G i v i n g d i r e c t i o n s t o r e a c h t h e e n d o f t h e p u r c h a s i n g f u n n e l
Online purchase intention for the e-tailer
b y
Roald Reurink
University of Groningen Faculty Economics and Business
Supervisor: Liane Voerman, Vakgroep Marketing External supervisor: Gertin Schraa, Philips Online Shop
Management summary
The internet as sales channel is booming, the e-tailer has expanded immensely in years and has a profound influence on the shopping process for many consumers (Brown et al., 2003). Products can be bought, sought and evaluated online. This unique ability of the internet has transformed the social and spatial aspects of shopping for many consumers (Brown et al., 2003). This research investigates and analyses the antecedents of online purchase intention in the online shopping process. Once a customer arrives on an e-tailer, the sales process mirrors that of a funnel. At each step the online shop can lose a customer, which are called the bounce and exit rates. So the entire sales process, from traffic (website visitors) to conversion rate (actual sales), involves a series of smaller conversions.
Based on prior research several antecedents have been found which are grouped into consumer characteristics (perceived risk, perceived trust, customer involvement, online shopping experience, online search experience, brand familiarity and awareness) and site characteristics (site interactivity, ease of use and web design). An online questionnaire has been designed to collect data on these antecedents and ask the respondents questions during an online shopping process on the Philips Online Shop. The antecedents are put in two regression analyses (bivariate and multiple) to see which antecedents have a significant influence on online purchase intention. For the bivariate regression analyses only customer involvement and awareness were not significant. For the multiple regression analyses there were three significant antecedents; perceived trust, perceiving the shopping experience as if a customer was buying in the physical store and brand familiarity.
Preface
I’m doing this research to understand the drivers that affect the sales performance and indirectly sales growth. My interest has been with e-commerce for some time now; at the age of 14 I started in a community called SimRacing. After taking over an online championship in one of the raceseries being on the Internet became a hobby. Since then the community passed over 150.00 members, it became more realistic and professional. I couldn’t imagine that an online gamer would be winning BMW car as prize, or earning a raceseat in NASCAR. After starting my Master Marketing at the University of Groningen I was happy that I found an internship where I could combine my marketing and e-commerce interest. The Philips Online Shop gave me the opportunity to conduct an empirical research to find the drivers of online purchase intention, comparing the outcome and process measures from their online shop. Combining an internship with writing a master thesis isn’t easy, but surprisingly I enjoyed it.
So, my student time has finished. Let’s start a new chapter with new challenges. I’m off to San Francisco now to do an internship at SimRaceWay.com, combining my passion for online racing and marketing. After that, travelling through California and eventually, go home.
I want to thank my supervisor Liane Voerman, mostly for helping me with the marketing research subjects such as the regression analyses. I want to thank my parents for their support during my studies, as without I would never had such an amazing student time! As last I want to thank my supervisor at Philips, Gertin Schraa, and the MARUG for giving the opportunity to have my best student year where I made best friends, had great parties, unforgettable memories, and giving me the interest in Marketing.
Table of content
Management summary ... 3
Preface ... 4
Table of content ... 5
Chapter 1: Introducing the online shopping world ... 7
1.1 Online purchase intention ... 7
1.1.2 Pros and cons for an e-‐tailer ... 9
1.1.3 Online shopping experience ... 10
1.2 Problem statement and research questions ... 11
1.3 Research questions ... 11
1.4 Case study Philips Consumer Lifestyle Online FlaghShip Shop (FSS) ... 11
1.5 Relevance ... 12
1.6 Structure of research ... 13
Chapter 2: The influences on online purchase intention ... 14
2.1 Consumer characteristics as antecedents of online purchase intention ... 14
2.1.1 Online search behavior ... 14
2.1.2 Perceived trust ... 17
2.1.3 Perceived risk ... 19
2.1.4 Online Shopping Experience ... 20
2.1.5 Online customer involvement ... 21
2.1.6 E-‐tailer awareness ... 23
2.1.7 Brand familiarity ... 25
2.2 Website characteristics as antecedents of online purchase intention ... 26
2.3 Conceptual Model ... 28
Chapter 3: Research design ... 30
3.1 Type of research ... 30
3.2 Operationalization of the antecedents ... 30
3.3 Data collection ... 34
3.4 Population and sample ... 35
3.5 Plan of analysis ... 36
3.6 Development of the research ... 38
Chapter 4: Results ... 39
4.1 General findings ... 39
4.2 Key results ... 41
5.1 Conclusions consumer characteristics ... 50
5.2 Conclusions site characteristics ... 53
5.2 Managerial implications ... 54
5.3 Limitations data and future research ... 57
References ... 59
Chapter 1: Introducing the online shopping world
“The Web as I envisaged it, we have not seen it yet. The future is still so much bigger than the
past.” Tim Berners-Lee1
The quote above from Tim Berners-Lee (founder of the Internet) shows that the Internet still continues to improve and this is similar for e-tailers, which undergo improvements to sell their products to the consumer. An e-tailer is a site in which customers can browse, evaluate, order, and purchase a product (Yoo and Donthu, 2001). But what drives these consumers to buy the products online? The research objective of this research is therefore to explore the antecedents that influence online purchase intention in consumer markets.
1.1 Online purchase intention
According to a survey by A.C. Nielsen (2007), over 627 million people have done online shopping. A research by Forrester (2006) reported that the e-commerce market would grow to $288 billion dollar in 2009, compared to the $228 billion in 2007. As reported by Jupiter Media Metrix (2005), the sales of e-tailers was about $65 billion in 2004, and it would likely grow to $117 billion in 2008 (Delafrooz, 2009). Delafrooz (2009) state, that by 2010, e-commerce will account for $316 billion in sales, or 13 percent of the overall retail sales. A total of 61 percent of the online users in the US will make their purchases through the Internet in 2010, as compared to merely 46 percent in 2004 (Delafrooz, 2009). Furthermore Delafrooz (2009) state in Neslin and Shankar (2009), that by 2011, 47% of all transactions are expected to be Internet-enabled. So, the internet as sales channel is booming, the e-tailer has expanded immensely in years and has a profound influence on the shopping process for many consumers (Brown et al., 2003).
That is why an increased understanding in the influences in online purchase intention can benefit in an increase of sales performance. Products can now be bought, sought and evaluated online. This unique ability of the internet has transformed the social and spatial aspects of shopping for many consumers (Brown et al., 2003). So, the online purchase intention is an important determinant of online shopping behavior and represents the best estimate of future
behavior available to market researchers (Delafrooz, 2009). Purchase intention is a part of the choice behavior of the consumer and is defined by Cobb-Walgren et al. (1995) as the intention to purchase a particular brand. Measures for purchase intention have been used regularly to identify buying likelihoods for products within defined time periods (Brown et al., 2003).
The definition of purchase intention used in this research is: “Purchase intention is the
purchase likelihoods for products within defined time periods” (Brown et al., 2003).
This research will focus on the online purchase intention for the e-tailer; once a customer arrives on a retail website, the sales process mirrors that of a funnel. At each step the online shop can lose a customer, which are called the bounce and exit rates. So the entire sales process is from traffic (website visitors) to conversion rate (actual purchase), which is defined as “the proportion of visitors who complete a desired action” (Ayanso and Yoogalingam, 2009). This sales process involves a series of smaller conversions (see figure 1) and at any point, the process may not support the customer’s needs and those customers will leave the online shop (Eisenberg and Novo, 2002).
Figure 1: Sales process
attract new customers as well as to retain existing customers is a major challenge. The paradox in online purchasing is that getting much traffic to an online shop can be meaningless and even expensive if customers do not take action or complete a purchase transaction (Ayanso and Yoogalingam, 2009). In other words, an online shop with fewer visitors but a high conversion rate can be much more profitable than an online shop with a lot of traffic and low conversion (Ayanso and Yoogalingam, 2009).
Concluding, it has been demonstrated that measures of online purchase intention do possess predictive usefulness regarding actual purchase behaviour, so that the higher the purchase intention is and thus the more customers will end up in the order process can lead to more purchases (Novak and Hoffman, 1999; Brown et al., 2003).
1.1.2 Pros and cons for an e-tailer
There are several opportunities for an e-tailer. First, whereas data captured from purchases in traditional stores only collect information concerning the buying behavior of their clients, online data provides much more information (Moe and Fader, 2002). This will result in products,
services and even marketing actions can be adjusted to the profile of visitors in order to influence (potential) customers’ visiting and shopping behavior (Poel et al., 2005). Visits that do not result in a purchase of one or more products are monitored that makes the customer picture, which firms are attempting to compose, more complete (Poel et al., 2005). Secondly, e-tailers enable consumers to access a greater amount of detailed information with regard to product attributes, availability, and overall value proposition and a wide selection of assortment (Brown et al., 2003; Yoon, 2002). For example, the number of available books at Amazon.com is more than 23 times larger than the number of books in a typical Barnes & Noble superstore, and even 57 times greater than the number of books in a typical large independent bookstore (see table 1,
Table 1: product variety comparison for Internet and Brick-and-Mortar channels (Brynjolfsson et al., 2003).
However, there are also cons for an etailer. First, the more products e-tailers make available, the harder it is for the consumers to locate the product they are interested in (Brynjolfsson et al., 2006). In fact, consumers can become overwhelmed when choices are poorly organized and actually reduce their purchases as a result (Brynjolfsson et al., 2006). An online shop can be poorly designed with little attention paid to functionalities that facilitate consumer decision-making (Ayanso and Yoogalingam, 2009). Secondly, shopping online is easier online because of the ease of access and comparison websites (Monsuwé et al., 2004). Price comparison sites such as Prijsvergelijk.nl and Kelkoo.nl enable consumers to quickly compare all the online prices. As consumers can compare these prices easily, they are likely to buy at the e-tailer with the lowest price. The low search costs and effort involved makes it easy to just visit online shop without a purchase intention (Ayanso and Yoogalingam, 2009). That is why consumers of online shops are rarely loyal to a specific website when searching for a particular product or category (Johnson et al., 2000).
Concluding, being active as an e-tailer does not necessarily implicate a bed of all roses. An e-tailer should be aware of the opportunities of the Internet, but should also be aware of the cons that reduces actual purchase.
1.1.3 Online shopping experience
Concluding, Verhoef et al. (2007) sees online shopping as enjoyment. Childers et al. (2001) found enjoyment to be a consistent and strong predictor of attitude toward online shopping. According to Monsuwé et al. (2004) online shopping seeks for the potential entertainment resulting from the fun and play arising from the online shopping experience. Consumers appreciate the online shopping experience for its own sake, apart from any other consequence like, for example, an online purchase that may result. The purchase of goods may be incidental to the experience of online shopping.
1.2 Problem statement and research questions
Time and money is spent by e-tailers to attract consumers to their online shop, which are optimized to deliver that positive shopping experience. E-tailers invest significant effort in managing functionalities that can attract and convert visitors (Ayanso and Yoogalingam, 2009). However, even when consumers visit the online shop with a purchase intention, many of them do not complete the transaction and abandon their intention just prior to the check-out (Cho, 2003). Research indicates that even 81 percent of those who browse web sites for goods and services do not actually make an online shopping (Lassar et al., 2005).
Since online shopping is getting matured and more accepted nowadays, it will be interesting to find the antecedents and to analyze how these antecedents influence online purchase intention. The academically contribution of this research is that it gives new knowledge about online purchase intention by looking at e-tailers. So, the purpose of this research is to investigate the antecedents of online purchase intention at e-tailers.
For this research the following problem statement is formulated:
Which antecedents influence online purchase intention an e-tailer?
1.3 Research questions
To answer the problem statement, the following research questions are formulated: 1. What kind of consumer characteristics influence online purchase intention? 2. What kind of website characteristics influence online purchase intention?
This section will give a short introduction of the Philips FSS, where the empirical data is collected from to show the results. The Philips Customer Lifestyle Online FlagShip Shop is an initiative that has been started at the end of 2007 and should lay the foundation for a worldwide direct sales platform. The front-end application (user-interface) of the Flagship Shops selected to support these ambitious sales targets is based on a standard e-commerce solution provided by third party Digital River, while look and feel of the shop was designed by Philips. Digital River is also contracted for the order creation process, including master data validation, fraud check and payment processing. All orders are fulfilled out of the central warehouse in Acht, Netherlands, which is managed by DHL and used as the stocking point for all products e.g. Consumer Lifestyle and Lighting products. Customer service is outsourced to third party Sitel, who has a dedicated team available in the call center in Eindhoven covering entire Western Europe.
The FSS team is responsible for marketing, planning and forecasting, overall activity and performance monitoring and related financial activities. The FSS is strongly marketing driven, but they are more and more taking a sales driven focus. However, it is not yet common for Philips to focus on the online market as direct sales channel.
Philips has a threat of intensive price competition, as it is unable to drop its prices because of possible conflicts with their offline retailers. The recommended retail prices are set by the business group of Philips for each country. The Netherlands and UK are similar; however each country set the real prices, for example to get rid of old stock. So it depends on the product life cycle, first the prices are set and hold for a certain time. A small dilemma for Philips is that it has to balance between prices for the online and offline retailer. The FSS is the only channel that directly sells to the consumer (B2C); all other offline channels are B2B.
1.5 Relevance
As online shopping gradually moves from a novelty to a routine way of shopping, it gets more matured and accepted. Here, the quality of the Internet sites (e.g. e-tailers) will play an important role in differentiating sites (Yoo and Donthu, 2001). According to van der Poel et al. (2005) predicting and understanding onlinbuying behavior is of utmost importance for e-tailers. A large number of consumers in the US and Europe frequently use the Internet for shopping purposes, however, it is not clear what drives them to shop online and whether these numbers could be even increased if more attractive online stores were developed. This raises the issue of examining what antecedents affect consumers to shop online (Monsuwé et al., 2004). This research aims to enhance our understanding of the antecedents which influence online purchase intention by analyzing the consumers online sales process. Analyzing these antecedents makes it possible to investigate the tradeoffs consumers make when selecting an e-tailer and the choices consumers have; when does a customer abort an online purchase and at which point in the sales funnel are customers leaving? Moe and Fader (2004) suggest that customers have different reasons for visiting a retail site and highlight the importance of understanding and accounting for various patterns in the relationship between visiting and purchasing.
With an online questionnaire the relative strength of the antecedents and relative importance of the antecedents of online purchase intention can be measured. This will give valuable information which can be used by e-tailers to effectively think about the importance of different antecedents which influence online purchase intention for their customers. This research leaves out auction sites such as eBay and Marktplaats.nl and solely focuses on purchase from e-tailers.
Concluding, what makes the consumer purchase online? This research identifies the antecedents of the buying process at an e-tailer and analyses their relationship with online purchase intention. The outcome of this research will provide antecedents that give e-tailers the knowledge to improve their actual purchase.
1.6 Structure of research
and discussion of the empirical data. Next, a short summary is given of the findings in this research. To conclude, a discussion, recommendations and further directions for research are presented.
Chapter 2: The influences on online purchase intention
In this chapter an exploratory literature is conducted regarding the online purchase intention which provides a background into the antecedents of online purchase intention. There are two main sections about the consumer side and website side of online purchase intention. These two main sections are further outlined using subsections that discuss each antecedent of online purchase intention. Each section ends with one or more hypotheses. Finally, hypotheses will be summarized into a conceptual framework which is the basis for the empirical research.
2.1 Consumer characteristics as antecedents of online purchase intention
This section will discuss the consumer antecedents of the online purchase intention. The following topics are discussed: the research shopper phenomenon, the attitudes towards online purchasing and the distinctions between online purchase behavior and the online shopping experience.
2.1.1 Online search behavior
Figure 2: distribution of conversion rates (Ayanso and Yoogalingam, 2009)
Despite some inconveniences (e.g. difficulty of assessing quality online, insecurity about
payments and postponed gratification), online shoppers generally indicate that shopping online is easier than offline due to the ease of access and comparison shopping (Children et al., 2001). Li et al. (1999) found that compared to offline shoppers, online shoppers were stronger motivated by search convenience.
Figure 3: Previous evidence of research shopping (Verhoef et al., 2007).
Verhoef et al. (2007) state that the vulnerability of the internet is because of channel lock-in is lock-insignificant for this channel, and there is a marglock-inally significant cross-channel synergy between online search and store purchase. Jayawardhen et al. (2007) indicate that consumer purchase orientations in both the traditional world and on the Internet are largely similar. Therefore, both academics and businesses are advised to treat the Internet as an extension to existing traditional activities brought about by advances in technology, i.e. the multi-channel approach (Jayawardhen et al., 2007). Figure 4 shows that two attributes that particularly hurt the Internet as a purchase channel are its lower average scores on service and privacy when compared to the store.
Figure 4: decreasing research shopping by improving Internet purchase attributes (Verhoef et al., 2007)
decreases by about seven percentage points. Another option for decreasing Internet to store research shopping is to increase Internet lock-in; this could be done by having the website remember a customer's previous orders, delivery addresses, and credit numbers, as is done by Amazon.com. The last option for decreasing Internet to store research shopping is to decrease cross-channel synergy. According to Shim et al. (2003) consumers’ intentions to use the same channel for both purchase and search reflect efforts to reduce costs (e.g., time, effort) for the entire search-purchase transaction. In some cases, consumers’ initial search effort may not be realized because unexpected events arise that increase the total transaction cost. According to Seiders et al. (2000) e-tailers that are convenient: are easy to reach (access convenience); enable consumers to speedily identify and select/order the desired products (search convenience); make it easy to obtain the desired products (possession convenience); and expedite the purchase and return of products (transaction convenience).
Concluding, Shim et al. (2001) state that consumers’ online search experiences at e-tailers are integral determinants of their online purchasing behaviors. The following hypothesis is formulated:
H1: online search experience is positively associated with online purchase intention
2.1.2 Perceived trust
relationship management and thus increase purchase intentions through increased consumer trust. If so, then adding into a website such features as virtual communities, toll-free numbers, responsive service through e-mail or Web-push technologies, or synchronous message boards where questions can be posted online to a sales-rep, may increase consumer trust and purchase activity (Gefen and Straub, 2003).
According to Kuan and Bock (2007) trust positively affects online purchase intention because the customer believes that the retailer’s online operations are able (because of
competence) and willing (due to benevolence and integrity) to deliver the products purchased. As such, if customers place their trust on the e-tailer, they rule out possible but unfavorable online actions of the retailer, leading to higher online purchase intention. According to Gefen and Straub (2003) consumers must trust the e-tailer not to engage in potential, but clearly
undesirable, opportunistic behaviors such as unfair pricing, violations of privacy, conveying inaccurate information, unauthorized tracking of transactions, and unauthorized use of credit card and purchase information. The consumer cannot be certain that the e-tailer will not pursue such unethical activities, and for this reason, trust and the building of trust is an essential element in e-commerce. Security and privacy policies performance and refund warranty, and quality of service are website characteristics that directly affect trust in the website, as they are signals of the firm’s capacity and good will (Martín and Camarero, 2008).
Concluding, it should be noted that trust in the company does not have to be a necessary condition to purchase online (Heijden et al., 2003). It has been argued that lack of trust in the organization can be offset by trust in the control system (Heijden et al., 2003). Such a control system would include the procedures and protocols that monitor and control the successful performance of a transaction, and could include the option to insure one against damage (Heijden et al., 2003). Someone may not trust the internet company, but we may trust the control system that monitors its performance (Heijden et al., 2003).
H2a: Perceived trust is positively associated with online purchase intention
2.1.3 Perceived risk
In general, the greater the risks consumers perceive, the more extensive is their information search prior to purchase (Dowling, 1986). Because online shopping is a new mode of shopping involving various and seemingly novel types of perceived risks the consumer is likely to place added importance on searching for information when using this channel (Shim et al., 2003).
As seen in the research shopper phenomenon by Verhoef et al. (2007) the Internet as a purchase channel is less effective because of lower privacy when compared to a physical store. So increasing privacy and improved service will reduces risk and increases trust. This will result on less research shoppers and a higher online purchase intention due to Internet channel lock-in, less free-riding and less cross-channel synergy (Verhoef et al. 2007). Other circumstances where perceived risk is higher is when there is little information available, the product is new, the product has a high price, the offering is technologically complex, the consumer has little experience in evaluating the product or opinions of others are important and the consumer is likely to be judged on the basis of the acquisition (Hoyer and MacInnis, 2009).
Concluding, the higher the perceived risk, the more trust is needed to make a purchase (Mayer et al., 1995). Also according to Hoyer and MacInnis (2009) consumers are motivated to search for information to reduce risk. Therefore, the following hypotheses are formulated:
H2b: Perceived risk in buying products online is negatively associated with online purchase intention H2c: Perceived trust is negatively associated with perceived risk
H2d: Online search experience is negatively associated with perceived risk
2.1.4 Online Shopping Experience
Novak (1996) research into consumer online behavior initially focused on the nature and consequence of staging compelling Internet experiences that were capable of inducing the optimal mental state known as "flow". They also found that when an Internet user’s personal skill and task-related difficulty is challenging, the stage named “flow”, it may have a bearing on his or her online purchase intention. According to Holzwarth et al. (2006) consumers report that online companies are impersonal, consumers feel helpless when shopping virtually in unfamiliar or complex product categories, and consumers want the customer assistance often found in a conventional shopping environment combined with the convenience of Internet shopping.
So, it can be concluded that online shopping experience should be considered. Improving the online shopping experience should improve the online purchase intention of potential buyers (Childers et al. 2001). Therefore the following hypothesis is formulated:
H3: Perceived online shopping experience is positively associated with online purchase intentions
2.1.5 Online customer involvement
As a new marketing channel, the Internet differs from traditional retail formats in many ways (Park et al., 2007). Consumers shopping online cannot touch or smell products, as would be possible in traditional retail outlets, so their purchase judgments must be based on the product information presented on the website. Online sellers seek to overcome this limitation by giving consumers the opportunity to share product evaluations online (Park et al., 2007). This consumer-created information is helpful in making purchase decisions because it provides indirect experiences of products. Information search has been shown to be sensitive to varying levels of involvement (Mathwick and Rigdon, 2004). Involvement is defined as “the perceived personal relevance of a product based on the individual consumer’s needs, interests, and values” (Park et al., 2007). As the personal relevance of a focal product or service increases, involvement tends to increase (Park et al., 2007).
can be done through online consumer reviews. Based on the definition of WOM by Westbrook (1987), electronic Word-of-Mouth (eWOM) can be defined as all informal communications directed at consumers through Internet-based technology related to the usage or characteristics of particular goods and services, or their sellers (Litvin et al., 2008). Customers who make a personal referral must not only believe that a company offers superior economic value but also feel good about their relationship with the company (Ramani and Kumar, 2008).
Consumers with higher product involvement generate increased WOM (Litvin et al., 2008). WOM refers to the spread of information about products, services, stores, companies, sales, or customer managers from one customer to another (Ramani and Kumar, 2008). According to Yoon (2002), recommendations or WOM have a positive effect (38%) on purchasing a product online. Sundaram and Webster (1999), who conducted a study on air conditioner purchase decisions by undergraduate students, demonstrated that the students’ evaluation of an unfamiliar brand was more susceptible to change from WOM than was their attitude toward a familiar brand (Litvin et al., 2008).
According to Martin (2009) users of social media visit social media sites in order to connect and share experiences (e-WOM) with friends and family over the web. Social media reaches consumers early in the purchase funnel, which is why most of the functionality provided by these sites does not lead directly to a commercial transaction. Rather, advertisers find these publishers valuable for their reach to a broader audience and ability to create WOM. As a result, social media sites tend to live high in their advertisers’ purchase funnels.
means approximate 55% will take those online comments into consideration while making purchase decisions.
According to Park et al. (2003) the quality of online consumer reviews has a positive effect on consumer purchasing intention. Reviews that are logical and persuasive, with sufficient reasons based on specific facts about the product, have a strong positive effect on purchasing intention. Also, consumers’ purchasing intention increases along with the number of reviews. The existence of many reviews indicates that the product is popular, and this is what increases purchasing intention. These two findings can be interpreted as the effects of the quality and quantity of online word-of-mouth messages.
Concluding, both online customer reviews and WOM affect the consumer in the online sales funnel. According to Martin (2009) WOM affects consumers early in the purchase funnel, where according to Huang et al. (2009) the presence of online consumer reviews enables consumers to interact with products before purchase. Therefore, the following hypothesis is formulated:
H4: customer involvement is positively associated with online purchase intention
2.1.6 E-tailer awareness
Reputation and image of e-tailers are often used as extrinsic cues for quality and consumers rely more on the e-tailers reputation, especially consumers with not so much of online shopping experience (Grewal, 1998). Wilson (1985) confirmed the importance of top-of-mind awareness in a study which found that the higher the position of the brand in the consumer's mind measured by unaided recall, the higher the purchase intention and the higher the relative purchase of the brand. Without awareness occurring, no other communication effects can occur and for a consumer to buy a brand they must first be made aware of it (Macdonald, 2003). Also attitudes cannot be formed, and purchase intention cannot occur unless awareness has occurred
(Macdonald, 2003). Yoon (2002) showed that company awareness and familiarity positively affect online purchasing. According to Keller (2003), awareness is important as there is a connection between awareness and purchase intention.
and we tend to like things that are familiar (Hoyer and MacInnis, 2009). As Klein and Quelch (1996) point out, new users tend to explore sites with familiar brands first. According to Allen and Fjermestad (2001) recent surveys have shown that 46 per cent of new online shoppers prefer to buy from e-tailers they had previously bought from offline. Even 34 per cent of repeat online shoppers preferred the familiar offline store sites. According to Cho (2008) customers strongly attached to shopping at physical outlets would not necessarily shy away from making purchases online. Brands that equate their products with an experience (e.g. feelings, associations, and memories) will likely be more effective than brands based on facts about a product and belief-based brands associate themselves with attributes such as high quality or reliability (Allen and Fjermestad, 2001). These attributes can be easily proven by an impartial display of the facts on a navigator’s website (Allen and Fjermestad, 2001). Even if the facts confirm the brand, it may only be rendering the brand redundant. Brands that are associated with a mixture of beliefs and experiences should play up the experiential side of the brand (Evans and Wurster, 1999).
According to Yoo (2008) consumers tend to employ a decision process that can be represented by a phased decision rule when involving many choice alternatives, i.e. consumers may simplify the purchase decision-making process by filtering available alternatives and then performing detailed analysis of the reduced number of alternatives. In this phased decision-making process, the term consideration set refers to the subset of all available brands brought to a consumer’s mind on a particular choice occasion. A consideration set defines the pool of brands from which the eventual choice is made. Therefore, accurate identification of the consideration set is essential to the practical success of the decision-making model. Consumers have to become aware of the e-tailer before having a purchase intention among the brands in their consideration set (Aaker, 1992).
Concluding, an important role for e-tailers will be branding. Brown and Hoyer (1990), state that awareness will keep the brand in an evoked set, which increases the probability to purchase at the e-tailer in the future. Therefore the following hypothesis is formulated:
2.1.7 Brand familiarity
Generating favorable attitudes is one of the most frequently used communication objectives of advertising, and prior research showed that favorable attitudes toward the brand positively affect click-troughs and purchase intention (Cho, 2003; Yoo, 2008). Furthermore, a number of
researchers have underlined the importance of consideration sets, e.g. Yoo (2008) pointed out that a brand that is not considered cannot be chosen. Brand familiarity is a continuous variable that reflects a consumer’s level of direct and indirect experiences with a brand (Ha, 2004).
Research evidence also indicates that brand familiarity reduces the need for information search, i.e. study revealed that consumers tend to spend less time shopping for a familiar brand than they do for an unfamiliar brand (Biswas, 1992 in Ha, 2005).
According to Kania (2001), familiarity with a company or brand generates higher trust, unless a person has a negative perception of a brand. Laroche et al. (1996) confirms the influence of brand familiarity on confidence, suggesting that a consumer's confidence toward a brand may result from his/her familiarity or experience with the brand. Also, according to Laroche et al. (1996) additional evidence shows that confidence in brand evaluation is one of the determinants of purchase intention. The latter is also influenced by the context, as defined by the set of alternatives under consideration. An obvious implication of these findings is that, in order to increase a consumer's intention to buy a specific brand, a marketer needs to enhance his/her confidence in the brand and may be realized by providing the consumer with more product related information, or direct experience.
Concluding, it seems that online purchase intention is higher for e-tailers who are more familiar. According to Kania (2001) familiarity with an e-tailer will generate higher trust, and according to (Biswas, 1992 in Ha, 2005) brand familiarity reduces the need for information search. Meaning, less search effort and more search convenience which is in this research the online search experience. Therefore, the following hypotheses are formulated:
H5b: brand familiarity is positively associated with online purchase intention H5c: brand familiarity is positively associated with perceived trust
2.2 Website characteristics as antecedents of online purchase intention
Song and Zahedi (2005), show that web design elements play an important role in shaping online customers’ attitudes and purchase behaviour. According to Hall and Hanna (2004) preferred colors (i.e. blues and chromatic colors) lead to higher ratings of aesthetic quality and purchase intention. Furthermore, ratings of aesthetic quality are significantly related to intention to purchase. This is underlined by Yoo and Donthu (2001) who state that e-tailers must be of high quality to attract consumers and influence their shopping decisions. Sismeiro and Bucklin (2004) show that page design can change preferences, and thus purchasing decisions, by influencing attribute importance.
According to Korzaan (2003) a webdesign must inspire users to experience the psychological state of flow. Hanson (2000) says that a well designed site has a number of beneficial impacts. It can build trust and confidence in the company; reinforce an image of competence, functionality, and usefulness; alert the visitor to the company’s range of products and services; and point out local dealers, upcoming special events, and reasons to come back again (Yoo, 2001). Hall and Hanna (2004) find that the degree to which consumers saw pages as pleasing and stimulating was linked with the degree to which they intended to purchase a given product. This relationship is relatively unexplored with respect to web pages. According to Hall and Hanna (2004) the visual aesthetics are a fundamental component in determining the effectiveness of e-commerce sites. Mathwick et al. (2004), state that a consumer can respond to the entertainment dimension of the aesthetic response. Both visual appeal and the entertainment dimension of the aesthetic response offer immediate pleasure for its own sake, irrespective of a retail environment’s ability to facilitate the accomplishment of a specific shopping task (Mathwick et al., 2001).
finding indicates that the designers of e-tailers need to pay particular attention to this category of web elements. So, providing comprehensive, complete, and rich information on a website is an important driver of online purchase intention. Websites contain not only product and service information with varying depth and breadth but also text, graphics, and multimedia about the product to compensate for the constraints inherent in online shopping. Offering rich resources for purchase facilitation enhances customers’ sense of control over the purchasing process.
Concluding, Sismeiro and Bucklin (2004) and Song and Zahedi (2005) both underline the importance of webdesign elements playing an important role in online purchase behaviour. Also, according to Hanson (2000) webdesign has an interaction effect with trust in a company. Recently, Verhoef et al. (2007) show that website investments in consumer trust can convert online searchers into online purchasers. So, the following two hypotheses are formulated:
H6a: An attractive web design is positively associated with online purchase intention H6b: Webdesign is positively associated with perceived trust
2.3.1 Site characteristics
Site characteristics like search functions, download speed, and navigation, play a role in shaping ease of use of the website (Zeithaml et al., 2002). Furthermore, a customer’s purchase intention may depend on the available opportunities and resources, such as the ability to view, evaluate, and test the product or service (Song and Zahedi, 2005). According to Sismeiro and Bucklin (2004) use of interactive decision aids improves the quality of purchase decisions and raises the possibility that the use of interactive decision aids on e-tailer website can lead to a higher online purchase intention.
According to van der Heijden et al. (2003) ‘perceived ease-of-use directly influenced the attitude towards purchasing online. However, van der Heijden et al. (2004) state that ease of use has a significant impact on the attitudes and intentions towards visiting a website. Whether these findings carry over to explain ‘the intention to visit a website to purchase online’ is not found. So, ease of use is an antecedent of enjoyment, rather than a direct antecedent of attitude towards purchasing.
et al. (2003; 2004) state the importance of ease of use in contribution towards online purchase intention, but van der Heijden (2004) also shows an interaction effect between ease of use and online shopping experience (enjoyment). So, the following hypotheses are formulated:
H6c: Ease of use is positively associated with online purchase intention H6d: Ease of use is positively associated with online shopping experience H6e: Site interactivity is positively associated with online purchase intention
2.3 Conceptual Model
Chapter 3: Research design
This chapter presents an overview of the research method that will be used to test the hypotheses of figure 5. From the exploratory and qualitative research in chapter 2, a conceptual model is build. This chapter ends with how the data is collected and the plan of analysis.
3.1 Type of research
A descriptive research will be performed to quantify the data and apply a statistical analysis. Descriptive research is characterised by the preceding formulations of hypotheses, therefore stressing the importance of clearly defined research problems (Malhotra, 2008).
3.2 Operationalization of the antecedents
Each antecedent of the concept needs to be operationalized. According to Steenkamp and Baumgartner (2000) multiple items need to be used to measure the antecedents in order to account for measurement error. Each antecedent is described in this paragraph and concludes with a table that displays the items used in this research, their sources and their item number. An overview of the operationalizations can be found in Appendix 1.
Customer involvement
The concept customer involvement with online shopping is measured using customer reviews using seven items from Park (2007) and Park (2008): 1) when I buy a product online, I always read reviews that are presented on the website, 2) when I buy a product online, the reviews presented on the website are helpful for my decision making, 3) when I buy a product online, the reviews presented on the website make me confident in purchasing the product, 4) if I do not read the reviews presented on the website when I buy a product online, I worry about my decision, 5) when I buy a product online, reading the reviews presented on the website impose a burden on me, 6) when I buy a product online, reading the reviews presented on the website irritate me and 7) I would highly recommend this site to others.
Perceived risk
monitor the safety and security of sending sensitive personal and financial information while shopping on the Internet (Martín and Camarero, 2008). The five items included in the
measurement of perceived risk are from Martín and Camarero (2008): 1) I feel like my privacy is not protected at this e-tailer, 2) purchasing through this e-tailer leads to uncertainties, 3) things can easily go wrong when I purchase through this e-tailer, 4) I think the site has mechanisms that warrantee the safe transmission of its users’ information and 5) I do not feel safe in my
transactions with this e-tailer.
Perceived risk is also measured using the term “product risk” from Verhoef et al (2007). Product risk is defined according Verhoef et al. (2007) as the perceived uncertainty in buying products through a specific channel, due to things such as payment issues, and lack of privacy. Verhoef et al. (2007) measure product risk according the following items: 1) there is a large probability that I do not get the right television when buying online, 2) it is difficult to judge the quality of the television and 3) there is a considerable chance that the television will be less than expected, when I buy through this e-tailer
Perceived trust
Perceived trust is measured according to Bart et al. (2005) with two items: 1) this site appears to be more trustworthy than other sites and 2) my overall trust in this site is high. The other two items that Bart et al. (2005) mention are not included in this research, instead one measurement from Martín and Camarero (2008) and one measurement from Kuan (2007) are included, these are respectively: 1) I think the information provided on this Web site is true and honest and 2) this site would act in my best interest.
Online shopping experience
Online search experience
Online search experience is measured using search effort and search convenience according to Verhoef et al. (2007), Bart et al. (2005) and Martín and Camarero (2008). According to Verhoef et al. (2007) search convenience is the perceived ease and speed at which consumers can gather information on products in the specific channel. This is measured by two items: 1) I can get information on the television on each time of the day in and 2) I can quickly get information on this television on the Philips Online Shop.
Search effort is defined as the perceived required time (time costs) and perceived
difficulty for consumers to gather information on the products and services according to Verhoef et al. (2007). This is measured using three items: 1) I think it costs a lot of time to search for information on the television on the Philips Online Shop, 2) I think collecting information on the television costs a lot of effort in the Online Shop of Philips and 3) I think it is difficult to collect information on the television on the Philips Online Shop.
Bart et al. (2005) measure search effort and convenience with four items: 1) the process for browsing is clear, 2) the site uses a layout that is familiar, and 3) information on the site can be obtained quickly and 4) I can get much information on the television in the Philips Online Shop. The last research which measures search effort and convenience used in this research is from Martín and Camarero (2008), who measures search effort with one item: website browsing is easy.
Site characteristics
Site interactivity is measured using Martín and Camarero (2008): 1) the intention is to promote interactivity with the visitors experience and 2) I perceive the shopping experience as if I were buying in the physical store.
Ease of use is measured according to Kuan (2007) and van der Heijden et al. (2003): 1) learning to use the website is easy and 2) it is easy to get the website to do what I want and 3) I can easily compare prices (Kuan, 2007)
Awareness
Previous research from Yoon (2002) measures awareness with the following item: I’m aware of this website of Philips. This is also how awareness is measured in this research.
Brand familiarity
According to Bart et al. (2005) brand familiarity can be measured using eight items: 1) I am familiar with the company whose site this is, 2) the site represents a quality company or organization, 3) the site carries products and services with reputable brand names, 4) I am
generally familiar with other brands being advertised on the site, 6) the site is consistent with my image of the company whose site this is, 7) before this survey, I was familiar with the site, 8) I have made a purchase on this site in the past.
Online purchase intention
Online purchase intention refers to the intentions to purchase online. Laczniak et al. (2001) state that purchase intention can be measured with one question: “if you were to buy a product from a certain category, how likely are you to purchase this particular brand?”. This question is transformed to match this research into: “if you were to buy a product from a certain online shop, how likely are you to purchase at this online shop?”
Additionally, Bart et al. (2005) used five items to measure online purchase intention: 1) I would purchase an item at this site, 2) I would recommend this site to a friend, 3) I am
item to measure online purchase intention. So, in total there are six items that measures online purchase intention.
3.3 Data collection
This research uses a survey technique by making use of an online questionnaire (see Appendix 2). The benefits of this data collection choice include the ability to ensure comparability of the data, increase speed and accuracy or recording and facilitate data processing (Malhotra, 2008). Whilst several types of surveys can be made, this study focuses on personal surveys and will be made on the Internet to reach an acceptable response rate (Malhotra, 2008). The questionnaire was setup using movies from an e-tailer website (Philips Online Shop) that guided the customer from the start (homepage) to the end (check-out page) of the online shopping process. Each movie lasted around 1.5 minute. In between the respondents answered questions about the certain web pages. Each movie had its own question-page and the respondent could only go to the next page after clicking a box saying ‘I have seen the movie’. This would ensure that all respondents have seen the movie before answering the questions about that part of the online buying process. Also, to ensure a high number of respondents, a discount voucher for an e-tailer was given to the respondents.
A structured-direct survey is conducted where the questions are fixed-alternative questions that required the respondent to select from a predetermined set of responses. This survey method has several advantages (Malhotra, 2008). First the questionnaire is simple to administer. Second, the data obtained are reliable because the responses are limited to the alternatives stated. Third, this provides the most flexible way of obtaining data and getting information on underlying motives. Finally, coding, analysis, and interpretation of data are relatively simple. Disadvantages are that respondents may be unable or unwilling to provide the desired information. The survey is filled in by 235 respondents, with a total of 153 respondents completely filling in the survey.
One question of the questionnaire has an interval measurement scale (age) and one question has a nominal scale (gender). All questions measuring the antecedents are of interval measurement scale. At last, collecting data from respondents, the anonymity should be protected and their names should not be turned over to the client.
3.4 Population and sample
The sampling design process has four steps: defining the target population, determining the sampling frame, selecting a sampling technique, and determining the sample size (Malhotra, 2008).
The target population is the collection of elements or objects that possess the information sought by the researcher and about which inferences are to be made (Malhotra, 2008). For this research the target population is the Internet consumers who make use of e-tailers for purchase intentions.
The second step is the sampling frame, which is according to Malhotra (2008) the
representation of the elements of the target population. It consists of a list or set of directions for identifying the target population. This research will be done online using online survey tool from surveytools.com. No registration is needed so the sampling frame consists of all Internet
consumers who participate in the research.
The third step is to select a sampling technique. This research will use a convenience sampling technique. However, a snowball-effect can happen when respondents forward the questionnaire to people they know, increasing the sampling size.
3.5 Plan of analysis
The analysis begins with data preparation and a preliminary check of the questionnaire for completeness and quality. If needed, the data will be edited and incomplete or inconsistent data will be deleted.
Reliability
To obtain the quantitative data an online survey is performed, which involves the direct
questioning of respondents. To test the reliability of the concepts the Cronbach’s alpha is used. This is the average of all possible split-half coefficients resulting from different ways of splitting the scale items. This coefficient varies from 0 to 1, and a value of 0.6 or less generally indicates unsatisfactory internal consistency reliability (Malhotra, 2008).
For the site characteristics there are three concepts. The web design (SC_WD) has an alpha of 0.869. For ease of use (SC_EoU) the item EoU4 is deleted to increase the alpha to 0.715. This item measures “I can easily compare prices in the Philips Online Shop”, which can be a reason that respondent thought they could easily compare prices between different e-tailers, instead of between product on the Philips Online Shop. The site interactivity (SI) item has an alpha of 0.413 which is too low to make it one concept. So, SC_SI1 and SC_SI2 are taken separately and are now respectively called site interactivity (visitor experience) and site interactivity (physical store).
Perceived risk (PR) has an alpha of 0.813, all items except item 4 (information warranty) have been rotated. Customer involvement (CI) has an alpha of .704, online shopping experience (SE) 0.809 and perceived trust (PT) 0.753. For online search experience (OSE) the alpha is 0.853, for this concept the items of OSE3_Info_cost_effort and OSE4_info_collection are reversed so that the higher the score, the more positive the answer is. For example, an outcome of a seven on the Likert-scale now means that searching the site costs minimum effort. The dependent concept of online purchase intention (OPI) has an alpha of 0.853. Appendice 1 shows a summary of the Cronbach’s alpha for each factor.
Regression
Subsequently, a regression analysis will be performed with online purchase intention as dependent concept and the antecedents as independent concepts. This research will use two regression analyses; bivariate and multiple regression analysis. If the t-test is associated with a significant β value (less than 0.05) then that predictor is making a significant contribution to the model. The smaller the significance, the greater the contribution.
Bivariate regression defines a line in a two concepts space by the equation Y=a+b*X; the Y concept is expressed in terms of a constant (a) and a slope (b) times the X concept. The
standardized beta values (βi) are used as these beta weights are standardized on Z-scores, have a
mean value of 0 and a standard deviation of 1. This way the dependents concepts can be compared to eachother and see which one is the stronger predictor.
Equation 1: Online purchase intention = α + β1* Webdesign; β2* Site interactivity (visitors
experience); β3* Site interactivity (physical store); β4 * Ease of use; β5 * Perceived risk; β6 *
Customer involvement; β7 * Perceived trust; β8 * online shopping experience; β9 * Awareness;
β10 * Brand familiarity; β11 * Online search experience + e
α = constant
βi = regression coefficient e = error term
The purpose of multiple regression is to learn about the relationship between several independent (X) concepts and a dependent (Y) concept. Multiple regression analysis is complicated by the presence of multicollinearity, which is a state of very high intercorrelations among independent concepts (Malhotra, 2008). Multicollinearity causes unclarity and decreases the size of the multiple correlation. This happens when the predictors correlate so closely that you can’t tell which predictor is doing the actual predicting. To check for multicollinearity the tolerance and VIF values will be taken into account
Equation 2: Online purchase intention = α + β1* Webdesign + β2* Site interactivity (visitors
Customer involvement + β7 * Perceived trust + β8 * online shopping experience + β9 * Awareness
+ β10 * Brand familiarity + β11 * Online search experience
α = constant
βi = regression coefficient
3.6 Development of the research
Chapter 4: Results
This chapter presents the results of the analyses. The results will lead to either rejecting or accepting the stated hypotheses in chapter 2. The general and key findings are discussed and are the starting point for the conclusion and recommendation chapter. The general findings, which are the demographic profiles of the respondents, can be found in table 3. The hypotheses are listed at the end of the chapter in table 8.
4.1 General findings
A total of 235 respondents filled in the online survey, however only 153 did complete the online survey (see table 3). The average age of the respondents is 26 years old and more males (65%) than females (35%) filled in the survey. The frequency that a respondent visits an e-tailer is 1-3 times a week with the highest being a bit over 30% of the respondents.
Table 3
Demographic profile of study participants (N = 153)
Gender Male: 65%
Female: 35%
Average Age 26 years
Frequency of e-tailer visits Never: 0.7%
Less than once a month: 14.4% Once a month: 18.3%
Every two weeks: 25.5% 1-3 times a week: 31.4% Once a day: 6.5%
Several times a week: 3.3%
5-6 years: 30.1%
More than 7 years: 16.3%
Frequency of e-tailer purchases Never: 2.6%
Less than once a month: 54.2% Once a month: 32.0%
Every two weeks: 9.2% 1-3 times a week: 2.0%
Channels Compare prices between offline and online: 23.5%
Buy products or services: 49.7%
Getting other information then prices about products: 26.8%
Important aspects of an e-tailer I don’t use online shops: 2.6%
Better service than other online shops: 5.9% Better reputation than other online shops: 17.6% Better website design than other online shops: 2.6% Easier to use than other online shops: 14.4%
Cheaper prices than other online shops: 27.5% Better assortment than other online shops: 29.4%
Collecting information for online purchasing
Asking their friends: 23.5%
Reading online customer reviews: 61.4% Searching the web for information: 79.7%
4.2 Key results
The results of the hypotheses will be discussed and explained in this chapter. The SPSS-output can be found in Appendix 3, same goes for the concepts and their meaning. The bivariate regression was done using online purchase intention as the dependent concept and an independent concept. The R2 shows the proportion of the total variance in online purchase intention is predictable from X1 (Malhotra, 2008). The standardized coefficient (often called Beta and given the symbol β) represents the correlation between the concept and online purchase intention (Malhotra, 2008). The F-rate is a measure of how much the model has improved the prediction of the outcome compared to the level of inaccuracy of the model (Malhotra, 2008). The outcome of the bivariate regression analyses can be found in table 4.
Concepts R2 F2 β Sig. Bivariate regression Y
Online purchase intention+ web design .083 13.664 .288 .000** Y = α + .288*SC_WD Online purchase intention + site interactivity
(visitors experience)
.079 13.015 .282 .000** Y = α + .282*SC_SI1
Online purchase intention + site interactivity (physical store)
.161 28.907 .401 .000** Y = α + .401*SC_SI2
Online purchase intention + ease of use .114 19.388 .337 .000** Y = α + .337*SC_EoU Online purchase intention + perceived risk .162 29.179 0.521 .000** Y = α + .521*PR Online purchase intention + customer involvement .016 2.470 .127 .118 (NS)
Online purchase intention + perceived trust .357 83.661 .597 .000** Y = α + .597*PT Online purchase intention + online shopping
experience
.173 31.549 .461 .000** Y = α + .461*SE
Online purchase intention + awareness .012 1.830 1.830 0.178 (NS)
Online purchase intention + brand familiarity .116 19.853 0.341 .000** Y = α + .341*BF Online purchase intention + online search
experience
.106 17.859 .325 .000** Y = α + .325*OSE
Online search experience (SE)
Online search experience has an R2 of 0.173, meaning web design accounts for 17.3% of the
variation in online purchase intention. The F is 31.549 and significant at p < 0.001. This means that there is less than a 0.1 % chance that an F-ratio this large would happen if the null
hypothesis were true. Therefore, the regression model results in significantly better prediction of online purchase intention then if the mean value of online purchase is used. The β is 0.461 for SE.
H1 is accepted.
Perceived trust (PT)
Perceived trust has an R2of 0.357, meaning web design accounts for 35.7% of the variation in online purchase intention. The F is 83.661 and significant at p < 0.001. This means that there is less than a 0.1 % chance that an F-ratio this large would happen if the null hypothesis were true. Therefore, the regression model results in significantly better prediction of online purchase intention then if the mean value of online purchase is used. The β is 0.357 for PR.
H2a is accepted.
Perceived risk (PR)
Perceived risk has an R2of 0.162, meaning web design accounts for 16.2% of the variation in online purchase intention. The F is 29.179 and significant at p < 0.001. This means that there is less than a 0.1 % chance that an F-ratio this large would happen if the null hypothesis were true. Therefore, the regression model results in significantly better prediction of online purchase intention then if the mean value of online purchase is used. The β is 0.521 for PR.
H2b is accepted.
Online shopping experience (OSE)
Online shopping experience has an R2 of 0. 106, meaning web design accounts for 10.6% of the variation in online purchase intention. The F is 17.859 and significant at p < 0.001. This means that there is less than a 0.1 % chance that an F-ratio this large would happen if the null
Customer involvement (CI)
Customer involvement is not significant and so H4 is not accepted.
Awareness (A)
Awareness is not significant and so H5a is not accepted.
Brand familiarity (BF)
Brand familiarity has an R2of 0. 116, meaning web design accounts for 11.6% of the variation in online purchase intention. The F is 19.853 and significant at p < 0.001. This means that there is less than a 0.1 % chance that an F-ratio this large would happen if the null hypothesis were true. Therefore, the regression model results in significantly better prediction of online purchase intention then if the mean value of online purchase is used. The β is 0.341 for BF.
H5b is accepted.
Site characteristic: Ease of use (SI_EoU)
Ease of use has an R2of 0.114, meaning web design accounts for 11.4% of the variation in online
purchase intention. The F is 19.388 and significant at p < 0.001. This means that there is less than a 0.1 % chance that an F-ratio this large would happen if the null hypothesis were true. Therefore, the regression model results in significantly better prediction of online purchase intention then if the mean value of online purchase is used. The β is 0.337 for SC_EoU.
H6a is accepted.
Site characteristic: Site interactivity (SC_SI)
Therefore, the regression model results in significantly better prediction of online purchase intention then if the mean value of online purchase is used. The β is 0. 282 and 0.401 are respectively for SC_SI1 and SC_SI2.
H6b is accepted
Site characteristic: Web design (SC_WD)
Web design has an R2of 0.083, meaning web design accounts for 8.3% of the variation in online purchase intention. The F is 13.664 and significant at p < 0.001. This means that there is less than a 0.1 % chance that an F-ratio this large would happen if the null hypothesis were true. Therefore, the regression model results in significantly better prediction of online purchase intention then if the mean value of online purchase is used. The β is 0.288 for SC_WD which is the one of the lowest of the significant results.
H6c is accepted.
Summary
The value of R2 is lowest for SC_SI1 (0.079) and the highest for PT (0.357), it can be concluded that these concepts account for 7.9% to 35.7% of the variation in online purchase intention in a bivariate setting. Most of the concepts account for approximately 10% of the variation in online purchase intention. The F is between 13.015 and 83.661 and significant at p < 0.001. This means that there is less than a 0.1 % chance that an F-ratio this large would happen if the null hypothesis were true. Therefore, the regression model results in significantly better prediction of online purchase intention then if the mean value of online purchase is used. In short, the regression models overall predicts online purchase intention significantly well.
Multiple regression
The multiple regressions are done using online purchase intention as dependent concept and tall concepts simultaneously as independent concepts, based on equation 2. For the overall multiple regression to predict online purchase intention from all concepts, R2 = .479 (see table 5). That is, when all concepts are used as predictors, about 47.9% of the variance in online purchase
intention is predicted. The overall regression is statistically significant, F = 11.776, p < .005.
In this multiple regression model, SC_SI2, PT and BF are significant predictors of OPI (see table 6). From the magnitude of the t-statistics we can see that PT has a slightly more impact than SC_SI2 and BF. SC_SI1 has a significance of almost 1, indicating it has virtually no impact whatsoever (note that its beta value is extremely close to zero).
Perceived trust (PT) is significantly predictive of online purchase intention: t = 4.802, p < .001. The positive slope for perceived trust as a predictor of online purchase intention indicates that there is about a 0.427 increase in online purchase intention for each 1-unit increase in perceived trust, controlling all other concepts. Brand familiarity (BF) is also significantly predictive of online purchase intention when all other concepts were statistically controlled: t = 3.180, p =.002. The slope to predict online purchase intention from brand familiarity weight was approximately b = +.234; in other words, there was about a quarter-point increase in online purchase intention in 1-unit increase of brand familiarity.
Concept R2 F2 Sig.
Online purchase intention + Webdesign + Site interactivity (visitors experience) + Site interactivity (physical store) + Ease of use + Perceived risk + Customer involvement + Perceived trust + Online shopping experience + Awareness + Brand familiarity + Online search experience
.479 11.776 .000**
Table 5: summary of multiple regression dependent concept = OPI
