WHAT TYPE OF RECOMMENDATIONS SYSTEMS AFFECT SATISFACTION AND LOYALTY OF ONLINE GROCERY SHOPPERS?

(1)

1

WHAT

TYPE

OF

RECOMMENDATIONS

SYSTEMS

AFFECT

SATISFACTION AND LOYALTY OF ONLINE GROCERY SHOPPERS?

and the mediation role of satisfaction

(2)

2

WHAT

TYPE

OF

RECOMMENDATIONS

SYSTEMS

AFFECT

SATISFACTION AND LOYALTY OF ONLINE GROCERY SHOPPERS?

and the mediating role of satisfaction

Master Thesis Marketing Management and Marketing Intelligence 2015/2016

June 20, 2016

Zaandam/Drachten, the Netherlands

Suzanne Hilde (S.H.) van Dijk Laagland 25, 9205 EX Drachten (NL) Student number: 1868489 s.h.van.dijk.2@student.rug.nl shvandijk@gmail.com +31 6 10 81 91 87 University of Groningen Faculty of Economics and Business

Department of Marketing

PO Box 800, 9700 AV Groningen (NL) First supervisor:

dr. J.E.M. van Nierop

j.e.m.van.nierop@rug.nl

Second supervisor: dr. H. Risselada

(3)

3

Management Summary

Online shopping is increasing in popularity. Also buying your groceries belongs to the opportunities nowadays and is increasing in market share. The current two big players in the Dutch grocery market are Albert Heijn and Jumbo but more and more parties are entering the market and trying to gain market share.

A lot of data is available about customers and for the supermarkets to survive in this market it is important to gather as much information as possibleabout their customers. When customers are buying groceries online their basket size and content can be analyzed easily so why not take advantage of it by recommending customers what they presumably want to buy? Different types of recommendations systems can help as a reminder, for example products that are bought on a weekly base are very likely to be in the basket next week as well. A central question in this research is whether consumers appreciate help while completing their grocery list. This research investigates the effect of weekly and monthly reminders, and additional and impulse product suggestions on satisfaction and loyalty.

Suggestions can be made based on the content of the basket of consumers. This can be done for additional products as well as for impulse suggestions. The effect of these type of recommendations systems on satisfaction and loyalty are also covered in this research. The effects of the different types of recommendations systems on satisfaction and loyalty are investigated by a regression analysis. Furthermore, for the additional product suggestions the effects of different unit prices are investigated by using a conjoint analysis.

With online data gathering privacy also becomes an issue so these effects are also accounted for by including privacy as an moderator on the relationship on satisfaction and loyalty. When assessing the different types of recommendations systems and their effect on satisfaction and loyalty the weekly reminder is used as the reference category. Furthermore, it can be concluded that satisfaction acts as a mediator on the relation between type of recommendations systems and loyalty. The effect for monthly reminders is completely mediated by satisfaction which means that before a monthly reminder results in loyal customers, the grocery store should make sure the respondent is satisfied. The effect for additional and impulse buying suggestions is partially mediated by satisfaction.

Preferences for different unit prices (high, medium, and low) are examined which eventually resulted in two segments with different preferences. These different preferences for unit prices are linked to customer satisfaction and customer loyalty. One segment has a clear preference for products with lower unit prices and is more satisfied and loyal.

Regarding the moderator effect, there is no evidence found that privacy influences the relationship between type of recommendations system and satisfaction or loyalty.

(4)

4

Preface

This Master thesis is the last step before graduating from my Master Marketing Intelligence and Marketing Management at the University of Groningen. My initial plan was to write my thesis during a graduate project at a large – preferably – retailing company in the Netherlands. Finding a suitable internship was more difficult than expected, so I decided to participate in the regular thesis program of the University. A week before starting writing my Master Thesis I heard that I was accepted for an internship at my #1 company, the E-commerce department of Albert Heijn Online at Ahold. This was a great opportunity to show them my qualities and hopefully it would open opportunities for a job after my graduation. I decided to accept the offer and to face the challenge of writing my thesis at the same time. Knowing in advance that it would be a tough semester, it really turned out the be a tough semester in the broadest sense. Not only because of the combination of a full-time internship and writing a Master Thesis, also because of setbacks in my personal life. Despite all setbacks writing my thesis was a good and pleasant “distraction” and I am very proud to present you my final thesis!

I would like to thank several people. First of all my first supervisor Erjen van Nierop for his open attitude, flexibility, willingness to provide useful feedback and his patience. I would also like thank my group members Justa Schouten and Jolien Mennink for their input, brainstorm sessions, and useful feedback and my dear friend Jacobien van Klinken for motivating me. A big thank goes to my parents Tjerk van Dijk and Tineke van Dijk-Hibma, and my sister Jessica van Dijk for their support and love during this last semester.

Last but not least, I would like to thank my colleagues at the E-commerce department of Albert Heijn Online for the educative internship, and for showing me how a large company operates in practice.

(5)

5

1 Introduction

Fifteen years ago, literature already supported that selling groceries online is growing at a phenomenal pace. Tat Keh and Shieh (2001) forecasted that as the internet continues to grow, so will e-commerce. Ever since, online shopping for groceries has increased dramatically over the last few years, and now tops the agenda of all major grocery retailers (Melis, Campo, Breugelmans and Lamey, 2015).

The trend shows that Dutch consumers are spending more and more money online, which is supported in a recently presented report by GfK. The report shows that Dutch consumers spend 16.07 billion online in 2015, whereof 4.71 billion in the last quarter of 2015. A striking observation according to statistics is that the biggest growth in online spendings are in the food/near food branch. In this research the fourth quarter of 2015 is compared with the fourth quarter of 2014. During this period there is an increase of 54% in the food/near food branch (for comparison: the increase for the home and garden branch is 38%, 37% for toys, household electronics 31%, and 29% for consumer electronics)1. Furthermore, the revenue of online grocery shopping in the Netherlands is 450 million, which is 1.3% of the total revenue of grocery stores2_.

Online grocery shopping is a trend that has not gone unnoticed by current retailers in the market. The Dutch grocery market contains many retailers. The two biggest players in the market (Albert Heijn and Jumbo, with market shares of relatively 35.1% and 17.4%)already offer their customers different possibilities to pick up groceries and to have groceries delivered at home. Some retailers offer the same possibilities (Plus and DekaMarkt), whereas other retailers only offer a part of the services (Coop, Deen, Poiesz and Hoogvliet). There are also some retailers left in the market without the possibility to order groceries online (Lidl, Emté, Aldi,Vomar and Dirk van den Broek). The state of affairs of the five largest players in the Dutch grocery market (market share > 5%) are shown in table 1.

TABLE 1

State of affairs Dutch retailers 3

Retailer % market-share Assortment online available Assortment online orderable Homedelivery service available Assortment Size

AH 35.1% yes yes yes 26.000

Jumbo 17.4% yes yes yes 19.000

Lidl 10% partly (only promotions)

no no 1.500

Aldi 7,3% partly no no 1.500

(8)

8

E-tailers in the grocery market already conquered their share in the current grocery market and they keep expanding their territory. Some consumers already buy their groceries online but there are definitely opportunities to grow, since 55% of the respondents - of a research conducted in Europe3 - state that they are willing to buy their groceries online in the nearby future. In order to keep improving the services for their customers retailers have many different options to embrace digital growth within their shopping environment, like mobile coupons, possibilities to make grocery lists online, store-apps, and Wi-Fi available in the store. With online grocery shopping consumers can use the latest technology and retailers can utilize many different and flexible options to improve the shopping experience, by using different channels and attract more customers to increase their market share and eventually generate more revenue. The growth of online grocery shopping is partly the result of the fact that digital “natives” – consumers who grew up with digital technology – reached the adulthood. These consumers have an unprecedented enthusiasm for technology and feel completely comfortable using it.4 Along the way retailers already improved their services, but with this ongoing trend, it is of crucial importance that retailers understand their digital consumer touchpoints and keep innovating along the route to purchasing groceries online. An important motivation to add an online alternative is that the extra online service can increase customer satisfaction and loyalty, and help to retain existing customers (IGD, 2012; Zhang, Farris, Irvin, Kushwaha, Steenburgh and Weitz, 2010).

Surveys show that consumers typically dislike grocery shopping, and those with busy schedules should increasingly find online purchasing a viable option. Unique selling point of selling groceries online are convenience and time savings, which appeal to busy consumers or those who simply dislike shopping for groceries (Tat Keh and Shieh, 2001). By using a combination of channels, retailers can better satisfy their customers’ needs by exploiting the benefits and overcoming the deficiencies of each channel. The traditional store provides certain unique benefits but on the other hand, for example, consumers need to spend time and energy visiting stores. The stores may not be open at convenient times for consumers. Purchasing groceries online can offer the convenience of buying merchandise whenever and wherever consumers want to and lower time to make purchases. In addition to these benefits offered, the internet enables customers to get as much information as they desire before making a purchase. Moreover, the web-based information can be tailored to the customers’ needs (Zhang et al., 2010).

One of the possibilities to extend the current e-commerce options for consumers and to keep innovating and fulfilling the changing needs of customers is to offer customers help with their grocery shopping list. When consumers are buying their groceries at the same retailer, some pattern arises which can be recognized by the retailer. Patterns can offer useful insights to the e-tailing business because in this way marketers have the opportunity to study consumers buying habits and target their promotions effectively. Online shopping enables manufacturers to link purchases to consumers in ways that traditional stores cannot (Tat Keh and Shieh, 2001).

(9)

9

Furthermore, by providing a greater array of benefits through multichannel operations, retailers can increase their share of customers’ wallets (Ansari, Mela, and Neslin, 2008). However, the downside is that shoppers are sensitive about revealing information regarding their buying behavior and lifestyles. Many are not willing to trade convenience for privacy (Tat Keh and Shieh, 2001).

Until now, a lot of research is done in the field of customer satisfaction and loyalty (Verhoef and Donkers, 2005; Gensler, Dekimpe, and Skiera, 2007; Neslin and Shankar, 2009; Taylor, Donovan and Ishida, 2014; Choi, 2015). This paper will contribute to the current literature by investigating how recommendations systems can contribute to the customer satisfaction and customer loyalty of online grocery shoppers. The research question derived is “What type of recommendations systems affect satisfaction and loyalty of online grocery shoppers?” There are multiple other types of grocery shoppers but discussing other types than the online customers is beyond the scope of this research. Furthermore, this research examines whether satisfaction acts as a mediating role on the relationship between the different types of recommendations systems and loyalty. Lastly, privacy related issues regarding list predictions and lower unit prices for value packs and their effect on satisfaction and loyalty will be discussed.

The remaining structure of this article will be organized as follows: chapter two discusses a theoretical framework concerning the research questions and hypotheses to finally derive a conceptual model. Chapter three explains the research methodology and data collection method. The chapter thereafter, chapter four, presents the results of the analysis and discusses the outcomes. Chapter five contains of the conclusions of the research and presents answers to the research questions as well as recommendations, limitations and suggestions for future research.

1.1 Aim of the research

Convenience is an important attribute in retail innovation. The highly challenging nature of food and grocery retailing has resulted in consumers being exposed to diverse retail format alternatives. Grocery retailers need to differentiate and innovate to serve the needs of customers better than competitors and make it easier, faster, and convenient for their customers. Innovation persuades shoppers to do something different, but newness wears off, and retailers must continually develop new ways to capture interest in order to stay ahead of the competition. Online loyalty is important because of the competitive nature of the dynamic online market and the ever-increasing number of online retailers. As mentioned before, when consumers start online grocery shopping, they tend to select the online store belonging to the same chain as their preferred offline store (Melis et al., 2015). When online grocery shopping experience increases, the focus shifts from a comparison within a chain across channels to a comparison across chains within the online channel. When shoppers gain more online buying experience, retailers should invest in satisfied online shoppers in order to keep customers loyal to the online store.

(10)

10

for data about their shopping behavior. Discounts are the most important motivation for customers to use a loyalty card, which explains the use of loyalty programs (Cedrola and Memmo, 2010). Vice versa, the e-tailer can use the collected information to learn from and to target and segment different customers. Customer loyalty cards are already introduced at for example Albert Heijn. A next step in the innovation process is introducing a recommendations system based on previous purchase behavior of customers. What the effects of these recommendations systems are will be the main contribution of this research.

2 Theoretical Framework

This chapter contains a theoretical framework of the concepts discussed during this research. First a profile of the online grocery shopper is outlined where the needs will be explained together with recent innovations and the responses of convenience oriented retailers. Next the different types of recommendations are explained, followed by the definitions of customer satisfaction and customer loyalty. Subsequently privacy is discussed, followed by unit pricing. Thereafter the mediator effect of satisfaction is explained and the conceptual model is presented at the end of this chapter.

2.1 The online grocery shopper

Every household needs groceries, but in the last decade the way in which the groceries can be purchased has dramatically changed. There are now different ways to purchase your groceries, which means there are different types of customers with different types of needs. There are multiple reasons for customers to buy their groceries online. First of all, the number of time-starved consumers is increasing (Tat Keh and Shieh, 2001). For example couples with two careers, children and above-average income want to spend their spare time with their families, so they are concerned more with convenience than price. As the role of woman keeps changing, the rise in the number of dual-income families is expected to continue. Such households, especially those with young children, will be more willing to embrace the concept of having their groceries delivered directly to their homes. Secondly, there are other types of consumers who also show a willingness to use online services (Tat Keh and Shieh, 2001). These include people who dislike grocery shopping and those with disabilities related to health and age. Besides these two clear reasons for online grocery shopping, McGovern (1998) reports that the demographics of consumers who demonstrate a willingness to use online services cut across all income and educational levels, age groups and locations. People across all market segments are becoming more and more comfortable with computers and the internet, expanding the limits of shopping and buying online beyond computer experts. It can be concluded that it is difficult to define one specific type of “online grocery shoppers” that includes all characteristics, meaning that the online grocery shopper has many different characteristics and needs.

2.1.1 Customer needs

(11)

11

also need to differentiate themselves from their competition. Many shoppers are focused on getting value for their money, and e-commerce has attracted a growing number of users.

2.1.2 Innovations for grocery shopping

New ideas in convenience make shopping faster and easier for consumers, but can go further to leverage customer data to predict purchases and fill needs. Innovation persuades shoppers to do something different, it only captures consumers’ attention for a brief time as it brings in new customers or improves loyalty. Newness wears off, and retailers must continually develop new ways to capture interest in order to stay ahead (Nielsen, 2014). Ma, Gill and Jiang (2015) investigated that the effect of innovation locus on consumer adoption intentions is contingent on the newness of the innovation. They defined the peripheral locus as a component of the product system that is physically detached from the core product and offers discretionary benefits. New products increasingly differ in innovation locus, partly because peripheral innovations are increasingly commonplace and are a significant source of revenues for firms. Also, those products which differ most within the category will be perceived as newer, and will gain consumer’s attention because this newness differentiates them from the product clutter (Radford and Block, 2011).

2.1.3 Convenience oriented retailers

Of these convenience-oriented retail innovations, the shift towards multichannel offline-online retailing is one of the most important and successful practices. Several of the large grocery retail chains (such as Walmart, Tesco and Ahold) now operate an online store next to their offline supermarket outlets (‘brick and click’ grocery retailers). By increasing their service levels, multichannel retailers aim to retain existing customers and gain new customers in the increasingly competitive retail environment (Chintagunta, Chu, and Cebolla 2012; Kabadayi, Eyoboglu, and Thomas 2007; Neslin and Shankar 2009; Zhang et al., 2010).

Online-offline channel integration is integrating access to and knowledge about the offline channel into an online channel. Although channel integration has been acknowledged as a promising strategy for retailers, its effect on customer reactions toward retailers and across different channels remain unclear and evidence for the success of multi-channel retailers remains scarce (Herhausen, Binder, Schoegel and Hermann, 2015).

Retail firms with physical stores realize that they may have a physical advantage over purely online players. However, some multi-channel retail firms with physical stores still struggle to answer the important yet unsolved question of whether they can create competitive advantage from a multi-channel strategy. The reason is often due to the lack of integration between a firm’s Internet and physical stores. Consequently, real collaboration between retailers’ physical stores and digital stores remains rare (Herhausen et. al, 2015).

(12)

12

Customers value an integrated internet store and online-offline integration leads to more favorable behavior toward a retailer and its internet store. Online-offline integration not only enhances search intention, purchase intention, and willingness-to-pay in the internet store but is also a source of competitive advantage for the whole firm (Herhausen et. al,, 2015). Research shows that there is no significant cannibalization of the physical store, so internet channels complement rather than substitute physical channels (Avery, Steenburgh, Deighton, and Caravella, 2012) .

2.2 Recommendations systems

A personalized recommender system can be used to suggest new products to grocery shoppers based upon their previous purchase behavior. The recommender can also mean to provide an alternative source of new ideas for customers who visit the store less frequently. Recommendations are generated by matching products to customers based on the expected appeal of the product and the previous spending of the customer. Products to be recommended can be determined by computing a measure of distance between vectors representing personal preferences and vectors representing products (Lawrence, Almasi, Kotlyar, Viveros and Duri, 2002).

There is a set of categories in which recommendations are more welcome than others. Customers want interesting recommendations. The list consists of recommendable products, emphasizing those product classes in which the spending percentage from the recommendation list exceeded that on the main shopping list, and de-emphasizing the others, with the aim of creating a more “fun” or welcome set of recommendable products. Recommendation algorithm combines aspects of content and collaborative filtering to rate new products for a customer based on their prior purchase behavior (Lawrence et al., 2002).

In grocery stores, large-scale transaction data with identification is being accumulated as a result of the introduction of frequent shopper programs. The accumulated data have been used to examine customer shopping behavior, especially by professionals in the marketing field. Recommendations based on this data are often adopted in e-commerce shopping stores, they are rarely introduced in face-to-face selling, such as in brick-and-mortar grocery stores. Therefore, introducing a system based on these recommendations to grocery stores could induce customers to visit the store of this particular retailer to make a purchase, either in a physical store or online (Sano, Machino, Yada and Suzuki, 2015). However, a downside is that an evaluation for product item recommendation is very sparse, since the number of product items a customer purchases is very few among all the products available in the assortment. This sparsity of evaluation values could result in product item recommendation performing poorly (Sano et al., 2015).

(13)

13

On the one hand, recommendations can help consumers in completing their basket to make sure they don’t forget anything. Here is also an advantage for the retailers, because – in an ideal situation - they can sell more to loyal customers. For example a product that consumers would not think of buying, products the retailer is selling and consumers are not aware of, or products consumers would otherwise buy at a competitor but can be persuaded to buy it at the same retailer. Retailers can also sell products they prefer selling, for example private label with higher margins.

The contribution of this research is to provide an overview of how a recommendations system based on previous purchase behavior of customers can improve the customer satisfaction and eventually customer loyalty. The first two recommendations systems used are weekly and monthly reminders. This includes reminders based on groceries bought on a regular base for the previous weeks or months. Next the two remainder recommendations systems will be explained.

2.3 Additional groceries and impulse buying

Kollat and Willet’s (1967) definition of impulse buying is an in-store decision that occurs without prior recognized need for the item in order to distinguish impulse purchases from unplanned reminder grocery purchases. This unplanned reminder purchase would be classified as a planned purchase if the shopper had remembered to put the item on his shopping list. A pure impulse purchase has no “reminder” component since there was no prior recognized need (Kacen, Hess and Walker, 2012). The unplanned reminder buy reflects a purchase decision made at a previous point in time (Stern, 1962). A true impulse purchase reflects an at-the-moment, in-store decision and is therefore subject to greater influence from the store environment, and the consumer’s current state at the time of shopping (Beatty and Ferrel, 1998; Cobb and Hoyer, 1986).

Offline entities, like consumer product giant Proctor and Gamble Co. spend millions on in-store marketing efforts, believing that the first three to seven seconds when a shopper notices a product on the shelf, what Proctor and Gamble Co. refers to as the “first moment of truth”, is critical to the purchase decision (Nelson and Ellison, 2005). Also brick and mortar supermarkets spend a lot of effort in persuading customers at the checkout to add a product to their basket. Online grocery shoppers can also be persuaded to add an extra product to their basket. A large advantage of online shopping is that the e-tailer knows exactly what is currently in the customer’s basket and an impulse product can be personalized right away based on the groceries in the basket or the past shopping behavior of the customer.

Therefore, the two remainder recommendations systems used are additional product suggestions – containing additional groceries that can be bought in combination with a product which is already in the basket and impulse buying product suggestions as previously described. Next the definitions of customer satisfaction and customer loyalty are explained in more detail.

2.4 Customer satisfaction definition

(14)

14

predictors (antecedents) of customer satisfaction in a typical food and grocery retail setting. Customer satisfaction significantly affects the store loyalty (behavioral and attitudinal), repurchase intention, positive word-of-mouth, price insensitivity, and complaint behavior (Jayasankaraprasad, Venkata and Kumar, 2012).

The highly challenging nature of food and grocery retailing has resulted in consumers being exposed to diverse retail format alternatives. As retail choices have proliferated, shoppers have undergone important changes in their expectations, shopping tendencies and strategies for where and how they will satisfy their ever increasing shopping needs (Sinha, Banerjee and Uniyal, 2002).

Due to the quick changes in the grocery retail market, retailers are facing challenges to differentiate themselves by serving the needs of customers better than their competitors (Hemalatha, Jagannathan and Ravichandran, 2010). Understanding and predicting consumer behavior-related issues (e.g., expectations, satisfaction, store loyalty, repurchase behavior) has become an interesting subject of research. Furthermore, given the wide diversity of choice in retail formats from which to choose, discerning consumers are continually making new choices of purchasing source that would give maximum value for their money, time and mental effort spent in shopping (Findlay and Sparks, 2008).

To test whether different types of recommendations systems have a different effect on customer satisfaction, the following hypotheses are formulated, where weekly reminder is used as a reference category.

H1a: Monthly reminders are expected to have a negative effect on satisfaction when compared to a weekly reminder.

H1b: Additional product suggestions are expected to have a negative effect on satisfaction when compared to a weekly reminder.

H1c: Impulse buying product suggestions are expected to have a negative effect on satisfaction when compared to a weekly reminder.

2.5 Customer loyalty definition

Internet retailing is growing rapidly in popularity among consumers in all sectors of retailing. A major challenge facing internet retailers is in the area of customer loyalty or e-loyalty which is defined as ‘the customer’s favorable attitude toward an electronic business, resulting in repeat purchasing behavior’ (Anderson and Srinivasan, 2003). The high importance placed on online loyalty is because of the competitive nature of the online market and the ever-increasing number of online retailers. The internet also makes it relatively easy and less costly for consumers to search for alternative suppliers and to comparison-shop, as well as giving them the ability to switch suppliers at the click of a button. This makes it even more important to build and maintain customer loyalty online (Rafiq, Fulford and Lu, 2013).

(15)

15

frequency, and as a result, in the levels of experience with buying in the online grocery channel. Multi-channel shoppers, at the start of online grocery shopping, tend to select the online store belonging to the same chain as their preferred offline store, especially when the online store is strongly integrated with the offline store in terms of assortment. When online grocery shopping experience increases, multi-channel shoppers’ focus shifts from a comparison within a chain across channels to a comparison across chains within the online channel, resulting in an increasing importance of offline assortment attractiveness and online loyalty when choosing an online store (Melis et al., 2015).

Relationship marketing theory suggests that it is more valuable for a retailer to invest effort in developing and maintaining close and long-lasting relationships with customers rather than attracting short-term, discrete transactions (Kumar, Bohling and Ladda, 2003; Reichheld and Schefter, 2000). Customers in such relationships are found to purchase more, willing to pay more for goods and/or services, to exhibit a high tendency to trust, to become emotionally attached to that firm and to refer customers to the firm.

Multi-channel retailers should be careful in assuming that their loyal offline customers will select and remain with their store when buying groceries online. Substantial differences between the online and offline store, or a less competitive online offer, may lead loyal customers to select – or switch to – a competing online store, especially when their online buying experience increases. When gaining more online buying experience, satisfied online shoppers may develop loyalty to the online store, reinforcing the tendency to revisit this store online (Melis et. al., 2015). But on the other side, Junhong, Arce-Urrizza, Cebollada-Calvo and Chintagunta (2010) show that households tend to be more brand loyal and size loyal in the online channel than in the offline channel.

Loyalty is highly important in a competitive world. With the increasing number of retailers and the easily accessible internet, there are low barriers to switch to competitors and it is easy to compare products and prices. For retailers it is important to keep customers loyal by constantly adjusting to their needs. Loyalty becomes especially important when online grocery shopping experience increases because customers’ focus tends to shift from a comparison within a chain across channels to a comparison across chains within the online channel. According to earlier mentioned literature, loyalty is ‘the customer’s favorable attitude toward an electronic business, resulting in repeat purchasing behavior’ (Anderson and Srinivasan, 2003). It assumes that a loyal customer will return to the retailer and buy his or her grocery again at the same retailer. As shown by previous literature, businesses try to retain customers by satisfying them, building loyalty (Ali and Muqadas, 2015) and maintain long-term relationships with customers (Gremler, Gwinner, and Brown, 2001). Therefore, the following hypotheses are established, where again weekly reminders are used as the reference category.

H2a: Monthly reminders are expected to have a negative effect on loyalty when compared to a weekly reminder.

(16)

16

H2c: Impulse buying product suggestions are expected to have a negative effect on loyalty when compared to a weekly reminder.

2.6 Privacy

Advances in ICT and network technologies have facilitated the lives of consumers in many ways but have at the same time aggravated uncertainties of invasion into their private lives (Kervenoal, Soopramien, Hallsworth and Elms, 2007). Consumers are concerned about how privacy will affect their usage of the internet in the future (BBC, 2006). Organizations need more comprehensive data to enhance their understanding of consumers’ decision-making processes. A wide range of technologies are accessible to track down user’s web surfing practices and purchasing tendencies (Boncella, 2001). Privacy is an important attribute which has a negative impact on the perceived “service quality” of commercial websites (Zeithaml, Parasuraman and Malhotra, 2002; Wolfinbarger and Gilly, 2003).

According to Berman and Bruening (2003) privacy entails an individual’s right to control the collection and use of his or her personal information, even after he discloses it to others. When individuals provide information, they expect that those companies will collect the information they need to deliver a service and use it for that sole purpose. Individuals expect that they have the right to object to any further use.

Whilst consumers express concern over privacy in general, giving basic information on “their privacy” is sometimes the only way in which they can take advantage of the online market environment (Sheehan, 2002), which means that they have to “give up” their privacy in order to participate in the loyalty program. If customers are not willing to give up on their privacy they cannot experience the benefits. To find out how privacy issues are related to grocery list prediction, privacy is a part of this research.

As mentioned before, consumers are concerned about their privacy. In order to make matching predictions and collect comprehensive data, it is important that the quality of the data is high. With highly qualitative data meaningful predictions can be made. In exchange for high quality data to make valuable predictions consumers should be willing to trade their data and feel like it is favorable to them as well.

Overall, it can be assumed that shoppers who are more concerned about their privacy will score lower on satisfaction as well as on loyalty. In this case privacy acts as a moderator on this relationship. Therefore, privacy is expected to negatively influence the relationship between the different types of recommendations systems and both satisfaction and loyalty. In order to find out whether this differs for the different recommendations systems, again weekly reminders are used as reference category and the remainder recommendations systems are interpreted by using weekly as a reference category. The following two hypotheses are formulated:

H3a: Privacy is negatively moderating the relationship between the different types of recommendations systems and customer satisfaction.

(17)

17

Normally, when a moderator moderates the effect of the independent variable on the dependent variable, the independent variable is a variable that really ‘varies’. The tough part about these hypotheses is that recommendation system is not really a variable: it is present or not and is therefore used as a dummy variable. Therefore, the results will show whether privacy is present in the specific types of recommendations systems instead of internet grocery shopping in general.

2.7 Unit pricing

As described before, online grocery shopping has many disadvantages and advantages when compared to brick and mortar stores. One advantage is that online shopping makes it easier to compare products. One way in which products can be compared is based on the price per unit. Displaying unit prices leads consumers to choose lower unit priced product options. The presence of unit prices increases the salience of price in decision making, making consumers more price-sensitive which in turn activates a greater motivation to select cheaper products. This motivational effect persists even when unit prices offer little cognitive benefit, such as when product options have identical sizes (Yao and Oppewal, 2015).

This research examines the effect of value packs, whether consumers prefer buying a value pack – which costs more money, but price per unit is cheaper, and whether there is a difference between fresh and preservable goods. This effect will be linked to customer satisfaction and customer loyalty in order to establish whether consumers who prefer a lower unit price are also more likely to be satisfied and or loyal.

When retailers decide what groceries will be presented to consumers this can lower the perception of the degree of influence by consumers. However, the product offered to the consumer may be more expensive, but the price per unit will be lower which is eventually cheaper for the customer. Whether lower unit prices will lead to more satisfied and loyal customers is tested by the following hypotheses:

H4a: Consumers with a preference for value packs with lower unit prices are more satisfied. H4b: Consumers with a preference for value packs with lower unit prices are more loyal.

2.8 Mediator effect of satisfaction

To compete and to have competitive advantage over competitors, businesses try to retain customers, satisfy them and build loyalty (Ali and Muqadas, 2015). Among the academic and professional fields, interest is growing in identifying the factors that influence customer loyalty, with the aim of developing the most appropriate market action strategies (Picón, Castro and Roldán, 2013). The literature establishes that satisfaction is the key determinant of customer loyalty (Oliver, 1999). The satisfaction level with the usual provider is the main factor in determining loyalty towards that provider. Satisfaction with the value of the product or service is the key determinant of customer loyalty (Zeithaml, Berry and Parasuraman, 1996).

(18)

18

time rather than buying from multiple suppliers within the category”5_{. Satisfied and loyal}

customers make purchases from the same manufacturer; they are less affected from the incentives and price promotions offered by the competitors (Dimitriades, 2006). Therefore, satisfaction, long term relations with customers and loyalty are considered as an important factor for owner’s success in today’s economic world. By fulfilling customer’s needs and by attracting them companies attempt to build loyalty among customers and have long term relationships with them (Gremler et al., 2001). Loyalty occurs as the result of customer satisfaction. The satisfaction may or may not result in loyalty (Oliver, 1993).

As previously shown, satisfaction is the key determinant of customer loyalty. To test whether satisfaction acts as a mediator for customer loyalty in this research, the following hypothesis will be tested:

H5: Customer satisfaction acts as a mediator on the relationship between the different types of recommendations systems to customer loyalty.

2.9 Conceptual Model

Based on previous described literature the conceptual model is shown in figure 1.

FIGURE 1 Conceptual model

The independent variable is the type of recommendations system, consisting of four different levels which will be manipulated; namely weekly reminder, monthly reminder, additional groceries and impulse buying, where weekly reminder is chosen as the reference category. This means that monthly reminders, additional product suggestions and impulse buying suggestions are interpreted by referring to weekly reminders as base category. During this research will be