How effective are the sales promotions utilised by Dutch food delivery services? : the effects of sales promotions on consumer behaviour

(1)

How effective are the sales promotions utilised by Dutch food

delivery services?

The effects of sales promotions on consumer behaviour

Msc. Business Administration – Entrepreneurship & Innovation Track Faculty of Economics and Business

University of Amsterdam

Student: Yannick Visbeek, 10084789 Supervisor: dr. A.S. (Alexander) Alexiev

(2)

Statement of Originality

This document is written by Yannick Visbeek who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

Abstract

In the growing Dutch food delivery market, delivery services frequently use sales promotions to attract customers and increase sales. This research studies and compares the effectiveness of using two different types of sales promotions, discounts and competitions, in the context of food delivery. As previous research indicates that the magnitude of the benefit that a promotion provides is an important determinant of promotional effectiveness, the effectiveness of using either discounts or competitions across three (low, moderate, high) different benefit levels is studied. When using discounts, another important aspect is the presentation of the discount. Hence, the differences between framing discounts either in percentage terms or cents terms are studied in this research. The results of the research indicate that discounts are more effective than competitions across all benefit levels. In terms of framing discounts, no difference was found in using either a percentage-off or a cents-off format. This in contrast with the expectation that using a percentage-off approach would be more effective, due to the relatively low costs of delivery.

(7)

1. Introduction

The market for food delivery in the Netherlands has been showing a steady growth over the last few years and this growth is expected to continue in the years to come (Driessen, 2016). In 2015, Rocket Investment, a company that has an impressive track record when it comes to rapidly scaling young internet businesses, introduced a new food delivery service in the Netherlands: Foodora. However, Foodora is not the first business operating in the Dutch food delivery market. Besides Foodora, there are numerous other food delivery services available in the Netherlands, of which the largest and most well-known are Thuisbezorgd, Foodora, Deliveroo, HelloFresh and UberEats. Because of the high number of delivery services, competition in this market is high (“Hoe Foodora Nederland wil veroveren”, 2016).

The largest delivery service in the Netherlands, Thuisbezorgd, which controls about 80 percent of the total Dutch food delivery market, delivers over one million orders each month. The growth of the Dutch food delivery market is reflected in the core numbers: the market for food delivery has almost doubled within the last three years and of the total amount of 57 billion euros that is spent on food each year, 1 billion is spent on home delivery of food (“Markt voor foodbezorging in ruim”, 2016).

Since competition in the market for food delivery is high, the companies that are active in this market utilise various methods to try and gain a competitive advantage. Examples are offering price promotions or offering exclusive deals with certain restaurants. UberEats, for instance, upon first entering the Amsterdam market, tried to gain market share by not charging any delivery costs and by not asking for a minimum order amount (“Uber gaat maaltijden bezorgen”, 2016).

In such an intensely crowded and competitive market, where switching costs, the costs of time, money and effort associated with changing service providers are low, how do food

(8)

delivery services gain market share and keep it (Chou & Lu, 2009)? Given the fierce competition in the food delivery market, businesses are looking for effective promotions to increase their market share.

Choosing and executing the right types of promotions can be a difficult task. W developing promotions, careful attention should be given to the characteristics of the market, competitors’ characteristics, behaviour and promotions, as well as to consumers’ characteristics (Alvarez & Vásquez, 2005).

1.2 Problem statement and research question

Almost all organisations engage in some form of promotion. By utilising promotional strategies, most of the time businesses try to convince consumers to take some form of action (Peattie & Peattie, 1995). Although the ultimate objective of executing promotional strategies is to directly influence consumer buying behaviour, examples of other promotional objectives, which are hoped to be eventually linked to sales, are improved brand recognition and creating a favourable climate for future sales (Rowley, 1998). Research shows that promotion directly affects consumer behaviour; it leads to repurchase behaviour, brand switching and increased purchase volume (Peattie & Peattie, 1995). Of all promotional strategies, sales promotions are the most used type of promotion.

As commercial businesses try to convince consumers of buying their products and services, the main tool that is used in measuring advertising effectiveness and marketing campaigns in general, is consumers’ purchase intention (Li et al., 2002). Increasing consumers’ purchase intention would very likely result in an increase of actual sales numbers, as is predicted by the theory of planned behaviour of Ajzen & Fishbein (1980). According to their theory, intention is the closest proxy for actual behaviour. Besides purchase intention there are

(9)

two other measures of the effectiveness of sales promotions: the perceptions of offer value and search intentions (Palazon & Delgado-Ballester, 2009).

Investing in sales promotions is not a guarantee for success. A business should plan its promotion activities carefully, taking into consideration the objectives of the promotion. It is essential for a business that is looking to promote its products and/or services that a promotional tool that is suited for the brand and the market in which they operate, is chosen (Peattie & Peattie, 1995; Alvarez & Vásquez, 2005).

In a market, crowded with businesses offering almost identical services, that keeps on welcoming new businesses, a business has the ability to, amongst others, differentiate itself from the others, attract untapped parts of the market and increase market share by employing the right types of promotion. But what are the right types of promotion in the food delivery market, a market that has low switching costs and is operating through the internet? This will be investigated in this thesis by studying the elements of, and finding an answer to the following research question:

How effective are the sales promotion activities employed by food delivery services in the Netherlands?

The following sub-questions will contribute to answering the research question:

- What is the effect of sales promotions on the perceived value of the offer? - What is the effect of sales promotions on purchase intention?

- What is the effect of sales promotions on search intention?

- What is the moderating effect of promotional benefit level on the effectiveness of sales promotions?

(10)

1.3 Relevance

The market for food delivery is growing and this promises an immense growth potential for both delivery services and restaurants, ABN-AMRO reported in 2016 (Driessen, 2016). In the report, it is stated that the food delivery market is supposed to grow by 3 percent next year, of which most of the growth comes from urban areas and student areas. This growth creates opportunities for new businesses and restaurants to generate additional revenue streams.

Marketers spend a lot of time trying to find out what consumers really want and what promotions will be most effective (Palazon & Delgado-Ballester, 2009). Sales promotions have become an important tool in marketing and have been increasingly getting attention from both scholars and practitioners (Nusair et al., 2010). This thesis will give insight into the effectiveness of different promotional activities that food delivery services currently employ. For many businesses promotional expenditures account for a large part of total marketing costs (Hardesty & Bearden, 2003). Hence, it is of particular importance to understand the effects these activities have on consumers. Knowing how consumers respond to promotional activities of different services will give a better understanding of the current existence of promotion thresholds and might be building blocks for general rules on sales promotions (Gupta & Cooper, 1992). One such general rule, for instance, is that when consumers are able to anticipate promotions, the effectiveness decreases, as consumers might incorporate the incentives of promotions into the characteristics of a product or service, which can result into consumers not buying the product when there aren't any promotions included (Lewis, 2006). So, it can be important for any type of business to know that the objectives of a sales promotion will be reached to a larger extent, when they are introduced sparingly and when consumers least expect it.

This research will contribute to current research on promotional activities and its effects on consumers in a different setting. Although research has already been conducted on

(11)

promotional activities and their effects, as Alvarez & Vásquez (2005) state, the effects are largely dependent on the market in which a business operates. The objective of this research is to determine the effects of promotional activities, specifically in the relatively little studied market for food delivery, on consumers’ purchase intention. The scope of this thesis is limited to sales promotion activities utilised by Dutch food delivery services and its effects on consumers who reside within the Netherlands.

(12)

2. Theoretical framework

In this chapter, relevant literature will be discussed. Here, the concepts that are relevant to this research will be explained. This will lead to the conceptual model and, based on the discussed literature, the formulation of the hypotheses on which this master thesis research is based.

2.1 Promotions

To communicate their offerings to customers or prospects, businesses use promotion, one of the four basic P’s of marketing (Rowley, 1998). Promotion was defined by Rowley (1998) as “the combined efforts of businesses to create channels in order to inform and persuade consumers of their products and services”. The definition of Peattie & Peattie (1995) is similar to Rowley’s definition. They defined promotions as “marketing activities, tied to a specific time period, place or customer group, that encourage a direct response from consumers or market intermediaries, through the offering of additional benefits”. Examples of promotional objectives are the increase of sales, increase of market share, improved brand recognition and the creation of a favourable climate for future sales (Rowley, 1998). Research shows that promotion directly affects consumer behaviour: it leads, amongst others, to repurchase behaviour, brand switching and increased purchase volume (Peattie & Peattie, 1995). The feedback that is received on the promotional strategy will largely be in the form of sales as it is difficult to differentiate between the effects of the different promotional activities and other marketing mix factors (Rowley, 1998).

The different channels that can be utilised to communicate promotional messages make up the promotional mix (Rowley, 1998). The promotional mix is based on the framework developed by Boone & Kurtz (2007). The elements that make up the promotional mix are advertising, direct marketing, public relations, personal selling, sponsorships and sales

(13)

promotions (Boone & Kurtz, 2007). The focus of this thesis is on the element of sales promotions, which is defined in the next section.

2.2 Sales promotions

Sales promotions, which have accounted for the largest share of the marketing budget in consumer packaged goods categories since the early 1970’s, are used as short-term incentives in order to encourage the purchase of a product or service or to get possible consumers to try the service or product (Currim & Schneider, 1991; Rowley, 1998; Pauwels, Hanssens, & Siddarth, 2002).

Sales promotions provide consumers with multiple benefits, such as monetary savings, convenience, self-expression, added-value, entertainment and increase of quality (hui et al. 2013). They have mostly been studied in a retail context, where it has been suggested that they affect short term consumer behaviour more than any other type of promotion (Laroche et al., 2003). Examples of sales promotions are short term price discounts, coupons and trial periods (Rowley, 1998). Offering acquisition discounts to potential customers lowers the risk that is associated with trial. By lowering this risk, offering trials and/or short-term promotional discounts can help acquire new customers, since customers are provided with a low-cost environment in which the product can be tested (Lewis, 2006). Besides encouraging trial of a product, sales promotions also have the ability to help smooth fluctuations in demand, to target different segments with different pricing strategies and give excitement to services that might otherwise be perceived as commodities (Farris and Quelch, 1987). Examples of other sales promotion strategies, that do not include a direct monetary reward, are the offering of free gifts and premiums (Hardesty & Bearden, 2003).

Price discounts are by far the most used form of sales promotions used by businesses (Darke & Chung, 2005). Since price discounts are used so often, it becomes easy for consumers to

(14)

predict them, which could lead to a lower perception of the value and quality of the product or service (Hardesty & Bearden, 2003). This effect can be diminished when other indicators of quality, such as the brand name, are provided. The possible lowering of quality perceptions has led to other pricing strategies, like every-day-low-price policies, as these convey value to consumers more directly, so product quality will not be undermined (Chandon et al., 2000). Framing the discount in a certain way can also maintain quality perceptions. An example of this is indicating both the initial selling price and the discount price (Hardesty & Bearden, 2003).

2.3 Sales promotion benefits

As research indicates that monetary benefits are not able to fully explain why and how consumers respond to sales promotions, Chandon et al. (2000) developed a framework of the benefits that sales promotions provide consumers. These benefits can be divided into two types, utilitarian (extrinsic) and hedonistic (intrinsic) benefits.

Benefits from sales promotions can be seen as utilitarian when they help consumers maximise their utility and efficiency when buying a product or service. Sales promotions are viewed as hedonistic when they make the buying process fun and provide intrinsic motivation. The benefits that are considered to be of utilitarian nature are the savings, quality and convenience benefits of sales promotions. The reason being that these help consumers increase the acquisition utility and enhance the efficiency of the shopping experience. The benefits that are considered to be hedonistic are the entertainment and exploration benefits. The reason behind this, is that they are intrinsically rewarding and are related to emotions, pleasure and self-esteem. Chandon et al. (2000) state that value expression benefits can be seen as both hedonistic and utilitarian considering that buying promoted products and services can give

(15)

consumers the moral satisfaction of being a “thrifty” shopper but it could also increase consumers’ prestige as they might be recognised as smart shoppers.

2.4 Sales promotion types

Sales promotions can be divided into two groups, monetary and non-monetary sales promotions (Peattie & Peattie, 1995; Chandon et al., 2000; Büttner et al., 2015). In this section, the most commonly used sales promotions are explained.

2.4.1 Monetary promotions

In monetary promotions, the most obvious benefit for consumers is in the saving of costs. These promotions try to increase the value of a product or service by manipulating the “price/value equation. These types of promotions are also labelled “value-increasing” promotions (Peattie & Peattie, 1995). Examples are reducing the price through temporary price reductions and coupons or increasing the amount a consumer receives for the same price by, for example, offering 10 percent extra for the same price (Büttner et al., 2015). Employing monetary promotions can increase sales numbers and market share and encourage product trial and brand switching, by inducing consumer buying behaviour in the short-term (Peattie & Peattie, 1995). However, offering monetary promotions can also have a negative impact on potential future sales. The use of promotions such as temporary price discounts can create or strengthen a “discount” association with a brand as it can negatively affect brand equity and price sensitivity (Büttner et al., 2015; Buil et al., 2013). Consequently, this affects consumer loyalty and how consumers respond to future price changes and other marketing efforts (Keller, 1993). Lewis (2006) states that customers might refuse to buy the product again when they have to pay the regular price, one that does not include the discount, so they might only buy when the product or service is discounted. Besides, monetary promotion campaigns are usually too short to create

(16)

long-term associations with brands and can create brand uncertainty. This will result in more negative perceptions of brand value (Buil et al., 2013). Below, the most common types of monetary promotions are explained.

2.4.1.2 Price discounts

Price discounts are by far the most used promotional tool across service industries (Nusair et al., 2010). By offering price discounts to consumers, businesses give consumers the ability to temporarily buy the same product or service at a reduced price. Research shows that this type of promotion is particularly effective in inducing immediate sales and trial of products and services (Shi et al., 2005).

2.4.1.3 Coupons

By offering coupons, businesses keep the original price of the product or service the same, but give consumers the option to receive the product or service at discount when redeeming the coupon (Shi et al., 2005). In general, consumers are more responsive to price discounts than they are to coupons as making use of coupons requires more involvement (Shi et al., 2005). As Appendix 1 shows, almost all sales promotions used by food delivery services utilise coupon style strategies where consumers need to redeem the coupon codes they receive through email or by physical mail in order to receive the promotional offer.

2.4.1.4 Buy-one-get-one-free

This type of promotion, which in research literature is labelled as a “bonus pack”, is usually offered at the original price but an additional amount of the product or service is added. This additional amount can persuade consumers into buying the product or service (Shi et al., 2005). Traditionally, researchers have considered bonus packs to be a non-monetary promotion, but

(17)

as Peattie (1998) states, adding an extra quantity of a product of service is value-increasing as it manipulates the price-quantity relationship the same way discounts do. Therefore, in this research, these promotional offers are considered to be of monetary nature.

2.4.2 Non-monetary promotions

Non-monetary promotions are becoming an increasingly employed tool in promotional strategies (Buil et al., 2013). These types of promotions represent the types of promotions that don't affect the price of a product, so they do not necessarily provide consumers with a direct financial benefit. These types of promotion bundle a product or service with something else and are also named value-adding promotions (Peattie & Peattie, 2015). Examples of non-monetary promotions are free gifts, free samples, complementary products and competitions (Büttner et al., 2015; Peattie & Peattie, 2015). Where monetary promotions are said to have a negative effect on brand association in the long term, non-monetary promotions have a more positive effect on brand equity and price sensitivity as they do not alter the internal reference price of a product or service (Buil et al., 2013). But, even when taking the above-mentioned advantages into account, monetary promotions are still used more often as non-monetary promotions are usually much less attractive for consumers than monetary promotions (Büttner et al., 2015). The most common offered non-monetary promotions are outlined below.

2.4.2.2 Free samples

By offering free samples to consumers, trial of a product or service is stimulated, with the intent that consumers will start buying once they understand the benefits of the product or service. It has been shown that making use of free samples improves quality judgement (Bawa & Shoemaker, 2004). Offering free samples is most effective when it is able to create brand

(18)

knowledge and awareness and is the preferred promotional strategy when introducing a new brand or product (Bawa & Shoemaker, 2004).

2.4.2.3 Competitions/sweepstakes

Competitions offer customers the possibility to win prices when buying products or services. Competitions have been shown to be an effective type of promotion, not only for products, but also for services. Possible outcomes of using competitions are that value can be added for consumers, a business can differentiate itself from competitors, generate interest among consumers and generate mailing lists for potential future promotions (Peattie & Peattie, 1995).

2.5 Conceptual framework

2.5.1 Monetary vs. non-monetary promotions

In research literature, purchase intention has been one of the most used indicators of promotional effectiveness, together with consumers’ perception of promotional value and consumers’ benefits of additional search intention, where purchase intention is positively related to the overall perception of value (Palazon & Delgado-Ballester, 2009). Search intention can be defined as “a consumers’ willingness to search for additional price information”. As prices in the market generally vary, consumers are often uncertain of what the lowest price is, which results in them searching for price information to lower this uncertainty. The willingness to search for other promotions is contingent on the perceived benefit of a promotion and on the trade-off a consumer makes between search benefits (e.g. money saved) and search costs (e.g. money and effort) (Grewal et al., 1998). This means that when consumers highly value a certain sales promotion they will have a higher purchase intention for that promotion and a lower interest in searching for other promotions (Palazon &

(19)

Delgado-Ballester, 2009). Hence, promotional effectiveness will be measured by three variables: perceived offer value, purchase intention and search intention.

In their utilisation of monetary promotions, Dutch food delivery services mostly use discounts and when using non-monetary promotions, they make use of competitions. As research suggests that monetary promotions are more effective than non-monetary promotions, the following hypotheses will be tested (Shi et al., 2005; Gilbert & Jackaria, 2002; Nusair et al., 2010):

H1a: Perceived offer value is higher for price discounts than for competitions

H1b: Purchase intention is higher for price discounts than for competitions

H1c: Search intention is lower for price discounts than for competitions

2.5.2 Sales promotion framing

Research findings have demonstrated, although mostly in retail settings, that consumers do not only respond differently to various types of sales promotions, but also to the presentation of promotions. Price discounts, for instance, can be presented as a percentage-off discount or as a cents-off discount (Hardesty & Bearden, 2003). Although there are multiple ways to frame discounts, research is mostly focused on the differences between the above-mentioned methods, as these are most commonly employed in practice (DelVecchio et al., 2007). It is suggested that framing affects consumers’ estimation of the value of the promotion, which affects the perception of the price. Hence, as perceived price determines consumers’ expectations of future prices, framing affects purchase intention. An advantage of presenting discounts in a percentage-off format, is that the promotion is bounded by two extremes, zero percent and one

(20)

hundred percent. This allows consumers to compare the discounts across a range of products (Chen et al., 1998).

The attractiveness of a discount and the influence of framing does not only depend on the discount amount, but also on the original price (Chen et al., 1998). Chen et al. (1998) give the following example: a 1000-dollar discount on a 20.000-dollar product seems significant, but the equivalent framed as a percentage discount on the other hand, five percent, seems less attractive. This in contrast with lower priced items: a fifty percent discount on a 0,50-dollar product seems a lot more attractive than the absolute discount amount of 0,25 dollars. As the spending costs in the food delivery market are relatively low, this leads to the following hypotheses:

H2a: Perceived offer value is higher for percentage-off discounts than for cents-off discounts

H2b: Purchase intention is higher for percentage-off discounts than for cents-off discounts

H2c: Search intention is lower for percentage-off discounts than for cents-off discounts

2.5.3 Promotional benefit level

Multiple studies have indicated that the way consumers evaluate and process promotions affects their value perception of a promotion. Consumers seem to evaluate information more extensively for moderate discount levels than for low or high discount levels, which can be explained as follows: when discounts are low, it is unlikely that consumers process information

(21)

extensively as monetary value is low. On the other hand, when price discounts are high, consumers are also unlikely to process information extensively as the benefits are usually easily recognisable. But when the benefit level of a discount is moderate, there is relatively high uncertainty regarding the deal, so consumers will process information more elaborately (Hardesty & Bearden, 2003; Palazon & Delgado-Ballester, 2009). This is what Grewel et al. (1996) explained to be the inverted-U relationship of consumer information processing regarding reactions to sales promotions.

In general, the benefits of non-monetary promotions are more difficult to process than the benefits of price discounts and the possibility exists that consumers are more sceptical of non-monetary promotions (Hardesty & Bearden, 2003). As these types of bias can be reduced by processing information more extensively, it can be hypothesised that consumers value comparable price discounts and competitions equally for moderate benefit levels and that consumers value comparable price discounts higher than competitions for low and high benefit levels. Hence, there will be tested for the following two hypotheses:

H3a: The perceived value, purchase intention and search intention of price discounts and competitions are similar for moderate promotional benefit levels

H3b: The perceived value, purchase intention and search intention are not similar for price discounts and competitions for both low and high benefit levels

As mentioned above, consumers are expected to process information more elaborately for sales promotions with moderate benefit levels than for sales promotions with low or high benefit levels (Hardesty & Bearden, 2003). With moderate benefit levels, consumers take more time

(22)

to evaluate promotions, which results in them attributing equal value to promotions that are framed in a percentage-off format with promotions framed in a cents-off format. Where with hypothesis 2a, 2b and 2c there is tested for the general difference in the effectiveness of framing, with both hypothesis 4a and 4b there will be tested for the moderating effect of promotional benefit level on the relationship between the different framing methods and sales promotion effectiveness. This will be tested using the following hypothesis:

H4a: The perceived value, purchase intention and search intention of discounts presented in dollar terms and discounts presented in percentage terms are

similar for moderate benefit levels

For sales promotions with a high benefit level, information processing will be low and decisions will be made on simple decision heuristics (Hardesty & Bearden, 2003). A high discount percentage, for instance, indicates a good deal. So, as costs are relatively low, it is expected that consumers will be more attracted to sales promotions presented in percentage terms than discounts presented in dollar terms. Therefore, the following hypothesis will be tested:

H4b: The perceived value, purchase intention and search intention are not similar for price discounts presented in percentage terms and discounts presented in currency terms for sales promotions with a high benefit level.

2.6 Control variables

In this thesis, there will be controlled for food delivery order frequency and income, as these are expected to have some influence on the effectiveness of promotional activities

(23)

3. Research Methodology

In this chapter, the research methodology that was used in this research is explained.

3.1 Sample & data collection

First, primary data was gathered by distributing an online questionnaire to inhabitants of Amsterdam, the city with the largest population in the Netherlands. The survey targeted consumers of all ages who either regularly, or occasionally make use of food delivery services. The reason that such a large group was targeted is that numbers show that in the Netherlands it is expected that most people have prior experience with food delivery services: in the Netherlands, about 315 million portions are delivered to peoples’ homes each year, excluding the 560 million yearly takeout portions that are picked up by consumers (Driessen, 2016). Although market research indicates that mostly young people and men make use of food delivery services, it is assumed that all adults have experience with it.

3.2 Pre-test

In order to validate the promotional values, a pre-test was conducted with eleven subjects who get food delivered to their homes on a regular basis. In the first part of the pre-test they were asked to provide low, moderate and high discount percentages for promotional offers, based on their experience with food delivery. The subjects were also asked to rank several non-monetary promotions based on how much they value the promotions. The final values used in this research are based on the pre-test that was conducted, on a review of the promotions used in the marketplace and on a review of the values used in previous literature (Alford & Biswas, 2002, Hardesty & Bearden, 2003 & Palazon & Delgado-Ballester (2009). The values are illustrated in tables 3.1 and 3.2.

(24)

The final survey was pre-tested with several subjects in order to check for mistakes and to test the survey flow. The subjects who participated in the pre-tests were excluded from the final survey sample in order to avoid bias. Both the pre-test and final survey are illustrated in Appendix 2.

3.3 The experiments

In this section, short descriptions of the conducted experiments are provided. The first experiment was designed in order to measure the effects of using different types of promotion; discounts vs. competitions (i.e. monetary vs. non-monetary). The second experiment was designed to measure how different methods (i.e. percentage-off vs. cents-off) of framing monetary promotions affect consumer behaviour.

3.3.1 Experiment One

In Experiment One, a 3 promotional benefit level (low, moderate, high) x 2 promotion type (discounts vs. non-monetary) between subjects experimental design was employed. Table 3.1 shows the different offers that were shown to the participants. The promotion values were based on the pre-test and a review of the promotions Dutch food delivery services utilise. An excerpt from sales promotions offered from the beginning of 2017 is presented in Appendix 1. In this experiment, the promoted discounts were presented in dollars. For food delivery, competitions often offer consumers the possibility to win order credit, which can be used to order extra products for free. Participants were asked to read the offer after which they were asked to respond to several statements that will be used to identify the effectiveness of the promotional offer.

(25)

Benefit level Promotion type

Discount Competition

Low €2,50 on €20 Win festival tickets

Moderate €4 on €20 Win €500 delivery credit

High €6 on €20 Win a trip to South Africa

Table 3.1 Experiment one promotions

3.3.2 Experiment Two

This experiment was designed to investigate the effect of price discounts and the effects of differences in the discount format (i.e. percentage, dollar). In order to do this, a 3 discount level (low, moderate and high) x 2 discount formats (percentage-off and cents off) between subjects experimental design was used (see table 3.2). This design was based on the research of Nusair et al. (2010), who sought to understand the effects of framing price discounts and discount benefit levels on, amongst others, consumer purchase intention. As in Experiment One, the participants were asked to read the promotional offer after which they were asked to respond statements about perceived offer value, purchase intention and search intention.

Benefit level Discount framing

Absolute number Percentage

Low €2,50 on €20 12,5% on €20

Moderate €4 on €20 20% on €20

High €6 on €20 30% on €20

Table 3.2 Experiment two promotions

3.4 Procedure

For each of the experiments, subjects were presented with one of six possible promotional offers. A participant could, for instance, get to see the moderate competition offer (“Win €500

(26)

delivery credit”) from table 3.1 in the first experiment and the low benefit level percentage offer (12,5% on €20) from table 3.2 in the second experiment. In both experiments, the participants were asked to respond to the statements explained in the next section. The participants were also asked to respond to some demographical questions and they were asked to indicate how often they get food delivered to their homes.

3.5 Operationalisations of dependent variables

This section explains how the dependent variables, the variables that are being used to evaluate promotional effectiveness, were operationalised in this research.

3.5.1 Perceptions of offer value

The perceived value of the promotions was measured by using seven items based on the research of Palazon & Delgado-Ballester (2009), Alford & Biswas (2002) and Nusair et al. (2010). The items were all evaluated based on a seven-point Likert scale, anchored by “Strongly disagree” (point 1) and “Strongly agree” (point 7). The items that were utilised were the following: (1) I like this type of promotion; (2) I wish there were more promotions like this; (3) This promotion offer incites me to buy the product; (4) This promotion offer is of great value.

3.5.2 Purchase intention

The dependent variable purchase intention, the intention of consumers to buy the promoted product, was measured by using a seven-point Likert scale, ranging from “Very low” (point 1) to “Very high” (point 7). The measurement items were taken and adapted from the research of Palazon & Delgado-Ballester (2009) and Alford & Biswas (2002) and were as follows: (1) If I

(27)

were going to order food, the probability that I would consider buying making use of this offer is; (2) The likelihood that I would make use of this offer is.

3.5.3 Search intention

The measures that were used to evaluate search intention were adapted from Palazon & Delgado-Ballester (2009). Participants were asked to respond to three items: (1) Before making a purchase decision, I would look at other delivery service providers to check for similar promotions; (2) Before making a purchase decision, I would need to search for more information about alternative promotions; (3) Before making a purchase decision, I would look at other delivery service providers looking for a better promotion. These items were also evaluated on a seven-point Likert scale, anchored by “Strongly disagree” (point 1) and “Strongly agree” (point 7).

(28)

4. Results

In this chapter, the results of the research that was conducted are illustrated and analysed. First, general demographics of the sample are discussed after which the hypotheses are tested.

4.1 Sample

In total, 230 people filled out the online survey, where the aim was to find around 200 people for a usable sample (Field, 2013). Most of the respondents were approached and asked to fill out the questionnaire through social media. Of these 230 respondents, 31 people did not complete the survey. Their (partial) answers were excluded from this research. Of the 199 respondents that did complete the survey, 9 were found to have provided unreliable data, so these were also excluded from the research. This resulted in a total usable sample of N=190, which is a representative sample size for a total population of 13.677.409 (18 years and older) people (Centraal Bureau voor de Statistiek, 2017). Cell sizes for each of the groups can be found in table 4.1.

Benefit level Promotion type Discount framing

Discount Competition Absolute number Percentage

Low 36 31 33 31

Moderate 31 32 32 33

High 28 32 29 32

Table 4.1 Cell sizes

4.2 Descriptive statistics – demographics

First, the total of 190 usable respondents were analysed by their demographical statistics. Below, in table 4.2, the distribution by gender is illustrated. This table shows that the

(29)

distribution between men and woman was not equal, but that that women were overly represented among the respondents.

Gender Frequency Percent

Men 76 40

Women 114 60

Total 190 100.00

Table 4.2 Gender

In table 4.3 the average frequency of orders per week is illustrated. The table shows that, on average, most of the respondents get food delivered to their homes once per week. 11 respondents never get food delivered. These respondents were still included in this research as these are also people who are targeted by sales promotions.

Amount of

orders Frequency Percent

0 11 5.8 1 166 87,4 2 10 5,3 3 1 0,5 4 or more 2 1,1 Total 190 100.00

Table 4.3 Orders per week

As indicated before, the group of people who get food delivered regularly exists mostly of men and students. This is reflected in the age groups, where 56% is younger than 25 and almost 80% is younger than 30 years old. It is assumed that most of these respondents are either still studying or have just finished studying and are working or looking for jobs.

N Min Max Mean Standard

(30)

Age 190 18 62 27.15 7.51 Table 4.4 Age (in years)

Table 4.5 shows that most of the respondents have either finished or are following higher education. As most people who get food delivered on a regular basis are students, this indicates that the sample is representative for the population that is being studied.

Income Frequency Percent

High school 17 8.9 Trade/technical/vocational training 16 8.4 Some College 31 16.3 Bachelor’s degree 82 43.2 Master’s degree 43 22.6 PhD 1 0.5 Total 190 100.00 Table 4.5 Education

Table 4.6 indicates that the net monthly income of 75% of the respondents is less than €2000, which is an indication that most of the respondents are either students who work part-time jobs or have entered a starting position. This also is an indication that the sample is a good representation of the population that gets food delivered on a regular basis.

Net monthly income Frequency Percent

€0 – €500 37 19.5 €500 – €1000 46 24.2 €1000 – €2000 59 31.1 €2000 – €4000 42 22.1 €4000 – €6000 4 2.1 More than €6000 2 1.1 Total 190 100.00

(31)

4.3 Construct reliability

In order to measure the reliability of the questions that were used to form the constructs, Cronbach’s alphas were calculated. An overview of the internal constructs is given in table 4.7 and a more extensive overview is provided in Appendix 4.

Construct Cronbach’s alpha Items

Experiment 1

Perceived offer value .917 4

Purchase intention .912 2

Search intention .888 3

Experiment 2

Perceived offer value .937 4

Purchase intention .946 2

Search intention .941 3

Table 4.7 Construct reliability analysis results

All constructs were found to be internally consistent, without having to make any adjustments or delete questions. For each the constructs, deleting any of the questions would have resulted in a lower alpha. As the reliability test showed that the measures were internally consistent, the constructs were transformed into variables by taking the average score of the questions that belonged to each of the constructs.

4.4 Hypothesis testing

In order to test the hypotheses, for both the first and second experiment a two-way multivariate analysis of variance (MANOVA) was carried out, using Promotion Type (Discount,

Competition) and Benefit Level (Low, Moderate, High) as factors and using perceived offer

value, purchase intention and search intention as dependent variables for the first experiment. For the second experiment, the same dependent variables were used and Framing Type

(32)

4.4.1 Assumptions

Before conducting the MANOVA, there were several assumptions that needed to be checked (Field, 2013). The experiments were designed to meet the first two assumptions; the independence of the observations and random sampling with measurement at an interval level. As SPSS does not allow for testing multivariate normality, there was tested for univariate normality. The reasoning behind this is that, although it does not guarantee multivariate normality, for multivariate normality to exist, there needs to be univariate normality. The results indicate that this assumption is not met for multiple groups, but as cell sizes for all groups are at least 28, the MANOVA procedure will still be robust. As MANOVA tests are sensitive to multivariate outliers, there was tested for these types of outliers by determining the Mahalanobis Distance. As both of the experiments included three dependent variables, the critical Mahalanobis Distance for each of the experiments equals 16.27. For both experiments, there were two observations that exceeded this critical value, but not by an extreme amount, so they were included in the MANOVA.

There was then tested for equality between groups using Levene’s tests as there needs to be univariate equality between groups for there to be homogeneity of the covariance matrices (Field, 2013).

For the first experiment the Levene’s test indicated that the answers had a similar variance among the groups. The outcome for perceived offer value was F(5, 184) = 1.37, p

= .24, F(5, 184) = .48, it was p = .79 for purchase intention and F(5, 184) = .76, p = .58 for

search intention. For the second experiment the outcome for perceived offer value was F(5, 184) = 1.25, p = .29, it was F(5, 184) = .15, p = .98 for purchase intention and F(5, 184) = .62,

p = .68 for search intention. In order to test for homogeneity of variance-covariance matrices,

the Box's M Test of Equality of Covariance Matrices was used. For both experiments the Box’s M test was insignificant; Box’s M = 23.41, p = 0.84 for the first experiment and Box’s M =

(33)

42.72, p = 0.09 for the second experiment. This indicated that equality of covariance matrices between the groups could be accepted for both experiments.

After, there was tested for linear relationships between each possible pair of dependent variables for all combinations of groups of the independent variables and for multicollinearity between the dependent variables. For both experiments the results showed that there exists a linear relationship between each possible pair of dependent variables and that the dependent variables are correlated, but not too strongly.

Finally, in order for the control variables (food delivery order frequency and income) to be added to the model as covariates, there also had to be tested for linear relationships between these variables and the dependent variables. For not one pair of covariates and dependent variables was there found a linear relationship. This resulted in the model being tested without including the proposed control variables

4.4.2 Two-way MANOVA

For experiment one, there was not found a statistically significant multivariate interaction effect between promotion type and benefit level on the combination of dependent variables (perceived offer value, purchase intention and search intention), F(6, 364) = 1.458, p = .192; Wilks' Λ = .954, η2 = .023. The results also indicated that was no significant multivariate main effect for benefit level, F(6, 364) = .626, p = .710; Wilks' Λ = .980, η2 = .010, but there was found significant a multivariate main effect for the type of promotion (discount vs. competition), F(3, 182) = 40.966, p = .00; Wilks' Λ = .597, η2 = .403. In this experiment, there were not found any significant univariate interaction effects. The univariate effects for promotion type were significant for all dependent variables. This means that there are significant differences between offering discounts and offering competitions. Table 4.7 displays the means and standard deviations for experiment one.

(34)

Benefit level Dependent variables

Perceived value Purchase intention Search intention

M (SD) M (SD) M (SD) Discount Low 3.35 (1.20) 3.50 (1.25) 3.18 (1.69) Moderate 3.82 (1.37) 3.74 (1.56) 3.02 (1.90) High 4.06 (1.20) 4.00 (1.41) 3.25 (1.63) Competition Low 2.02 (1.55) 2.18 (1.31) 2.27 (1.74) Moderate 1.47 (1.50) 1.69 (1.25) 1.80 (1.54) High 1.50 (1.23) 2.11 (1.40) 2.51 (1.65)

Table 4.7 Experiment one means and standard deviations

The differences in means can be examined in figures 4.1 through 4.3. In figure 4.1 the perceived offer value means across different benefit levels are illustrated. From this graph, it can be concluded that, as the difference in means was found to be significant and perceived offer value means were higher for discounts than for competitions across all benefit levels, hypothesis 1a could be accepted; Perceived offer value is higher for discounts than for competitions.

Figure 4.1 Interaction plot of benefit levels and promotion type for perceived offer value

1.00 2.00 3.00 4.00 5.00 L O W M O D E R A T E H I G H PE R C E IV E D O FFE R V A L U E Discount Competition

(35)

Figure 4.2. Interaction plot of benefit levels and promotion type for purchase intention

After looking at figure 4.2, which depicts purchase intention means across different benefit levels, the same could be concluded for hypothesis 1b; Purchase intention is higher for monetary promotions than for non-monetary promotions.

Figure 4.3. Interaction plot of benefit levels and promotion type for search intention

The difference between search intention means for discounts and competitions was found to be significant, although not in the way that was predicted; Search intention seemed to be higher

1.00 2.00 3.00 4.00 5.00 L O W M O D E R A T E H I G H PU R C H A SE I N T E N T IO N Discount Competition 1.00 2.00 3.00 4.00 5.00 L O W M O D E R A T E H I G H SE A R C H I N T E N T IO N Discount Competition

(36)

for discounts than for competitions, which resulted in the rejection of hypothesis 1c, which stated that search intention is lower for discounts than for competitions. This can be derived from figure 4.3.

Also, a significant interaction effect was found between benefit level and promotion type on perceived offer value, F(2,184) = 3.850, p < .05, η2 = .040, but the interaction effects were insignificant for purchase intention and search intention.

In order to test for hypothesis 3a and 3b, independent samples t-tests were conducted. The first t-test was conducted in order to determine whether or not the means for perceived offer value, purchase intention and search intention were similar at moderate benefit levels for both discounts and competitions. The results indicated that mean differences between discounts and competitions were significant for all three dependent variables. This resulted in the rejection of hypothesis 3a; the perceived value, purchase intention and search intention of price discounts and competitions are not similar for moderate promotional benefit levels.

Benefit level and promotion type

Dependent variables

Perceived value Purchase _intention _intentionSearch

Low Discount M (SD) 3.35 (1.20) 3.50 (1.25) 3.18 (1.69) Low Competition M (SD) 2.02 (1.55) 2.18 (1.31) 2.27 (1.74) t 3.886 4.231 2.158 sig. < .05 < .05 < .05 Moderate Discount M (SD) 3.82 (1.37) 3.74 (1.56) 3.02 (1.90) Moderate Competition M (SD) 1.47 (1.50) 1.69 (1.25) 1.80 (1.54) t 6.486 5.781 2.805 sig. < .05 < .05 < .05 High Discount M (SD) 4.06 (1.20) 4.00 (1.41) 3.25 (1.63) High Competition M (SD) 1.50 (1.23) 2.11 (1.40) 2.51 (1.65) t 8.164 5.203 1.739 sig. < .05 < .05 .087

(37)

This also meant that hypothesis 3b had to be accepted; The perceived offer value, purchase intention and search intention are not similar for price discounts and competitions for both low and high benefit levels. The results of the t-tests are illustrated in table 4.8.

Experiment two, which measured the effects of framing price discounts, formatted either as a percentage-off discount or as a cents-off discount, indicated no significant multivariate interaction effect, F(6, 364) = .434, p = .856; Wilks' Λ = .986, η2 = .007. The main multivariate effect for framing promotions was also found to be insignificant with F(3, 182) = 1.561, p = .2; Wilks' Λ = .975, η2 = .025, but in this experiment, in contrast to experiment one, the main multivariate effect for benefit level was found to be significant, F(6, 364) = .2.727, p = .013; Wilks' Λ = .916, η2 = .043.

Benefit level Dependent variables

Perceived value Purchase intention Search intention

M (SD) M (SD) M (SD) Cents-off Low 3.45 (1.44) 3.39 (1.23) 3.28 (1.81) Moderate 4.28 (1.46) 3.93 (1.44) 2.94 (1.73) High 4.41 (1.47) 4.40 (1.35) 2.98 (1.81) Percentage-off Low 3.96 (1.54) 3.68 (1.32) 2.74 (1.83) Moderate 4.27 (1.01) 3.79 (1.19) 2.72 (1.55) High 4.55 (1.33) 4.23 (1.39) 2.90 (1.86)

Table 4.9 Experiment two means and standard deviations

As in experiment one, in this experiment there were not found any univariate significant interaction effects but there were found significant main univariate effects of promotional

(38)

benefit level on perceived offer value, F(2, 184) = 5.338, p < 0.05, η2 = .055 and purchase intention, F(2, 184) = 5.473, p < 0.05, η2 = .056. The fact that there were not found any significant effects for promotional framing, resulted in the rejection of hypothesis 2a, 2b and 2c. The means and standard deviations for experiment two are illustrated in table 4.9.

For post-hoc analysis, Tukey tests were executed in order to determine the nature of the differences in the means for promotional benefit levels that were found to be significant. The Tukey tests indicated that the low benefit levels differed significantly from the moderate and high benefit levels for perceived offer value at p < .05. For purchase intention, the Tukey test indicated that only the low benefit level and high benefit level differed significantly p < .05.

Hypotheses 4a and 4b were tested by conducting independent samples t-tests. As predicted, the results of the t-test indicated that the means for moderate discounts formatted as a cents-off and as a percentage-off were not significantly different; the perceived value, purchase intention and search intention of discounts presented in dollar terms and discounts presented in percentage terms are similar for moderate benefit levels. Hence, hypothesis 4a was accepted.

Benefit level and format

Dependent variables

Perceived value Purchase _intention Search _intention

Moderate Cents-off M (SD) 4.28 (1.46) 3.93 (1.44) 2.94 (1.73) Moderate Percentage-off M (SD) 4.27 (1.01) 3.79 (1.19) 2.72 (1.55) t .027 .459 .541 sig. .978 .648 .591 High Cents-off M (SD) 4.41 (1.47) 4.40 (1.35) 2.98 (1.81) High Percentage-off M (SD) 4.55 (1.33) 4.23 (1.39) 2.90 (1.86) t -.372 .461 .150 sig. .711 .647 .881

(39)

The t-tests resulted in the rejection of Hypothesis 4b, as there was not found a significant difference in means for percentage term discounts and dollar term discounts with a high benefit level; the perceived value and purchase intention for price discounts presented in percentage terms are similar to discounts presented in dollar terms for sales promotions with a high benefit level. The results of the t-tests are depicted in table 4.10.

(40)

5. Conclusion

In this research, it was studied how the sales promotion efforts utilised by food delivery services in the Netherlands, affect consumer behaviour. Two different experiments were conducted in order to determine the effects of differences in types of promotions and differences in the framing of promotions in the context of food delivery in the Netherlands. In the first experiment, the difference between utilising either a price discount or a competition as a promotional method, across different benefit levels, was investigated by using a 2 x 3 between subjects design. The second experiment, designed to investigate the differences in framing promotions across different benefit levels, also employed a 2 x 3 between subjects design. For both experiments, three indicators of promotional effectiveness on consumer behaviour were used as dependent variables: perceived offer value, purchase intention and search intention.

5.1 Conclusion and discussion

First, several types of promotions were identified, after which the promotions that are used most by Dutch food delivery services, discounts and competitions, were used as proxies for monetary and non-monetary promotions. The effectiveness of both promotions was studied and compared.

One of the most important findings of this research is that using price discounts as a promotional method seems to be significantly more effective than using competitions. As predicted, consumers value discounts more and demonstrate a higher purchase intention for discounts than they do for competitions. Consumers might prefer a direct benefit in the form of an immediate discount, rather than ordering food with the possibility of winning a prize. As is explained by Shi et al. (2005), discounts provide consumers with transaction utility and consumers find them relatively easy to understand.

(41)

It was, however, also predicted that search intention would be lower for price discounts than for competitions, but the results indicate the opposite; search intention was higher for price discounts than for competitions. As search intention also did not differ across benefit levels, this is an indication that in the context of food delivery, consumers may always feel that there is a better offer somewhere else.

It was also predicted that the differences between price discounts and competitions would differ across different discount levels (benefit levels). The results indicate that there are no significant differences across benefit levels for perceived value, purchase intention and search intention. The results do indicate a steady increase in perceived offer value and purchase intention across benefit levels for discounts, with these values being lowest for low benefit levels and highest for high benefit levels. But for competitions these values are lowest at moderate benefit levels. This might be an indication that, as Hardesty & Bearden (2003) stated, there is a relatively large uncertainty at moderate benefit levels, even more so for competitions than for discounts, which can result in increased consumer miscomprehension and consumers being more sceptical about the promotion, which results in lower perception of the offer value and lower purchase intention.

As consumers were thought to process information more elaborately for promotions at moderate benefit levels, it was predicted that promotional effectiveness would be similar for both discounts and competitions at this level. The results indicate that at moderate benefit levels, discounts are still more effective than competitions, as is the case with low benefit levels and high benefit levels. Consumers seem to be more certain of the value of discounts and the immediate benefit it offers them than they are of competitions.

As was found in previous studies that have investigated promotional effectiveness in the context of several different markets, in this research it was also predicted that there would be a difference in the framing of discounts. Discounts are usually framed either in percentage

(42)

terms or in absolute value terms. It was predicted that, as percentage-off discounts seem more attractive for lower priced items than cents-off discounts, perceived value and purchase intention would be higher and search intention would be lower for percentage-off discounts than for cents-off discounts. Results from the current research indicate that differences in framing promotions do not affect perceptions of the offer value. Differences in discount framing also do not influence the intention to order food, nor do they influence consumers’ intentions to search for other food delivery options. The results however did show differences in the effects different benefit levels have on purchase intention and perceived offer value. For both framing options sales promotions with low benefit levels consumers seem to perceive the offered value to be significantly lower than promotions at moderate and high benefit levels. Purchase intention is also lower for low benefit levels than for high benefit levels.

As with the different types of promotions, with regard to the framing of promotions, it was predicted that promotional effectiveness would be similar for both framing types for moderate benefit levels, which was found to be true. It was also expected that for high benefits levels there would be a difference in effectiveness between percentage-off and cents-off discounts, which was found not to be true. A possible explanation for there not being a difference for high benefit levels, is that percentage-off discounts require additional processing in comparison with cents-off discounts (DelVecchio et al., 2007). As for moderate benefit levels, consumers evaluate the promoted deal more extensively since there is uncertainty regarding the deal, they might do the same at high benefit levels. Not because there is uncertainty but because percentage-off discounts are more difficult to processes. So, consumers might calculate the exact dollar amount of the discount when faced with a percentage-off discount, which results in them equating similar discounts that are framed differently.

Remarkable are the results regarding search intention. Although search intention was expected to be lower for discounts than or competitions, this was not the case. Also, search

(43)

intention was expected to decrease as benefit level increased but this was not the case in both experiments, as when benefit levels increase, promotions will become more attractive and consumers will probably be less inclined to search for alternative promotions (Palazon & Delgado-Ballester, 2009). A possible explanation for this, as explained before, is that when it comes to making use of food delivery services in the Netherlands, as the different services promote sales promotions so often, consumers might almost always think that there is probably a better deal somewhere else.

5.2 Managerial implications

Given the amount of resources that is spent on promotional activities, understanding when to use certain promotional activities remains important. The results of this research suggest that Dutch food delivery services will benefit from using discounts over competitions when they are trying to attract customers. But, as search intention stays high, regardless of benefit level, food delivery services should be aware of the benefits their promotions provide with regard to the competition in order to actually convince consumers to make a purchase. As promotions using discounts will usually cost a food delivery service more than utilising competitions, it is important that the right promotion type is used for the right end.

This study also suggests that for discounts, it makes no difference whether delivery services frame discounts in percentage terms or in cents terms as there was found that consumers do not value the two types of framing differently.

5.3 Limitations and suggestions for future research

The current research has tried to provide more insights in consumers responses to sales promotions used by Dutch food delivery services. This research does, however, have some limitations. Firstly, discounts and competitions were used as proxies for monetary and

(44)

non-monetary promotions, as they are the most used types by Dutch food delivery services. But, given the various types of promotional tools that are available to businesses, it might not be possible to make generalisations with regard to monetary and non-monetary promotions, using only price discounts and competitions. So, it could be an option to include the full range of possible promotions in future research. Including more promotional tools could provide more dependable insight into the particular effects of certain promotions and the strength of these effects. Although several demographical statistics were included, more could be added to gain more insight into how responses to food delivery service promotions differ across differ segments of the population.

As multiple studies point out, there are several variables that might influence how consumers respond to sales promotions, such as price consciousness and sales proneness and with regard to competitions, belief in luck (Palazon & Delgado‐Ballester, 2009). These variables might strongly moderate the relationship between the promotional activities and promotional effectiveness, so including these provides a great opportunity for future research.

(45)

6. References

Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behaviour.

Alford, B. L., & Biswas, A. (2002). The effects of discount level, price consciousness and sale proneness on consumers' price perception and behavioral intention. Journal of Business

research, 55(9), 775-783.

Bawa, K., & Shoemaker, R. (2004). The effects of free sample promotions on incremental brand sales. Marketing Science, 23(3), 345-363.

Buil, I., De Chernatony, L., & Martínez, E. (2013). Examining the role of advertising and sales promotions in brand equity creation. Journal of Business Research, 66(1), 115-122.

Büttner, O. B., Florack, A., & Göritz, A. S. (2015). How shopping orientation influences the effectiveness of monetary and nonmonetary promotions. European Journal of

Marketing, 49(1/2), 170-189.

Centraal Bureau voor de Statistiek. Bevolking; geslacht, leeftijd en burgerlijke staat, 1 januari [Dataset]. Retrieved July 3, 2017, from

http://statline.cbs.nl/Statweb/publication/?DM=SLNL&PA=7461bev&D1=0&D2=0 &D3=19-133&D4=l&HDR=G3%2cT&STB=G1%2cG2&VW=T

(46)

Chandon, P., Wansink, B., & Laurent, G. (2000). A benefit congruency framework of sales promotion effectiveness. Journal of marketing, 64(4), 65-81.

Chen, S. F. S., Monroe, K. B., & Lou, Y. C. (1998). The effects of framing price promotion messages on consumers' perceptions and purchase intentions. Journal of

retailing, 74(3), 353-372.

Chou, P. F., & Lu, C. S. (2009). Assessing service quality, switching costs and customer loyalty in home-delivery services in Taiwan. Transport Reviews, 29(6), 741-758.

Currim, I. S., & Schneider, L. G. (1991). A taxonomy of consumer purchase strategies in a promotion intensive environment. Marketing Science, 10(2), 91-110.

Darke, P. R., & Chung, C. M. (2005). Effects of pricing and promotion on consumer perceptions: it depends on how you frame it. Journal of Retailing, 81(1), 35-47.

DelVecchio, D., Krishnan, H. S., & Smith, D. C. (2007). Cents or percent? The effects of promotion framing on price expectations and choice. Journal of Marketing, 71(3), 158-170.

Dodds, W. B., Monroe, K. B., & Grewal, D. (1991). Effects of price, brand, and store information on buyers' product evaluations. Journal of marketing research, 307-319.

How effective are the sales promotions utilised by Dutch food delivery services? : the effects of sales promotions on consumer behaviour