• No results found

An Analysis of the effects of the

N/A
N/A
Protected

Academic year: 2021

Share "An Analysis of the effects of the "

Copied!
72
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Ikea

An Analysis of the effects of the

“woonsdagaanbieding” on the category Tv Benches and Media furniture

Master thesis

Rijksuniversiteit Groningen Faculty of Economics

H. Boschma July 2005

First supervisor: Dr. J. Wieringa Second supervisor: Dr. K.J. Alsem

(2)

Measuring the effects of Promotions at Ikea

An Analysis of the effects of the

“woonsdagaanbieding” on the category Tv Benches and Media furniture

H. Boschma

Rijksuniversiteit Groningen Faculty of Economics

9700 AV Groningen

The Netherlands

July 2005

(3)

Summary

In the past few years the retail market has changed drastically for Ikea. The retail expenditures by consumers have been declining for several years. Nowadays, consumers have less money to spend on consumer goods. To anticipate on the lower consumer’s budget, Ikea spends more time and effort to strengthen its low price image. The most important vehicle to communicate the low price perception is the “Woonsdag”. On this special day (every last Wednesday of the month) a wide range of activities and promotions are organised to attract customers. The results of the”Woonsdag” 2004 are very promising, some indexes showed a 150 % increase in turnover and paying customers. However, the effects of this and other promotions are not well known to the management.

To estimate the effects of promotions on the sales of the Category TV Benches & Media furniture and its main products a case study (the “Woonsdag” of 28th April) is used. The model building process resulted into three models that estimate the promotional effects at a daily level; one model that deals with the Category sales of the TV Benches & Media furniture, a model that describes the total sales of “Woonsdagaanbieding” and a model that describes the baseline sales of the “Woonsdagaanbieding”. The three models are estimated with the use of daily data. Most studies use weekly or data with longer time intervals, and these are only dealing with the effects of promotions on fast moving consumer goods, whereas this study focuses on durable consumer goods.

The main conclusion of this study is that not all promotions have the same impact on sales.

Some promotions have a larger impact than others. A price discount is the most effective way to stimulating short-term sales. The “Woonsdagaanbieding” resulted into an increase in its own sales and the Category sales as well. It resulted also into a negative lead effect and a large positive lagged effect. After the promotion the sales increased, even its baseline sales increased. The negative lead effect is probably caused by featuring the discount. A feature that advertises a discount leads to anticipatory effects by consumers. Features of other

products that were not under study did not lead to cross brand effects, and had no effect on the category sales. The employment of display resulted in an increase with 30 % of own product sales. Surprisingly, the lagged effects were larger, it was expected that the lagged effects would be smaller than the display effect. The introduction of the new catalogue resulted in a boost in sales of nearly all the products under study and the Category sales. Its lagged effects are even stronger because it takes time for consumers to plan a visit to the store. Brochures which are distributed in the first week of every month did not have any effect on the sales of the products under study. Out of stock caused a drop in sales of nearly 50 % of own product sales. The Category sales decreased with nearly 20 %. Holidays have an important influence on the sales. The magnitude of its effect on sales depends on the type of holiday, Christmas holidays have the largest effect on Sales. In contrast to the general opinion within Ikea, the weather did not seem to influence sales.

(4)

Acknowledgements

This thesis is written as a part of my completion of the study MBA Marketing, at the Faculty Economics, University of Groningen, the Netherlands. This research is the result of a research assignment at Ikea Groningen in the period between April 2004 and March 2005.

The supervision of this process was in the hands of Dr. J. Wieringa and Dr. K.J. Alsem of the University of Groningen. I owe them my thanks and gratitude for their critical remarks and suggestions. I would also like to thank the management team of Ikea Groningen who made this opportunity possible. Especially, I owe thanks to the department SCA who delivered the data and the department of controlling who collected the vouchers. And I would like to thank all my colleagues at Ikea who helped me with useful suggestions and insights. Last but not least I would like to thank drs. H.J. Slomp for his time and effort in helping me with the difficult process of writing this thesis in English. Furthermore I would like to thank my cat Gijs who persistently wanted to help me, by putting his paws on the keyboard and printer, which was sometimes quite inconvenient.

Groningen, July 2005 Henk Boschma

(5)

Contents

Chapter 1 Introduction 1

§ 1.1 Object of study IKEA 1

§ 1.2 Research approach 2

§ 1.3 Research set up 3

§ 1.4 Variables researched 5

§ 1.5 Scientific relevance 5

Chapter 2 Promotions 6

§ 2.1 Promotions 7

§ 2.2 Types of sales promotions 8

§ 2.3 Target groups of sales promotions 8

§ 2.4 The rise of Sales Promotions 9

Chapter 3 How promotions affect Sales 10

§ 3.1 Decomposition of Sales Promotions 10

§ 3.2 Deal Effect Curve 13

§ 3.3 Empirical Findings 14

Chapter 4 Sales Analysis 17

§ 4.1 Sales analysis of the Category TV Benches 17

§ 4.2 Category Sales of the TV Benches 19

§ 4.3 Sales of the Ikea ps category 21

§ 4.4 Sales of the Oppli category 23

§ 4.5 Sales of the Stormark 24

§ 4.6 Sales of the Timrå 25

§ 4.7 Sales of the “Other” Category 25

Chapter 5 Introduction into Model building 27

§5.1 Steps in Model building 27

§5.2 Typology of models 27

§5.3 Model building criteria 28

Chapter 6 Building a model 30

§ 6.1 Specification of the models 30

§ 6.2 Parameterisation of the models 34

§ 6.2.1 Estimation Procedures 34

§ 6.2.2 Model testing 34

§ 6.2.3 Estimation of the parameters 35

§ 6.3 Classification of the models 41

§ 6.4 Drawbacks of OLS 41

§ 6.4.1 Heteroskedasticity 43

(6)

§ 6.4.2 Serial Correlation or Autocorrelation 44

§ 6.4.3 Multicolinearity 44

§ 6.5 Validation 46

6.5.1 Face validity 46

§ 6.5.2 Statistical Validity 49

§ 6.5.3 Predictive validity 50

§6.6 Implications of the models 50

Chapter 7 Conclusions 52 Appendix 57 Appendix I Market share of the products under study 57 Appendix II market share of all the products 57 Appendix III Total units sold of all the products under study 58 Appendix IV Total turnover per product under study 58 Appendix V Model I original SPSS outcomes 59 Appendix VI Model I SPSS model outcomes 62

Appendix VII Model II original SPSS outcomes 67 Appendix VIII Model II SPSS model outcomes 64

Appendix IX Model III original SPSS outcomes 69

Appendix X Model III SPSS model outcomes 72 Appendix XI Durban Watson Table 74

References 75

(7)

Chapter 1 Introduction

“May you live in interesting times” is an old Chinese saying and indeed most marketers live in interesting times; especially those who work in the retail business. Last year, the business underwent a dramatic change. Where once the full service stores ruled, now the price fighters dominate. Today, marketing managers are focusing more than ever on the capital “P” of Price.

This sudden change in interest is a result of changing consumer behavior. Due to economic depression, consumers are more and more looking for bargains. In a way, they have become more price sensitive. For some retailers it is imperative to have a low price perception in the minds of the consumers. To maintain traffic in stores, retailers drop their prices at a

spectacular rate. Besides lowering prices, retailers demand contribution from the

manufacturers to finance promotions. Although, price cuts and promotions are frequently used it is remarkable that its effects are often unknown to retailers. Van Heerde (1999) estimates that less than one percent of all sales promotions in the Netherlands are evaluated.

Abrahams and Lodish (1993) find in their study that only one out of ten promotions is profitable. Due to involvement of high costs it is important for marketing managers to know the effects of promotions. Knowing the effects of promotions, is knowing whether to support a price cut with an feature or display or both or leave it unsupported or not to do price

promotions at all. Choices like these are also made in the furniture sector. This study is about the effects of promotions at Ikea Groningen.

1.1 Object of study: IKEA

Ikea was founded nearly six decades ago in Sweden. Since its founding, it is highly successful selling furniture to its customers. Today, Ikea is active in 44 countries and employs nearly 76.000 people. The Ikea stores are located in nearly 32 countries and regions. Last financial year 2004 the sales of Ikea Group totalled 12.8 billion euro. Europe is its most prominent market which is accountable for nearly 81% of the total turnover. Ikea owns in the Netherlands 11 stores; these are located in the city’s of Amsterdam, Barendrecht, Breda, Delft, Duiven, Eindhoven, Groningen, Haarlem, Hengelo, Heerlen, Utrecht, and Sliedrecht (see figure 1).

The smallest store, with 13.800 m2, is located in Groningen (www.Ikea.nl). Despite its small size, the store’s market area is relatively large compared to other stores. The market area covers nearly all three Northern provinces and parts of northern Germany

City Founding Floor area in m2

Amsterdam 1982 32 700

Duiven 1983 20 300

Breda 2003 26 000

Eindhoven 1992 28 600

Goningen 1997 13 800

Haarlem 2005 28 500

Heerlen 1994 19 600

Hengelo 2002 23 200

Barendrecht 2001 27 200

Sliedrecht 1978 14 600

Utrecht 1996 17 500

Table 1.1 Ikea stores in the Netherlands (www.Ikea.nl) Figure 1 Ikea stores (www.Ikea.nl)

(8)

The total turnover at the Groningen location was last financial year 52 million euro and the store was visited by more than 900.000 people. Due to space limitations the store is to be relocated and expanded to 29.500 m2, this coming summer.

1.2 Research approach

The success of an Ikea store depends on its ability to generate high traffic volume, high conversion rates and a large turnover per customer. A conversion rate is defined as the percentage of visitors that is induced to a purchase at the store. Turnover per customer is the average spending of each customer at the register. A store can be very successful in attracting large crowds to the store, but if conversion rates and turnover per customer are low, than the store does not make a lot of profit. An important aspect of the Ikea philosophy is the low price image in the minds of the consumers. Low prices attract customers. Over the years “Ikea” as a brand has proven to be very successful in establishing a low price image in the minds of the consumer. Hence, consumers often refer to the high quality and the low prices.

These prices are communicated throughout the stores, by means of streamers (large banners), price tags and displays, brochures, radio and TV commercials, the internet, and most

important the catalogue. Besides the communication instruments, themes or events are organised in the stores, to entertain the visitors. The most important theme of Ikea is the

“Woonsdag”. This day is dedicated to the founding of Ikea in the Netherlands. On this special day (every last Wednesday of the month) Ikea organises a wide range of activities to attract customers. The main traffic builder is the “Woonsdagaanbieding”. The “Woonsdag-

aanbiedingen” are popular products that are discounted; the discount can amount to 65 % of its original price. The “Woonsdagen” have become very popular among shoppers. Only after a few hours, the promoted product is sold out. The results of the”Woonsdag” 2004 are very promising, some indexes showed an 150 % increase in turnover and paying customers. The increase of turnover did not only depend on the “Woonsdagaanbieding”, but it is also

depended on the increase in additional spending by customers. Nearly two third of the visitors bought a product that was not planned. The “Woonsdag” resulted not only in high traffic volumes but also resulted in delayed purchases of the promoted product (Ikea (a) 2004).

The aim of this investigation is to study the effects of promotions on the category TV benches

& Media furniture by using a case study, the “Woonsdag” of 28 April 2004. The aim of the study is formulated as follows:

“To develop a marketing model to estimate the effects of promotions on the category TV benches & Media furniture and its main products within this specific category at Ikea Groningen”

The marketing manager of Ikea has a vast array of communication marketing instruments at its disposal. Within the stores all different types of promotions are used, but its effects on sales are rarely know. The deployment of communication instruments in the stores or price discounts is mainly based on experience rather than hard facts. For instance, the full effect of the “Woonsdagaanbieding” is often not known, only when the promotion is over the

management knows whether it had enough of the promoted product in stock or not. The effects on the long term are also not known, for instance substitution or category expansion effects are often neglected. The effects of other promotions, such as a display or

advertisement are estimated, based on experience. However, measuring the promotional effects rather than to estimate its effects by experience, would greatly improve the knowledge

(9)

on sales promotion. This would likewise increase the efficiency of the deployment of promotions and reduces the cost of it.

The research question is formulated as follows:

“What effects do promotions have on the sales of the category TV benches and Media furniture and its main products within this category at IKEA Groningen?”

To answer the research question several questions need to be answered. These questions are stated below:

1. What are promotions?

There is often confusion about the meaning of promotions. It sometimes refers to the last P of the marketing mix or it is referring to the marketing communication instrument “sales promotion”. The difference between these two is discussed in Chapter 2.

2. What are the effects of promotions on sales?

There are many published articles on the effects of promotions on sales. The larger part of published articles deals with the effects of sales promotions on fast moving consumers goods. However, this research deals with the effect of promotions on durable consumer goods. It is therefore interesting if the effects of promotions on fast moving goods are the same as on durables. The effects of promotions on fast moving consumer goods are discussed in Chapter 3.

3. What is the decomposition of sales promotions?

The decomposition of sales promotions is important; it reveals the sources of the increase in sales. There are several decompositions used and these are discussed in Chapter 3.

4. How do the sales of the products under study develop over time?

Chapter 4 discusses the sales development of the products under study. This chapter gives possible explanations for the spikes and dips in sales.

1.3 Research set up

The store level dataset that is investigated consist of daily data on sales of the category TV &

Media furniture and promotions that are held during the studied period. This period runs from 25 March 2004 to 3 February 2005, thus it contains 270 observations; there are five missing values due to errors.

The category under study consists of 22 different items. Ikea introduced eight new products, in this period (Ikea/ps Silver, Stormark, Oppli Silver, Benno TV Bench 120 x 60, Timrå TV meubel and the Klubbo black brown, the Kaxås TV Bench), and six products were withdrawn (Benno Tv bench 118 x 50 birch veneer, black brown veneer, and beech veneer and Kaxås 50

(10)

x 42 Foil Dark metallic and Kaxås 97 x 43 Foil Dark metallic). A total of 11,502 units, with a sales value of € 561,380, were sold.

(11)

An overview of the products under study is presented in table 1.2. The products are divided into group 1 and 2. Group 1 consists of TV benches that are frequently supported by

promotions and/or belong to the Ikea concept of “Aces and Kings”. Aces are responsible for a large part of the turnover and/or have a high gross margin and/or are extremely low priced to compete with competing stores. “Kings” have the potential to become “Aces” this implies they receive a lot of attention from the sales force (see also table 1.3). Products from group 2 do not meet the criteria that are mentioned above. Products from group 1 are considered to be more important than products from group 2.This study will focus on products from group 1 for two reasons; they have higher sales and are frequently supported with marketing

communication. Products that belong to group 2 are in this research aggregated and are addressed as “Other”.

Group 1 Group 2

IKEA/PS KAST 119X63 RD BENNO TV-MEU 118X50 BER FI IKEA/PS KAST 119X63 W BENNO TV-MEU 118X50 BEU FI IKEA/PS KAST 119X63 ZILVERKL BENNO TV-MEU 118X50 MBR OPPLI S TV-MEU&WIE 90X59X40 ZILVERK BENNO TV-MEU 120X60 BER FI OPPLI TV-MEU&WIE 150X50X40 BER FI ILEN ROLTAF/TV-MEU 50X50X56 FOL BL OPPLI TV-MEU&WIE 60X50X40 BER FI ILEN ROLTAFEL/TV-MEU 50X50X56 FOL W STORMARK TV-MEU 101X48 ZILVER/GLAS IVAR TV-MEU&WIE 104X40 ONB GREN TIMRÅ TV-MEU&WIE 117X59X50 BERKPATR KAXÅS TV-MEU 97X43 FOL DGS METALLIC

KAXÅS TV-MEU 97X60 FOL DGS METALLIC

KAXÅS TV-MEU&WIE 50X42 FOL DGS META

KIVIK TV-MEU 60X45X36 BEUPAT/FOL

KLUBBO SALONTAF/TVMEU 97X60X27 ZWBR

KLUBBO SALONTAFEL/TV-MEU 97X60X27 W

TUNSTA TV-MEU&WIE 81X54 ZILVERKL

Table 1.2 Products researched

Nearly 520 Ikea ps units were sold on the “Woonsdag” of 28th April. However, the supply of Ikea ps present at Ikea Groningen was not large enough to satisfy the demand. Costumers arriving when the product was sold out received a voucher. The vouchers could be used to purchase a red or white Ikea ps with discount. More than 900 vouchers were distributed. Due to administrative errors vouchers for a red Ikea ps were also used to purchase a white Ikea ps, and vice versa. To reduce measurements errors in estimating the effects of the “Woonsdag” of 28th April, the Ikea ps red and white from group 1 are aggregated, and they are referred in this study as Ikea ps.

1 Oppli Silver is referring to the Oppli 90 x 59 x 40 Silver

2 Oppli small is referring to the Oppli 60 x 50 x 40 Birch Veneer

3 Oppli large is referring to the Oppli 150x 50x40 Birch Veneer

High turnover ace Gross margin Ace King

Stormark Ikea ps Red Oppli silver1

Ikea ps silver Oppli small2 Oppli large3

Tabel 1.3 Aces and Kings (Ikea (b), 2004

(12)

1.4 Variables researched

The studied variables are listed in table 1.4. The variables can either have positive or negative effects on sales. Additionally, some variables have lead and lagged effects. If a variable does not have any effect on the sales it will be deleted from analysis. More detailed information is found in chapter 6.

1.5 Scientific relevance

In comparison to other studies, this study is unique in two ways. First it studies daily data to measure the effects of promotions on durable consumption goods. Most promotions only last a week or several days; this study investigates a promotion that lasts only one day. As a consequence, the length of the time interval (month, week or day), has a direct effect on the outcomes of any research concerning the effects of promotions. Research suggests that the usage of small data intervals (e.g. daily or weekly) improves the model outcomes. Besides a better fit, the sample size and the variation within the data increase (Blatberg and Neslin 1990); (Leeflang et al. 2000). In marketing literature a few studies are known that use daily data; Kondo and Kitagawa (2000) and van Dijk (2004). Kondo and Kitagawa (2000)

modelled category sales for a single store. The study of van Dijk (2004) concerns modelling category dependencies. Second, the effects of promotions on durables are investigated.

Publications of the effects of promotions on durables using daily data are a rarity in marketing literature. Because of this uncommonness it is interesting if the findings of this study can be compared to other findings of studies dealing with the effects of promotions on fast moving consumer goods.

Display Events

Feature

Type of Holiday Out of stock Price Rain

Streamer (large sales sign) Sunshine

Sales Period Temperature

Table 1.4 Researched variables

(13)

Chapter 2 Promotions

This chapter introduces the concept of Promotions. The first paragraph explains the difference between promotion and “sales promotion”. The types of promotions are discussed in

paragraph 2.2 and the last paragraph deals with promotional trends in the marketing world.

2.1 Promotions

The word “promotion” is frequently used, but there remains much confusion about its definition. Most confusion arises when English texts are translated into Dutch. A Promotion can be referring to the last “P” of the marketing mix or to the marketing mix instrument of sales promotion. Promotions can be used in a broad- or narrow context. If the word

“promotion” is used in a broad context then it is referring to the last “P” of the marketing mix.

Promotion used in this way, is a marketing communication instrument that belongs to the class of “verkoopbevordering”. The main purpose of “Verkoopbevordering” or the last “P” of the marketing mix is to inform, persuade, or remind buyers about the characteristics or

qualities of a brand on the short and long term. Promotions used in a small context are referring to sales promotions (Leeflang 1994).

Sales promotion belongs to the class of the action marketing communication mix. The action marketing communication mix and thematic marketing communication mix form together the communication mix (Floor en Van Raaij 2002) or marketing communication instruments (Leeflang, 1994). For more information, see also figure 2.1.

Thematic marketing communication tends to influence the knowledge and/or attitude of the consumer on the long term. Marketing instruments that are used to influence knowledge and/or are attitude are: advertising, sponsoring, publicity, cooperate design, and public relations.

The main purpose of action marketing communication or sales promotions is to influence the purchase behavior of consumers. Promotion or a sales promotion used in this context is a temporarily change in the price/value ratio.

The focus of this investigation lies on promotions used in a narrow context. Sales promotion is often defined as an improvement of the price/value ratio. The following definition of sales promotion is frequently used. Sales Promotion is:

“ A temporarily improvement of the price/value ratio of products or services to generate a short-term increase in sales” (Gussekloo and Strating 1985, p16).

This definition, however, excludes several types of marketing instruments used at Ikea.

Promotions at Ikea are often more than just an improvement of the price/value ratio; it also includes features, displays and other communication instruments. The definition of Kotler is more flexible and therefore more suitable for this study. Kotler defines sales promotion as:

“Sales promotion consists of a diverse collection of incentive tools, mostly short-term, designed to stimulate quicker and/or greater purchase of a particular product by consumers or the “trade” (Kotler 1988, p 645)

(14)

2.2 Types of sales promotions

Sales promotions can be divided into classic and thematic promotion. Classic sales

promotions are a marketing tool to increase temporarily sales (Floor and Van Raaij 2002). In this respect promotions can be regarded as a temporarily change in the four P’s of marketing.

Price is sometimes lowered, the package is changed and distribution is adjusted. If the promotional effort does not lead to higher sales, it is unsuccessful (Leeflang 1994).

Promotions that are used to enhance the meaning of a brand refer to thematic sales

promotions. The purpose of thematic sales promotion is two folded. First it tries to generate incremental sales. Second it is used to influence the perception of a brand. A promotion can support a low price image or it can be used to strengthen the relationship between consumers and a brand (Floor and van Raaij 2004). Promotions held at Ikea are often thematic of nature.

Establishing a low price perception is its most important goal and its next goal is to generate incremental sales. The larger parts of the promoted products at Ikea are loss leaders. Loss leaders are products sold below production price (Foekens 1995).

For a long time, Thematic- and Action marketing communication were considered to be competitive instruments (Leeflang 1994). The two marketing instruments could not be used together. Thematic marketing communication is focused on the long term, while those of the action marketing communication are supposed to be more successful on the short term.

Marketing Mix

Sales Promotions

Price Discounts

Coupons

Bonus Packaging

Sampling

Presents with payment

Free Premiums

Advertising

Sponsoring

Publicity

Cooperate design

Public Relations

Price

Product

Place

Promotion

Marketing communication instruments

1. Thematic marketing communication

2. Action marketing communication

Figure 2.1 Overview of the marketing communication mix

(15)

A temporarily price discount has a short term effect on sales, whereas advertising has a long term effect. However, opposite effects can also occur; a price discount can damage a carefully build quality image of a product or service within weeks. There exists a negative correlation between price and quality. Products with high prices are regarded to have higher quality standards. Hence, many people believe that a discounter does not sell quality products. In practice, though, the distinction between action and thematic marketing communication is difficult to recognize. If for example, a brand is on promotion that is supported with a feature, the consumer sees advertising. The brand is in reality advertising a promotion. In some cases the consumer may forget the promotion, but maybe the consumer will remember the brand name when he or she is visits the store. By remembering the brand it meets the goal of advertising (Shapiro, 1990).

2.3 Target groups of sales promotions

Promotions can target three different groups, namely: consumers, distributors, and sales force.

Promotion aimed at consumers is referred as Consumer promotions. They are designed to persuade new category users, induce trial, award existing users or to change purchasing behavior. Trade promotions target distributors. Its main purposes is to maintain or increase existing distribution opportunities, increase shelf spacing or to persuade the distributor to sell more products of the manufacturer. Sales force promotions are like trade promotion, targeted

Price Discounts

Coupons

Bonus

Packaging

Sampling

Presents with Payments

Free premium

Thematic communication

Display in store

In-store

communication

Packaging

External media

2. Classic Sales Promotion 1. Thematic Sales

Promotion

Long-term Short-term

Sales Promotions

Figure 2.2 Overview of “sales promotions”

(16)

at the supply side. Their main function is to motivate the sales force in selling more products of the manufacturer (Blattberg and Neslin1990).

(17)

2.4 The rise of Sales Promotions

Sales Promotions have become immense popular among marketeers. In the United States it is estimated that over a decade, the marketing budget spend on sales promotions have increased from 60 to 70 % (Ailiwadi et al. 2001). This is also accompanied by the increase in tailor made promotions. Tailor made promotions are developed by the manufacturer in cooperation with a specific retail chain. It enables the retailer to distinguish themselves from competitors.

Successful tailor made promotions can result in store switching. This type of promotion is also beneficiary for the manufacturer. The brand manager can influence on how his or her product is promoted. The increase in popularity of promotions is often attributed to:

The increase of the number of products and brands that are sold. The parties that are involved in the retail chain from manufacturer to consumer have to divert their attention to more products and brands.

Brands have difficulties in distinguishing themselves from competitors.

Thematic communication like advertising has a long term effect. Usually managers want quick results.

The effects of promotions are easily calculated. On the other hand the effects of advertising and sponsoring and marketing PR are difficult to calculate (Floor and van Raaij 2003).

Three quarter of the decision making by the consumer is made in the stores.

Consumers are learning to anticipate on promotions.

Retailers are becoming more powerful; they demand that manufactures participate in promotions. (Leeflang 1994).

(18)

Chapter 3 How promotions affect sales

This chapter discusses the effects of promotions and how they are affecting sales. The first part of this chapter introduces several decompositions of sales promotions. The sales decomposition of Blattberg and Neslin (1990), the elasticity based, and the unit sales based decompositions are discussed. The second part discuses the deal effects curve of promotions.

This chapter concludes with the main empirical findings concerning promotions.

3.1 Decomposition of Sales Promotions

Sales promotions can generate short term sales increases. To determine whether a sales promotion is beneficial from a retailers or manufacturer perspective the sales promotion is decomposed. The sales increase of a brand can come from other brands, from other time periods, and from category expansion. From a retailer perspective brand switching is not beneficial, except for possible differences in margins. However, a manufacturer can obtain a benefit from cross brand effects. Neither retailer or manufacturer benefit if the sales increase comes from other time periods, unless they get higher margins during the promotion or the stockpiling is intended to take the consumer out of the market. Blattberg and Neslin (1990) propose a composition that consists of four components;

1. Brand switching 2. Repeat Purchasing 3. Purchase acceleration 4. Category expansion.

1. Brand switching:

This occurs when a consumer purchases a different brand that is on promotion, instead of purchasing the usual non promoted brand. Brand switching effects can be aggressive or defensive. Aggressive switching effects occur when the promotion induces the consumer to buy a different brand then he or she otherwise would have bought. Defensive switching effects occur when the consumer buys the same non promoted brand rather than a different brand that is on promotion. From the manufacturers perspective it is critical to know whether the promotion attracts new consumers or rewards existing consumers. For the retailer is it important in which store the promoted brand is purchased. A consumer may respond by switching stores, to buy the brand that is on promotion at a store that he or she otherwise would not visit. A consumer might due to a promotion not only consider switching brands, but he or she may also switch stores. The latter confirms that there is a relation between store switching and brand switching.

2. Repeat Purchasing

Repeat purchasing refers to the consumer’s probability of buying the promoted brand again after the promotion. Repeat purchases are affected by promotions in two ways. Promotions can affect repeat purchasing because consumer’s learning and habit formation is triggered every time when a consumer purchases a brand. Over time he or she learns about the brand and forms a habit in buying the brand more often. The habit forming would not have occurred if the brand would not have been promoted. Hence, the promotion proved to be the incentive

(19)

to try the brand. If sales promotion induce purchases which otherwise would not occur, then it is known as the purchase effect. The second effect occurs when the purchase probability changes due to a promotion. A promotion may weaken the relation consumer and product. His or her attitude changes towards the brand and this may decrease the probability of buying the brand in the future. This is also known as the promotion usage effect.

These two effects can occur separately or simultaneously. A consumer may buy a product that is on promotion (formation of a habit, as a result the purchase probability increases), but at the same time the consumer might evaluate the purchased brand lower due to the promotion (purchase probability decreases).

For durable consumer goods the concept of repeat purchasing is, in the context of cross selling, important as well. A satisfied consumer may increase his or her purchase quantity. If Ikea is taken as an example: a consumer may buy a couch that is on promotion, and if he or she is satisfied it may also purchase a bed or a complete kitchen at the same store. Although, this is not completely the same as repeat purchasing, as in buying the same product, but we consider it to be the same as repeat purchasing effects.

3. Purchase Acceleration

Purchase acceleration occurs when consumers are purchasing different quantities of the product, or at different times in response of a promotion. Consumers have the tendency to buy more of the same product and purchase it less in the future. If it concerns retailers it is known as forward buying whereas for consumers it is referred as stockpiling. Purchase acceleration may also result in a shift of the time of purchase, and then it is not increasing the total sales, because the purchase would occur anyway. This is also known as sales displacement.

However, there are also advantages of purchase acceleration. First it takes the consumer temporarily out of the market, thereby preventing the consumer to switch brands. Consumers can be induced by a promotion to consume more of the purchased product.

This concept can be applied on durable goods by the concept of stock*. For example furniture that is owned by consumers, is the number of furniture products owned times the average quality per furniture. Quality is related then to age, performance, reliability, etc.

Stock t = Stock t-1 – Retired stock t – depreciation t + new stock t

Retired stock refers to furniture that is no longer used. Depreciation is the deterioration due to usage of the product. New stock represents new furniture that is purchased. In period t, the stock present with consumers can increase or decrease. When it increases consumers are buying more furniture, do not retire much furniture and do not wear down their furniture as usual. A sales promotion can influence this process. Consumers can be induced by promotion to buy more furniture and as a consequence retire more furniture. A promotion could also induce consumer to buy higher quality furniture. Due to acceleration there will be more furniture in stock and lesser retirements. As a consequence the demand for new furniture will decrease. Of course consumers can be encouraged to buy furniture more often than they used to do and get rid of the old furniture sooner.

* Blattberg and Neslin use in their example automobile whereas this study uses furniture as an example. The concept is applicable for most durable goods.

(20)

4. Category expansion

Category expansion is the increase in category sales during the period surrounding the promotion. That is the increase in sales that is not borrowed from other brands or time

periods. Promotion can stimulate sale by means of creating new opportunities to purchase the product, increasing the consumption rate, and to accelerate the purchase timing. However, consumers have the tendency due to promotions, to stockpile and repurchase the category later or purchase less of the category in the future. To prevent this, a brand can stimulate consumers to use the product in different circumstances, and then they otherwise would do.

As a result of this consumers learn more of the product and how to use it more. Blattberg and Neslin (1990) argue that buyers of durable consumers goods are easier induced to accelerate their inter purchase times. For example they can be persuaded by home styling programs to change their furniture according to the latest fashion.

Besides the decomposition of Blattberg and Neslin (1990), two other types of decompositions exist: gross (elasticity based) and net (units sales based). Promotions bases on sales elasticities are decomposed in terms of brand choice, category incidence and quantity elasticities (van Dijk 2004). Gupta (1988) introduced the gross decomposition, and it is also used by Bell et al.

(1999). Gupta (1988) decomposes the increase in sales into three components; Brand switching, purchase acceleration and stockpiling. The author finds that brand switching was the most important contributor to the increase in sales. His study shows that 84 % is due to brand switching, purchase acceleration in time accounted for 14 % and 2 % is due to stockpiling effects. Bell et al. (1999) show that the price promotion elasticity can be decomposed into 74 % to brand switching, 15% in quantity elasticity, and 11 % purchase incidence elasticity.

Van Heerde et al. (2003) use the net approach and proposes to decompose the unit sales into three components; cross-brand effects, cross period effects, and category expansion effects.

1. Cross brand effects

Cross brand effects can occur within and between brands. The first is also known as Cannibalization. This occurs when consumers purchase a different product that is on

promotion, but of the same brand. The second occurs between brands; the consumer switches brands.

2. Cross period effects

Cross period effects are also known as stockpiling effects. These are the losses in turnover before and after a promotion. It includes the pre-promotional effects due to purchase deceleration and post- promotional effects as result of purchase acceleration.

3. Category expansion

Category expansion effects are decomposed into cross-store and market expansion effects.

Market expansion effects are the results of extra purchases from consumers that come from another category. Cross store effects are an increase in sales as a result of extra consumers who usually shop at other stores.

(21)

The framework of van Heerde et al. (2003) is extended by van Dijk (2004). Van Dijk (2004) adds a fourth component, which is suggested by Neslin (2002), the cross category effects.

4. Cross category effects

Cross category effects are the change in sales of items in other categories. The effects are separated into an own-, cross item- and cross category effects these effects can be positive or negative. Positive effects occur when a promoted products results in an increase in sales of other related products. Negative effect occurs when a promoted products stimulates switching behavior from one category to another.

Promotional elasticity is frequently misinterpreted by researchers. It is believed that the increase in sales represent consumers switching away from other brands. This means if a promoted brand increases its sales with 100 units, the other brands would lose together 75 units. However, this approach does not account for category expansion, the elasticity percentages are not equal to unit sales percentages. This means that if 75 % is attributed to brand switching this does not result in a gain of 75 units for the promoted brand and likewise 75 unit losses in sales of competing brands (Van Heerde et al. 2003).

The net approach accounts for the increase in purchase incidence that benefits all brands in the category. This means if a brand is promoted, the category sales will also increase.

Because, if the category sales also increase, it will compensate the effects for the loss due to brand switching. Thus, if a promoted brand gains 100 units, the net loss for other brands is 33 units. This means that 33 % of the sales increase, if one accounts for the compensating effect, is due to brand switching or cross brand effects. Van Heerde et al. (2004) show in their study that one third of the unit sales increase is attributed to cross period effects and the last third is attributed to category expansion effects. Van Dijk (2004) concludes that considering cross category effect has a serious impact on the evaluation of sales. Omitting the cross category effects may lead to either over- or under estimation of the price promotion’s effect. The author finds that positive and negative effects across categories are both about one third of the within category effects.

3.2 Deal Effect Curve

The deal effect curve shows how sales respond to temporary price discounts. Due to the nonlinearities and interactions in the deal effect curve the shape of it has become a complex issue in marketing. Whether the shape is linear, or has another shape such as: concave, convex, or S-shaped is not known (Blatberg, Briesch and Fox 1995). For retailers and manufacturers it is important to know the shape of the deal effect curve. If it is convex the retailer will run steeper deals then if it is concave

The shape of the deal effect curve is determined by five components. These are Threshold effects, Saturation effects, Cross-item effects, Interaction effects between deals of different items, and interaction effects between deals and promotional signals (Van Heerde et al. 2001).

1. Threshold effects

(22)

This is the minimum value of a price promotion to induce consumers to change its purchase behavior. Research shows that a price discount has to be at least 15 % (Della et al. 1980), 10% (Van Heerde et al. 2001), or 15 %, (Gupta and Cooper 1992), otherwise the consumer does not notice that the item is on promotion. Gupta and Cooper (1992) find in their study that this differs per brand.

2. Saturation effects

Saturation effects influence the deal effect curve in two ways. One is the limit of items a consumer can stockpile (Blattberg, et al. 1995) and or the amount it can consume. Some items are difficult to stockpile, like large items (Van Heerde et al. 1997). The other refers to

consumer behavior to discount the discount, this increase with higher advertised savings.

Consumers have internal reference prices to which current prices are compared (Kalwani et al. 1990). The perceived discount is a result of the expected savings from the internal reference price. If the advertised discount by retailers increases, the perceived discount increases likewise. However, large advertised discount are distrusted by consumers. Cooper and Gupta (1992) argue that discounts higher than 40 % are not more effective than discounts around 40 %. The saturation level of discounts is around 40 %.

3. Cross-item Deal Effects

Consider how the relative price effects affect the cross-item deal effects. Thus, which effects does a price discount of a high priced brand have on the sales of a low priced brand?

Sethuraman (1996) finds that low priced brands sales decrease if a higher priced brand is below the low priced brand.

4. Interaction effects between deals of different items

The shape of the deal effect curve is also influenced by other product’s discounts. The discounts of other products influence the sales of the focussed product (Van Heerde et al.

2001).

5. Interaction effects between deals and promotion signals

Consumers react on the presence of a promotional signal. Some consumers react to promotional signals, such as a display, without concerning the price information. In their opinion a display is a cue for a price cut. If a discount is supported with a display the response is even greater (Inman et al. 1990). Van Heerde et al. (2004) find in their study that the effect of price discounts are influenced by the type of support; a feature is more effective than display at low discounts, whereas a display is on the other hand more effective for higher discounts. Price discounts which are featured may result in cross period effects. Thus, if a price discount is featured it can result in brand switching effects or stockpiling effects.

Display supported discounts have category expansion effects; displays seem to induce trial.

3.3 Empirical Findings

The effects of promotions can be studied at the short and long term. Short term effects are those effects that occur during the promotions or several weeks after the promotion. Long

(23)

term effects occur after several weeks and last several months. On the short term promotions can increase sales dramatically (Van Heerde et al.1997).

Most of the discussed empirical findings that are mentioned below are based on the article of Blattberg, Briesch and Fox (1995). They describe the empirical generalizations in sales promotions up to 1995. This is supplemented by other useful articles for this study. The findings in this paragraph deals predominantly with fast moving consumer goods, because there is almost no literature that studies the effects of promotions on durables. However, it is useful to study the effects of promotions on fast moving consumer goods, because it gives an insight on how promotions work.

A temporarily price cut can increase sales. Sales promotions cause a spike in the deal effect curve.

Higher market-share brands are less influenced by promotions. Thus, brands with a large market-share tend to have lower elasticises even though the brand in focus may attract a larger proportion of switcher of other brands than lower share brands do.

The frequency of sales promotions influences the reference price. If a product is

discounted frequently this influences the price that the customer is willing to pay for the product. This is the most important explanation for the loss of brand equity when brands are heavily promoted. The lower the reference price, the lesser the premium that can be charged for the brand, which results in the loss of brand equity.

The more frequent a price promotion is held, the lower the deal spike will be. This is due to the expectations consumers have regarding promotions and changes in reference prices.

Cross promotional effects are asymmetric; stronger brands have a stronger impact on sales of weaker brands. When a strong brand is on promotion it will result in more switchers from a lower quality brand to the brand on promotion, than when the weak brand is on promotion. This is easily explained by brand equity. Weaker brands have lower brand equity, thus when a stronger brand is on promotion it will effect the sales of a weaker brand negatively.

Display and feature have strong effects on product sales. Simultaneous use of feature and display leads to synergy. If price, feature and display are used together it results in higher sales increase than when they are used separately.

Advertised promotions can influence store traffic. Some brands that are on promotion can attract new customers to the store.

Promotions affect sales in complementary and competitive categories (Blattberg et al.

1995). For instance when a couch is on promotion this can also result into an increase in the category TV-benches but it can also reduce the sales of a sofa.

Consumers may respond different to sales promotion. Some segments are more sensitive to sales promotions than others (Bucklin and Gupta 1998). If a sales promotion is held at a store it may attract different types of costumers than it otherwise would when there would not be a promotion.

(24)

Long term effects of promotions are not persistent. The effects on brand choice, category sales and profit are unknown (Dekimpe and Hanssens 1999). Although, Nijs et al. (2001) find in their study that the long term category expansion effects due to promotions are stationary and dissipate within 10 weeks, the long term impact is nearly zero.

The amount of discount influences the price discount elasticity. If a promotion has a large price cut than the next time consumer expect the same discount (Foekens et al.1999).

Promotions create lead and lagged effects; they anticipate on promotions in the future and are stockpiling (Van Heerde et al. 2001).

There are threshold and saturation effects. Promotions less than 10% do not generate enough sales to be noticed by the consumer. Saturation effects also exists, discounts above 25% are not more successful than those who are discounted around 25% (Van Heerde et al. 2001). Gupta and Cooper (1992) find levels of 15 % and 40 %, respectively.

(25)

Chapter 4 Sales Analysis

This chapter studies the data set that is collected from the Ikea Groningen database. The data is explored by calculating descriptive statistics and examining the sales development. The outcomes of this chapter are used to build the conceptual model in chapter 6. In the first part of this chapter the descriptive statistics of the products under study are discussed. This chapter ends with the discussion its time series plot of the investigated products.

4.1 Sales analysis of the Category TV Benches

Figure 4.1 presents the total sales per unit of the category TV Benches & Media furniture.

During the study, a total of 11.502 units were sold. The first two bars of figure 4.1 represents the total and baseline sales of the Ikea ps. Baseline sales are defined as products sold without discount. According to figure 4.1, the Ikea ps is the most frequently sold product. With a total of 2963 units sold, the Ikea ps makes up for nearly a quarter of total unit sales. The difference between the baseline and total sales of the Ikea ps, is caused by the promotion. Nearly one third of the total unit sales is attributed to vouchers (991) and more than one sixth (523) is attributed to the “Woonsdag” of 28th April 2004. Thus, more than half (1514) of the Ikea ps unit sales were sold with a discount. Another large category is the category Other, this makes up for more than 20 % of the total unit sales. The third largest unit sales are the Oppli’s. If the Oppli’s are taken together into one category, then they represent nearly 30 % of total unit sales. This means that one out of three TV benches is sold is an Oppli.

(26)

Total unit sales TV Benches & Media furniture

2963

1951

1168 1134

1032

508

190

2565

1972

0 500 1000 1500 2000 2500 3000 3500

Ikea ps Total

Ikea ps baseline

sales

Oppli Small

Oppli Large

Oppli Silver

Stormark Ikea ps Silver

Timrå Other

Units

Figure 4.1 Total units sold

Important introductions in the category are the Stormark and the Oppli silver. During the summer of 2004 these products were introduced and in a relative short period, they have managed to gain a market share of 10 % each. Other new introductions like the Ikea ps silver and the Timrå have been less successful (See Appendix I).

When the unit sales are compared to total turnover there are little changes. The most

important generator of sales is the Ikea ps (see figure 4.2). With turnover of nearly € 145,000, it alone accounts for 25 % of the total turnover of the whole category. The Ikea ps baseline sales amounted nearly € 100,000. The difference between the base line and Ikea ps total amounts nearly € 45,000. The Category Other contributes a little bit less than a quarter to the total sales. With a total turnover of more than € 180.000 the Oppli’s are important generators of sales. Due to the increase of the price of the Oppli small and large, the Oppli category’s contribution to the total sales has increased.

Referenties

GERELATEERDE DOCUMENTEN

Hierna volgen lineaire afbeeldingen (eigenwaarden en bijbehorende eigenvectoren, kern). Hoofdstuk 10 bespreekt matrices en determinanten, lineaire vergelijkingen, de algebra van

ELECTRICAL CONDUCTIVITY STUDY OF DEFECTS IN ZIRCONIUM DOPED YTTRIUM ALUMINIUM GARNET CERAMICS..

search and publish in high quality journals such as International Marketing Review, Journal of Busi- ness Research, European Journal of Marketing, Journal of Business

In addition, practical relevance comes from the fact that this study adds to the knowledge of the (Dutch) pharmacy channel, especially regarding the channel’s sales

Research in the following fields is explored ; (1) Performance measurement systems, (2) Behavior-based and outcome-based control, (3) Agency theory, (4) Organizational theory,

We find that the stringency of measures and fear of the virus had a substantial positive effect on sales and that these relationships were moderated by country and product

Economy private labels have been advanced as a potentially effective defensive tool for traditional grocery retailers against the rise of hard discounters like ALDI

Four different models are constructed to investigate sequence effects on different forms of purchasing behavior: if sequence effects play a role in whether households purchase