of products, which lead to more diversity and niche markets or “long tail” markets (Anderson, 2006). Deregulation has also increased the number of choices; more competition has lead to more variety (Kezunovic et al., 1998). Another driver is the so called experience economy (Pine and Gilmore, 1998). The social experience a consumer gets when buying products leads to more variety, because every experience is experienced in a unique way.

When firms adopt high variety strategies, they fulfil two distinct goals (Kahn, 1998 in Desmeules, 2002). First, more variety makes it more likely for consumers to find exactly the option they were looking for, such as customized products. Second, it allows consumers to enjoy variety over time, so they can discover new products. These strategies enable long-term relationships with consumers and learning consumers’ preferences over time (Desmeules, 2002). The process of learning preferences can be influenced by the way product information is presented (Huffman and Kahn, 1998).

Information products, or digital products, are experience goods (Nelson, 1970; Shapiro and Varian, 1999). This means that consumers do not know what it is worth to them until they experience it. There is a relationship between variety and the positiveness of a consumption experience, when the evaluative task is performed by cognition (Desmeules, 2002). This relationship is shown in Figure 2.4.

There are three sections to the curve on the graph. The first, section 1, is an upward sloping portion, followed by a relatively flat line, section 2. Consumer satisfaction increases with increased variety. Adding more options increases the positiveness of the experience because the chances that the consumers find what they want increase. There is a point where the consumer experience is satisfied, called point of satisfaction. This point can be reached with a single option to choose from, or in other cases with multiple options to choose from. As satisfaction is reached, at section 2 in the figure, adding more variety does not influence the positiveness of the consumer experience much. According to Desmeules (2002), this is the optimal amount of perceived variety.

To measure variety, I follow Blecker et al. (2006). Blecker et al. (2006) introduce a key metrics system to control variety induced complexity in mass customization, which is based on three assumptions. The key metrics system is first based on the assumption that the product family is built around platforms and modules. The second assumption is that the interaction process is carried out over the Internet, so that the consumer is involved in the production process. Both assumptions are in line with the mass customization classification of Duray et al. (2000). Third, they assume that the mass customizer has not implemented a cost calculation system that enables one to accurately evaluate the costs triggered by variety induced complexity.

Blecker et al. (2006) revealed that multiple use, interface complexity and platform efficiency are key metrics that directly influence the extent of product variations that can be offered by the mass customizer. In mass customization, products have to be developed around common parts without restricting the range of end product variants required by consumers. Commonality of components or modules can considerably reduce overhead, because it reduces internal variety and thereby variety induced complexity (Blecker et al., 2006). However, the fulfilment of individual needs of the consumer, which requires external variety, is the most important objective of mass customization. To achieve this objective, the supplier has to strive for individualizing products while using only a few modules. The multiple use metric provides a measurement of the number of product variants required by consumers as compared to the total number of modules. The higher the value of this metric, the better since it indicates a higher flexibility of the product assortment (Blecker et al., 2006). The second metric that directly influences the possible variety, is the complexity of the interface. In order to generate a wide range of product variants by only mixing and matching a few modules, an optimal design of module interfaces is necessary. Interface complexity reduces the extent of product variations, and should in turn be reduced by standardization of interfaces (Blecker et al, 2006). The third factor that directly influences the possible variety is the product platform. A platform is the core module of a product family. Because of this, several product variants based upon a particular platform will be introduced or eliminated during the product lifecycle (Blecker et al., 2006). The complexity of deriving new products on the basis of one platform should be kept as low as possible. Blecker et al. (2006) introduce the average platform cycle time efficiency as a key metric that directly influences the possible variety, which measures the average elapsed time to develop a derivative product in comparison to the elapsed time for the development of a product platform.

Variety and complexity

One limit of mass customization often quoted is that excess variety may result in an external complexity. The burden of choice can lead to an information overload, resulting from a limited capacity of humans to process information (Choo, 1998). As a result, the configuration process may last quite a while, and consumers may experience an increase in uncertainty during the transaction (Piller et al., 2005).

Too much variety can lead to complexity. As shown in Figure 2.4, there is a point where the positiveness of the consumption experience decreases when variety increases (Desmeules, 2002). This point on the inverted U-shaped figure is called the point of regret. This is the point in the amount of variety where variety alone brings about doubt and regret avoidance mechanisms (Desmeules, 2002; Schwartz, 2000). After the point of regret, the positiveness of the consumption experience goes down because of stress and frustration caused by heightened expectations and the inability to make a choice.

There are possibilities to move the point of regret to the right. This does not indicate what the optimal amount of variety should be, but there are techniques to increase the amount of variety before the point of regret is reached. Presenting information by attributes rather than by alternative increases consumer satisfaction, and also enables the consumer to be more ready to make a choice (Huffman and Kahn, 1998). As variety is high, consumers should be explicitly asked about their preferences rather then just listing all available options (Huffman and Kahn, 1998). Variety alone is not enough when the amount of variety exceeds a point where the positiveness of the consumption experience decreases. Since the positiveness of the consumer experience depends on the perceived variety, it remains possible for the consumer to experience the consumption in a positive way. This consumption experience can be influenced by how the information about the product class is presented (Huffman and Kahn, 1998). When variety is low, the positiveness of the consumers’ experience can be maximized just by offering the products by alternative. Presentation by alternative is preferred when assortments are small. However, in complex situations, the consumer can benefit from learning his or preferences first (Huffman and Kahn, 1998; Louviere et al., 1999, Kurniawan, 2006). Huffman and Kahn (1998) also found that satisfaction levels for consumers can be higher when they are asked to explicitly state their within-attribute preferences. In order for consumers to process a complex, high variety choice set, the consumption experience should not be frustrating.

Kurniawan et al. (2006) conducted a study on differences between product configuration and product selection, in terms of consumers’ decision quality. Their results show that product configuration offers consumers greater satisfaction during the process and in the products of their choices than product selection does. These results are comparable with the results of Huffman and Kahn (1998). One aspect on decision quality with product selection is product accuracy. Product accuracy is the closeness between preferred product and the product that is actually bought or selected.

Dellaert and Stremersch (2004) addressed consumer preferences for mass customization. They mention ways in how mass customization configurations may differ, which have influence on the perceived complexity of mass customization configurations. One of them is about the company presenting a default version which consumers may then customize, or the firm may not show a default version and let consumers start from scratch in composing the product. Their results show that firms should offer a default version that consumers can use as a starting point for mass customization, which minimizes the complexity for consumers. This default version should not be too advanced, because consumers are more willing to switch up than they are willing to switch down.

Consumer expertise also influences the perceived complexity (Dellaert and Stremersch, 2004; Franke and Piller, 2003; Huffman and Kahn, 1998; Piller et al., 2005). Consumers with high product expertise experience lower complexity when participating in mass customization than consumers with low product expertise. They are also more able to analyze information and to select that information which is most important and task relevant.

When consumers are involved early in the mass customization process, the consumer interacts with the mass customizer. The interaction process, or interaction sub-process (Blecker et al., 2006), can be supported by interaction systems to guide the consumer through the interaction process. It is difficult or even impossible to know what consumers want at the beginning of the interaction process (Franke and Piller, 2003). Therefore it is strongly recommended to implement immediate feedback tools for mass customization toolkits (Von Hippel, 2001; Franke and Piller, 2003). The main problem with these interaction systems is that configuring a product can become quite difficult, frustrating, and time consuming for the customer (Stegmann et al., 2006). To overcome these difficulties, Stegmann et al. (2006) propose two methods of consumer support. The first is automatically generated product and component recommendations, which depends directly on the quality of available customer information (Balabanovic and Shoham, 1997; Resnick and Varian, 1997). The second is direct recommendations among consumers.

Offering customized products requires an individual, or one-to-one, relationship between the consumer and the supplier. Besides the external complexity and the lack of consumer knowledge, Piller et al. (2005) identified another problem category. This third limit of mass customization is the information gap regarding the behaviour of the supplier. For many consumers, customizing a product is still an unfamiliar process, and can be compared with the principal-agent problem. Asymmetric information distribution as well as deviations between the goals of the principal, or the consumer, and the agent, or the supplier, which cause the latter to behave in an opportunistic manner (Wigand et al., 1997). This behaviour incurs costs for the consumer. For digital products this is a serious problem, because these products are experience goods.

The above uncertainties can be dealt with by collaborative consumer co-design in online communities (Piller et al., 2005). Communities for consumer co-design offer three contributions to reduce mass confusion problems (Piller et al., 2005). The first is the generation of consumer knowledge. Consumer knowledge reduces the burden of choice and transfers personal needs into product specification, as is visualized in Figure 2.5. Offering users a starting solution based on the profile of the consumer increases the personalization of the co-design process, and increases the customization possibilities. Lacking a proper consumer profile leads to ineffectiveness of the co-design process (de Vries, 2003). The second is the provision of support for interactive, collaborative filtering where consumers directly interact on the co-design platform. Consumers may mutually support each other in finding a solution which best fits their needs (Piller et al., 2005). This second contribution reduces mass confusion in the same way as the first. The third contribution is trust building and the reduction of risk. It reduces mass confusion by transferring personal needs into product specification and reduces uncertainty about the behaviour of the provider.

Blecker et al. (2006) identified key metrics that can be used during the interaction process to measure the average interaction length of time, and the abortion rate, which could give some answers on the questions raised by Franke and Piller (2003). The interaction system is the primary instrument for reducing the consumer’s costs arising from a principal-agent constellation that is inevitable in mass customization (Franke and Piller, 2003). Because of extensive product assortments in mass customization, consumers may experience high complexity during the web-based configuration process. A high range of product variety is relevant for the fulfilment of different consumer needs, but the experienced complexity that comes with it is something that has to be reduced. To determine the optimal level of variety from the consumers’ perspective, the key metric ‘used variety’ was introduced (Piller, 2002, in Blecker et al., 2006), see Figure 2.6. The used variety metric compares the variety that is actually perceived by consumers to the theoretically possible product variants. Low values of this metric indicate that a large number of product variants are unperceived or uninteresting for consumers. To measure the perceived complexity due to variety, Blecker et al. (2006) proposes two key metrics. The first is the average interaction length of time, which measures how much time consumers need on average to completely configure a product variant. The second key metric is the abortion rate. If consumers are uncertain about their choices or are overwhelmed by the interaction process, it is more likely that the consumer will abort from the interaction process.

This section addressed some methods to overcome variety induced complexity from the consumers’ perspective. These methods should be used in a way to move the point of regret of the inverted U-curve (Desmeules, 2002) to the right.