BI Norwegian Business School - Thesis
CUSTOMER SATISFACTION:THE ROAD TO SUCCESS IN A RECESSION.
AN INVESTIGATION OF HOW CUSTOMER SATISFACTION AND FIRM SIZE MODERATE THE EFFECTS OF ADVERTISING AND R&D SPENDING ON FIRM PERFORMANCE IN RECESSIONS.
31-07-2012
By:
Jeroen Kromme 0934256 Sander Spinder 0934262
Program:
Double Degree Strategic Marketing Management Supervisor:
Associate Professor A. Hunneman, PhD
This thesis is a part of the MSc programme at BI Norwegian Business School. The school takes no responsibility for the methods used, results found and conclusions drawn.
Customer satisfaction: The road to success in a recession.
An investigation of how customer satisfaction and firm size moderate the effects of advertising and R&D spending on firm performance in recessions.
Hand-in date: 31-07-2012
University: BI Norwegian Business School
Campus: BI Oslo, Nydalen
Program: Double Degree Strategic Marketing Management Exam code and name: GRA1900 – Master Thesis
Supervisor: Associate professor A. Hunneman, PhD
University: University of Groningen (RuG)
Department: Marketing
Qualification: Master Thesis
Program: Double Degree Strategic Marketing Management
Exam code and name: EBM867A30 – Master’s Thesis BA Marketing; Double Degree Supervisor: Assistant professor of Marketing Dr. Ir. M.J. Gijsenberg
Jeroen Kromme Sander Spinder
Student number BI: 0934256 0934262
Student number RuG: S1687700 S1539647
Address: Kleine Badstraat 19
9726 CG Groningen The Netherlands
Westindischekade 98 9715 TE
Groningen The Netherlands
Phone number: +31 653440966 +31 636147332
E-mail: krommejeroen@hotmail.com sanderspinder@gmail.com
Page i 31 July 2012
EXECUTIVE SUMMARY
The goal of this study is to examine why some firms have to cease operations in recessions, while others flourish. Extending the work of Boulding and Staelin (1995) and Srinivasan, Lilien and Sridhar (2011), we investigate the moderating effect of firm-level contingency variables, customer satisfaction and firm size, on the effects of advertising and R&D expenditures on firm performance in times of recession. We model the effect of marketing activities on four different performance metrics: stock returns, net income, ROA, and EBITDA margin.
We use the concept of consumer confidence to identify recessionary periods, a concept which is, t our knowledge, never used before in this context. Consumer confidence grasps the actual psychological results of the recession in the mindset of the consumer and is therefore better than a macro variable as GDP.
Using a multi-level approach for repeated measures, we find that customer satisfaction plays a vital role in times of recession. Advertising and R&D is more effective for firms with high customer satisfaction. The rationale for this is that satisfied customers are less price-sensitive, engage in more positive word-of- mouth, are more likely to try and adopt new products, and are less likely to switch to competitors. Moreover, we find that, in a recession, R&D is more effective for small companies, while advertising is more effective for large companies.
We also investigate whether small and large firms with high or low customer satisfaction under- and/or overspend on advertising and R&D in times of recession and non-recessionary periods. We find that small firms, in comparison to bigger firms, generally under- or overspend to a lesser extent, as these firms are closer to the market and more flexible, and are, therefore, better able to (re-) allocate their budgets. Furthermore, as recessionary periods are characterized by change, it is harder to allocate budgets properly, which leads to more under- and overspending in times of recession for both small and large firms. We also find that firms underestimate their customer satisfaction or the power of customer satisfaction, since many firms with satisfied customers substantially overspend.
Page ii 31 July 2012 We develop a webtool with which managers can easily replicate this analysis. By providing their advertising and R&D budgets managers can benchmark their results against comparable firms and see whether they are under- or overspending. This helps managers to decide whether the current (budget allocation) strategy is suitable and to prioritize investments. The webtool is accessible at www.scima.nl/thesis.
Page iii 31 July 2012
ACKNOWLEDGEMENT
This thesis is the final step in the achievement of a double degree in Marketing, combining the Master in Marketing Research and Management at the University of Groningen, and the Master in Strategic Marketing Management at BI Norwegian Business School.
This thesis also marks the end of our student-life, of which we spent the last year in Oslo, Norway. Our exchange year in Oslo has given us many things, both on a educational and personal level. We have seen many sides of the beautiful Norwegian landscape, experienced the Norwegian culture in many different ways, and made life-lasting friends.
We want to thank our thesis supervisor Associate Professor Auke Hunneman for his valuable feedback and suggestions. During the process, he has been of constant help. We also want to thank the University of Groningen and BI Norwegian Business School for giving us the opportunity to take part in the double degree program; it has been an enrichment of our lives.
Page iv 31 July 2012
CONTENTS
1 Introduction ... 1
2 Literature review... 5
2.1 Periods of economic downturn... 5
2.1.1 What happens in recessions? ... 5
2.1.2 Advertising in times of recession ... 6
2.1.3 R&D in times of recession ... 7
2.1.4 Firm-level contingencies ... 9
2.2 Customer satisfaction ... 10
2.2.1 Satisfaction and advertising in times of recession ... 12
2.2.2 Satisfaction and R&D in times of recession ... 14
2.3 Firm size ... 15
2.3.1 Firm size and advertising in times of recession ... 16
2.3.2 Firm size and R&D in times of recession ... 18
2.4 Conclusion ... 19
3 Methodology ... 21
3.1 Model specification ... 21
3.2 Data ... 23
3.2.1 Dependent variables ... 24
3.2.2 Independent variables ... 25
3.3 R and the ‘nlme’-package ... 29
4 Validation and results ... 30
4.1 Validation ... 30
4.1.1 Multicollinearity ... 32
4.1.2 Heteroscedasticity ... 33
4.1.3 Autocorrelation ... 34
4.1.4 Normality of residuals ... 35
4.2 Results ... 36
4.2.1 Results of Stock returns model ... 37
4.2.2 Results of EBITDA margin model ... 38
Page v 31 July 2012
4.2.3 Results of net Income model... 39
4.2.4 Results of ROA model ... 39
4.2.5 Summary of the results ... 40
5 Discussion ... 41
6 Managerial implications ... 46
7 Limitations and further research ... 51
7.1 Limitations ... 51
7.2 Further research... 52
8 References ... 53
Appendices ... 62
Appendix 1: List of companies used in analysis ... 62
Appendix 2: Correlation matrices before and after centering ... 63
Appendix 3: Level 1 Heteroscedasticity ... 65
Appendix 4: Homoscedastic level 2 residuals ... 66
Appendix 5: ACF and PACF for all models ... 67
Appendix 6: Residuals for stock returns, net income and ROA ... 68
Appendix 7: Normality of random coefficients... 70
Appendix 8: Marginal effects ... 71
Appendix 9: Interface for managers ... 72
Preliminary thesis report ... 73
Page 1 31 July 2012
1INTRODUCTION
“Everything in the world may be endured except continual prosperity”
Johann Wolfgang von Goethe (1749 – 1832)
As Goethe expresses, prosperity does not last. This quote is highly applicable to economic prosperity, as we encounter a recession roughly every ten year (Claessens, Kose and Terrones 2009). Recessions can have enormous and enduring impacts on economies (Claessens, Kose and Terrones 2009), companies (Perry and Schultze 1993), and individual consumers (Shama 1978, 1981). In 2007, for instance, the USA was struck by a severe recession due to the collapse of the housing market, with rigorous implications for other countries around the world. Within only one year, housing prices in the US dropped by 17%, stock prices on the S&P 500 decreased by 37%, and unemployment rose from 5% to 7.4%1. In the UK, GDP fell by 6% and employment declined with 1.6 percent points in the 2007-2009 period, and similar effects are observed in, for example, Japan and Sweden2.
In recessions, companies encounter changes in the market, and these changes can have a substantial influence: while some companies go bankrupt or are forced to fire employees due to necessary budget cuts, others flourish and gain superior results (Srinivasan, Rangaswamy and Lilien 2005). One explanation for these differences in firm performances is that companies differ in how they deal with marketing activities in times of recession (Tellis and Tellis 2009).
To maintain liquidity, companies often cut down their advertising and R&D expenditures as these are easy to cut and produce limited short-term cash flows (Lamey et al. 2007; Srinivasan, Lilien and Sridhar 2011). However, numerous studies reveal that the opposite, or holding budgets constant, can be more effective. Advertising in times of recession can lead to increased earnings (Graham and Frankenberger 2011), sales (Kambler 2002) and stock performance
1 National Bureau of Economic Research: Effects of the Financial Crisis and Great Recession on American Households (2010)
2 BBC News: http://www.bbc.co.uk/news/uk-14879003/
Page 2 31 July 2012 (Deleersnyder et al. 2009). Also, R&D expenditures in times of recession can boost firm performance (DeDee and Vorhies 1998).
There are mixed findings in the literature about whether firms should de- or invest in times of recession. Because of these mixed findings, Srinivasan, Lilien and Sridhar (2011) propose that the effect of advertising and R&D depends on firm-level contingencies. They successfully demonstrate that the effect of advertising and R&D in times of recession is moderated by a firm’s market share, its financial leverage, and its product-market profile (B2B versus B2C, and services versus goods). Moreover, Boulding and Staelin (1995) claim that firms have different firm performances, because companies differ in their motivation (e.g. the second wants to become the best) and ability (e.g. its resources and managerial abilities).
In our study, we explain the different effects of advertising and R&D expenditures in times of recession, based on a firm’s size and customer satisfaction. We build on the work of Boulding and Staelin (1995) and Srinivasan, Lilien and Sridhar (2011), however, we innovate by operationalizing the constructs in different ways. As a proxy for ability we use customer satisfaction; a metric that grasps the realized outcome of a company’s behavior. We use firm size as a proxy for motivation.
Furthermore, we use consumer confidence as an indicator of economic prosperity. Consumer confidence is based on consumers’ direct appraisal of the current economic state and expectations about the future economic state, it therefore grasps the actual psychological results of the recession in consumers’
minds. As marketing instruments focus on changing consumer preferences and behavior (Acemoglu and Scott 1994; Ang 2001; Keller 2008; Shama 1978), consumer confidence is a much better indicator than a simple dummy variable which most other researchers use (e.g. Mela, Gupta and Lehmann 1997; Lamey et al. 2011; Srinivasan, Lilien and Sridhar 2011). As far as we know, no other research in our field has ever used consumer confidence as a metric to indicate a recession.
Page 3 31 July 2012 To sum up, we examine how a company’s size and customer satisfaction influence the effect of advertising and R&D expenditures on firm performance.
To indicate a recession we use consumer confidence, a metric which is, to our knowledge, never used before for this purpose. Hence, we identify the following research question for our study:
“How does a company’s customer satisfaction and firm size influence the effect of advertising and research and development on firm performance in times of
recession?”
According to Verhoef and Leeflang (2009), marketing is losing its influence in the boardroom; the gap between executives and analytical marketers grows. One of the reasons for this trend is that senior management and finance executives do not understand the metrics used by marketers (McAlister, Srinivasan and Kim 2007), and it is this accountability which is one of the key drivers of marketing’s influence on top management (Verhoef and Leeflang 2009).
In our study, we try to bridge this gap by using input variables marketers use as independent variables (advertising and R&D expenditures), and link these to performance metrics used by senior management. We incorporate both contemporaneous metrics (net income, EBITDA margin, and ROA), and a forward-looking metric (stock returns) to capture short- and long-term effects.
This increases relevance to senior management.
We use customer constructs that are close to the market and directly provided by consumers (customer satisfaction and consumer confidence) to better capture the real drivers of these constructs. In addition, we translate our research findings into managerial implications in chapter 6. We develop a webtool with which companies can benchmark their firm performance against comparable companies. Managers can, with this tool, also analyze whether they are currently over- or underspending on advertising and R&D.
By providing strong empirical evidence for the link between marketing variables and performance metrics, we demonstrate the impact of marketing on firm
Page 4 31 July 2012 performance. This should help marketing to regain its importance and bring marketing back into the boardroom.
To sum up, we try to bridge the gap between senior management and marketers by (i) using marketing variables as independent variables, (ii) incorporating actual customer metrics (consumer confidence and customer satisfaction), and (iii) incorporating an extensive overview of firm performance metrics frequently used by senior management as dependent variables.
The thesis continues with a literature review in which we hypothesize the effects of advertising and R&D expenditures on firm performance in times of recession.
We then discuss the moderating effects of customer satisfaction and firm size and introduce our conceptual framework. Chapter 3 describes the methodology and data, and in chapter 4, we check the model assumptions and present the results. This is followed by a discussion of the results in chapter 5. As we want to bridge the gap between marketers and senior management, we incorporate a separate chapter with managerial implications (chapter 6). In this chapter, we provide an analysis of marginal effects and investigate whether small and big firms, with high or low customer satisfaction, over- or underspend in times of recession and in non-recessionary periods. The webtool is also introduced in this chapter. The thesis ends with a discussion of the limitations of our research and areas for further research.
Page 5 31 July 2012
2LITERATURE REVIEW
2.1PERIODS OF ECONOMIC DOWNTURN
There is no universal definition for a period of economic downturn or recession, however, most researchers follow the definition of the US National Bureau of Economic Research (NBER) (e.g. Claessens, Kose and Terrones 2009; Graham and Frankenberger 2011; Srinivasan, Lilien and Sridhar 2011). The NBER defines a recession as: ‘A significant decline in activity spread across the economy, lasting more than a few months, visible in industrial production, employment, real income, and wholesale-retail trade’ (NBER 2001).
Recessions are common, widespread, and inevitable and can have a serious and long-lasting impact on economies (Claessens, Kose and Terrones 2009), companies (Perry and Schultze 1993), and individual consumers (Shama 1978, 1981). Claessens, Kose and Terrones (2009) investigated 21 advanced economies in the 1960-2007 period and found a total number of 122 recessions, of which 30 were severe. This means that economies are about 20% of the time in a state of recession. Because of its impact and omnipresence, recessions have been widely studied (Blanchard 1993; Claessens, Kose and Terrones 2009; Perry and Schultze 1993). We will give a brief overview of what happens in a recession and then discuss its impact on advertising and R&D spending.
2.1.1WHAT HAPPENS IN RECESSIONS?
Recessions have various causes; for example, underconsumption is assumed to be the cause of the 1990-1991 recession (Blanchard 1993), while the collapse of the US mortgage market was the cause of the most recent crisis (O’Malley, Story and O’Sullivan 2011). Despite recession-specific characteristics, all recessions have the following in common: a decline in investments, lower consumer consumption, increasing unemployment rates, lower credit growth, lower import, and a decline in credit and housing prices (Claessens, Kose and Terrones 2009).
Consumer confidence plays a major role in inducing and fuelling a recession, as it initiates a vicious circle (Dhalla 1980; Perry and Schultze 1993). Due to reduced consumer confidence, consumers spend less, companies’ inventory levels rise,
Page 6 31 July 2012 and workers are fired, which, in turn, reduces consumer confidence even more (Dhalla 1980).
Moreover, during an economic downturn, consumers become more price conscious because their feeling of job security lowers (Shama 1981). Lamey et al.
(2007) state that, because of this increased price sensitivity, consumers often switch from well-known brands to private labels in recessions and that it takes longer to return to these brands when the recession is over.
Recessions imply change and present thus also opportunities. During the 2001 recession in the US, more than 20% of the companies in the bottom quartile of performance in their industry managed to climb to the top quartile (Srinivasan, Lilien and Sridhar 2011). Economic downturns provide unique opportunities to gain market share, sales and higher returns, which are more difficult to achieve in times of an economic upturn (Kamber 2002; Tellis and Tellis 2009). So, although recessions may have negative influences on a micro and macro level, they also entail opportunities.
2.1.2ADVERTISING IN TIMES OF RECESSION
In times of an economic downturn, marketing budgets (including advertising budgets) are often a prime target for cost cutting (Srinivasan, Rangaswamy and Lilien 2005). The rationale for this is that marketing budgets are easy to cut and can boost short-term earnings in an effort to survive the crisis (Lamey et al.
2007). Indeed, Kijewski (1982) finds that cutting advertising budgets in a recession does not result in lower profits or lower market share. However, numerous other studies reveal that the opposite, or holding budgets constant, can be more effective. Graham and Frankenberger (2011), for instance, find that for both consumer and industrial product firms, increasing advertising spending during a recession leads to earnings greater than the same increase in the advertising budget in ‘normal’ economic states. Kamber (2002) reports similar positive effects on sales, and Deleersnyder et al. (2009) find that companies that set their advertising expenditures independent from the business cycle achieve higher stock performance than companies that strictly tie their advertising expenditures to the business cycle.
Page 7 31 July 2012 One of the reasons why an increase in advertising expenditures might be more effective in a recession is that most companies set their advertising budget based on sales; i.e. in times of a recession, when sales decline, they cut back on advertising spending (Deleersnyder et al. 2009). When most companies limit their advertising spending, maintaining the level of advertising spending will automatically lead to a larger share of the total sum of advertising spending in the industry (Graham and Frankenberger 2011). The result is more exposure, attention, persuasion and eventually purchases (Tellis and Tellis 2009).
Additionally, as the costs for advertising also fall, the exposure per advertising dollar increases even more (Bromiley, Navarro and Sottile 2008).
Another reason for advertising becoming more effective in a recession is that consumers may become more sensitive to advertising (Mela, Gupta and Lehmann 1997). Because of lower consumer confidence, consumers want more secure and trustworthy products. They become more comparative shoppers, and judge products and services in a different way (Shama 1978). By increasing their advertising expenditures, companies send a reassuring signal to their customers, which might prevent them from switching to competitors (Srinivasan, Rangaswamy and Lilien 2005).
In conclusion, we see that most of the research indicates an increase in advertising effectiveness in times of recession. We therefore assume the following:
H1: In times of recession, companies with higher advertising expenditures will have better firm performance than companies with lower advertising
expenditures
2.1.3R&D IN TIMES OF RECESSION
For most companies, innovation is a key success factor: it raises the quality of products, lowers prices (Hauser, Tellis and Griffin 2006), and it is therefore associated with positive short- and long-term effects on firm performance (Pauwels et al. 2004). However, as with advertising budgets, R&D budgets are also often one of the first to be cut during a recession (Barret, Musso and Padhi 2009). As R&D investments yield almost no short-term cash flows, reducing the
Page 8 31 July 2012 budget is an easy way to generate some cash (Srinivasan, Lilien and Sridhar 2011). Yet some research demonstrates that innovation can play a key role in surviving a recession. For example, DeDee and Vorhies (1998) find that for small companies an increase in R&D is positively related to firm performance.
One of the characteristics of recessions is that they often cause changes in consumer preferences. Shama (1978) finds that customers change their habits and preferences, look for more durable products, and become more do-it- yourself persons. Ang (2001) shows that these findings also hold for Asian consumers. The most recent recession points to a change in customer preferences as well: customers seek more value for money and change from luxury to more basic products (Piercy, Cravens and Lane 2010). To meet customers’ new preferences, companies have to change their product portfolio and their relationship with customers (Shama 1978). Companies with a high level of R&D spending can deal better with these changes: ensuring a match between customer needs and product characteristics increases the success of incremental as well as radical innovations (DeBrentani 2000).
Henard and Szymanski (2001) show that the fit between new products and customer needs is one of the dominant drivers of new product performance.
DeBrentani (2000) shows that this holds for incremental as well as radical innovations: client/need match is the most important factor of innovation success for incremental innovations, and the second most important factor for discontinuous innovations. In times of recession, customers’ preferences and habits change, and by introducing incremental product innovations companies can re-establish a good match between products and customer needs.
Moreover, recessions open up possibilities for breakthrough innovations.
Breakthrough innovations offer highly unique benefits, have a high product quality, and are superior in the eyes of the customer (Kleinschmidt and Cooper 1991).
Srinivasan, Rangaswamy and Lilien (2005) show that, for firms with a strategic emphasis on marketing, pursuing a proactive marketing approach yields superior business performance in times of recession. They define proactive marketing as:
Page 9 31 July 2012
‘The organization’s interpretation of the recession as an opportunity and the development and execution of a response to capitalize on the perceived opportunity created by the change’. Companies with a strategic emphasis on marketing have established brands, differentiated products, targeted communications and make better and more effective use of R&D (Srinivasan, Rangaswamy and Lilien 2005). R&D helps in developing, executing, and capitalizing on opportunities induced by the recession.
So, although many companies cut back on R&D expenditures in a recession, there is abundant research that shows that an increase in R&D expenditures during a recession can significantly increase firm performance. We therefore state the following hypothesis:
H2: In times of recession, companies with higher R&D expenditures will have better firm performance than companies with lower R&D expenditures
Figure 2.1 gives an overview of the hypothesized relationships.
Figure 2.1
The effect of advertising and R&D expenditures in times of recession
2.1.4FIRM-LEVEL CONTINGENCIES
In the previous subsections, we hypothesized the effects of advertising and R&D expenditures on firm performance in times of recession. We stated that advertising and R&D spending will, in general, lead to higher firm performance.
Page 10 31 July 2012 As we already mentioned, however, the findings are somewhat mixed, arguing for a deeper understanding of the relationships.
Srinivasan, Lilien and Sridhar (2011) find that the effect of advertising and R&D spending depend on firm-level contingencies. The contingency-theory, proposed by Zeithaml, Varadarajan and Zeithaml (1988), states that the effect of a company’s actions on its performance is moderated by both firm and market characteristics. Srinivasan, Lilien and Sridhar (2011) find that the effect of advertising and R&D spending in times of recession is moderated by a company’s market share, its financial leverage, and its product-market profile (B2B versus B2C and services versus goods). Moreover, Boulding and Staelin (1995) claim that companies have different firm performance because they differ in their motivation (e.g. the second wants to become the best) and ability (e.g. its resources and managerial abilities).
In the next two sections, we introduce two different firm-level contingencies that influence the effect of advertising and R&D on firm performance in times of recession. As a proxy of ability, we use customer satisfaction. Customer satisfaction is the outcome of customer expectations, perceived quality, and perceived value (Fornell et al. 1996), and it is, therefore, a metric which grasps the realized outcome of a company’s behavior. As a proxy of a firm’s motivation we use firm size.
2.2CUSTOMER SATISFACTION
Customer satisfaction has been the subject of many studies (e.g. Aksoy et al.
2008; Fornell, Rust and DeKimpe 2010; Gupta and Zeithaml 2006), and there is a general consensus that it leads to better firm performance. Customer satisfaction results, for example, in an increase in customer retention (Gupta and Zeithaml 2006), positive word-of-mouth (Anderson 1998), and a decrease in customer complaints (Bolton 1998). Moreover, it is positively linked to financial outcome metrics such as higher profits (Fornell et al. 2006), return on investments (Anderson, Fornell and Rust 1997), and stock performance (Aksoy et al. 2008).
Page 11 31 July 2012 One frequently used measure of customer satisfaction is the American Customer Satisfaction Index (ACSI), developed at the University of Michigan’s Ross School of Business. The ACSI model is a cause-and-effect model, which specifies both the antecedents and consequences of customer satisfaction (Fornell et al. 1996).
As can be seen in figure 2.2, there are three antecedents of customer satisfaction, namely: customer expectations, perceived quality, and perceived value. Customer complaints and customer loyalty are the outcomes of customer satisfaction.
Figure 2.2
The cause-and-effect model of customer satisfaction (Fornell et al. 1996)
Customer expectations consist of two parts: a customer’s previous consumption experience with the firm, including non-experiential information such as advertising, and a forecast of the probability that the firm will be able to deliver quality in the future. Perceived quality is a consumer’s product evaluation of, on the one hand, the fit between his own needs and the product offering (customization), and, on the other hand, the reliability with which the firm will be able to deliver this offering (does it work as it is supposed to, and is it free from failures?). Perceived value is the subjective evaluation of the quality related to the price of the product. Customer complaints are simply the number of complaints a customer has filed with the company within a certain time period.
And customer loyalty is the likelihood that a customer will repurchase from the same company and the extent of a customer’s price tolerance.
We already stated that customer satisfaction is positively related to customer loyalty (retention) and that is leads to less customer complaints. In addition,
Page 12 31 July 2012 Fornell et al. (1996) also assume that there is a relationship between customer complaints and customer loyalty. They argue that the relationship will be positive if the firm is able to successfully handle complaints and turn complaining customers into loyal ones, and that it will be negative if the firm handles complaints unsuccessfully.
Customer satisfaction thus leads to loyalty (Gupta and Zeithaml 2006), positive word-of-mouth (Anderson 1998), a reduction in customer complaints (Bolton 1998), and ultimately, to higher profits (Fornell et al. 2006), return on investments (Anderson, Fornell and Rust 1997), and better stock performance (Aksoy et al. 2008). Although we are mostly interested in the moderating effects of customer satisfaction on advertising and R&D expenditures, we also investigate its main effect on firm performance. Therefore, we propose the following hypothesis:
H3: All else being equal, higher levels of customer satisfaction will lead to better firm performance
2.2.1SATISFACTION AND ADVERTISING IN TIMES OF RECESSION
We already discussed that advertising can be effective in times of recession. In an economic downturn, customers become less secure about their purchases, more price conscious and they engage in more comparative shopping. Firms with an effective advertising campaign reassure consumers in their choice and prevent them from switching to competitors.
Customer satisfaction moderates the impact of advertising in three ways. First, through customers’ expectations, second, through a smaller increase in price- sensitivity, and third, through word-of-mouth.
Firms with a higher overall consumer satisfaction have, by definition, more satisfied customers. In times of recession, customers seek for more secure and trustworthy products, and firms thus have to reassure their customers and prevent them from switching to competitors. A customer’s expectations consist of his previous experiences with the company, including non-experiential information such as advertising, and by his expectation whether the company
Page 13 31 July 2012 will deliver to its promises. Satisfied customers have, by definition, better experiences with the company and more trust in the company (Fornell et al.
1996), and are therefore easier to prevent from switching. For companies with higher customer satisfaction, advertising will thus be more effective in reassuring customers.
Although there is not much evidence for a drop in disposable incomes in times of recession (Dhalla 1980; Ergungor and Oliver 2011), consumers become more reluctant to spend money (Shama 1978). Shama (1978) found that over 90% of the consumers belief that a recession results in paying higher prices, and over two-third thinks that prices will remain high in the future. Anderson (1996) shows that satisfied customers are less price-sensitive, and have an increased willingness to tolerate price increases. Companies with high customer satisfaction are thus less inclined to engage in price promotions.
Mela, Gupta and Lehmann (1997) show that price promotions make loyal and non-loyal customers more price-sensitive, while non-price promotions make (non-)loyal customers (more) less price-sensitive. While they find that recessions increase price sensitivity in the loyal segment (recessions do not impact the non- loyal segment), price promotions still result in higher price-sensitivity in the non- loyal segment opposed to the loyal segment. Recessions also reduce non-price promotion sensitivity in the loyal segment (again recessions do not impact the non-loyal segment), however, non-price promotions still result in less price- sensitivity in the loyal segment, and more price-sensitivity in the non-loyal segment.
We thus see that companies with higher customer satisfaction are less inclined to engage in price promotions, and that price and non-price promotions are more effective for companies with high customer satisfaction. Companies with higher customer satisfaction can rely on non-price promotions, which allows them to rely less on price promotions (which will reduce their profit margin), which will result in a higher profit per advertising dollar.
Advertising can also help in acquiring new customers; when consumers have limited resources and their confidence is low, they need much more attention to
Page 14 31 July 2012 become buyers (Bernoff 2009). Bernoff (2009) states that word-of-mouth can be extremely effective in times of recession, and that companies have different strategies to evoke and steer word-of-mouth. Anderson (1998) investigated the relationship between customer satisfaction and word-of-mouth and found a U- shaped relationship. Customers who are either extremely satisfied or extremely dissatisfied engage most in word-of-mouth, with the difference that satisfied customers’ word-of-mouth is positive, and dissatisfied customers’ word-of- mouth is negative. As word-of-mouth becomes more important during recessions, trying to influence and evoke word-of-mouth through advertising will be more effective for companies with higher overall satisfaction, as their customers are more likely to engage in positive word-of-mouth (Anderson 1998).
In conclusion, we assume that advertising will be more effective for companies with (more) satisfied customer as: (i) these companies are better able to reassure their customers, (ii) satisfied customers are less price-sensitive, and (iii) satisfied customers engage in more positive word-of-mouth. We therefore state the following hypothesis:
H3a: In times of recession, advertising will be more effective for companies with high customer satisfaction than for companies with low customer satisfaction
2.2.2SATISFACTION AND R&D IN TIMES OF RECESSION
We discussed that in times of recession, customers want more value for their money, that their habits and preferences change, and that new opportunities may arise in the marketplace. R&D can be an effective instrument to deal with these changes. By modifying and upgrading existing products, companies can add value to these products, and are thus better able to cope with changing customer preferences. Moreover, recessions open up possibilities for breakthrough innovations created by the R&D department.
Much research has been done on the role of customer satisfaction on R&D effectiveness. Satisfied customers are more likely to try brand and category extensions (Srivastava, Shervani and Fahey 1998), they buy more add-on services (Anderson, Fornell and Mazvancheryl 2004), and they engage in more cross- selling (Loveman 1998). Verhoef, Franses and Hoekstra (2001) show that cross-
Page 15 31 July 2012 buying is positively linked to the length of the relationship with the firm, which means that the longer a customer’s relationship with the company, the higher the amount of cross-buying. Moreover, satisfied customers are less likely to switch to competitors when new products are introduced, and less satisfied customers are more likely to switch to competitors when new products are introduced (Lam et al. 2010).
Companies with highly satisfied customers will benefit more from R&D spending in times of recession for the following reasons. First, their customers are more likely to try and adopt incremental as well as breakthrough innovations. In times of recession, customers look for more durable products (Shama 1978), and want more value for their money (Piercy, Cravens and Lane 2010). R&D helps to adapt products to these changing needs, thereby increasing their value.
Second, as less satisfied customers are more likely to switch when new products are introduced (Lam et al. 2010), companies with satisfied customers can more easily attract new customers. We already mentioned earlier that, in times of recession, word-of-mouth is an effective tool to attract new customers. When there is a new product introduction, and less satisfied customers of a competitor are in doubt of switching, positive word-of-mouth of the incumbent firm can give these customers the final push. Given that satisfied customers are more likely to engage in positive word-of-mouth (Anderson 1998), this will lead to higher firm performance for companies with higher customer satisfaction.
Third, satisfied customers are less likely to switch to competitors (Lam et al.
2010), making R&D expenditures of competitors less effective. We therefore state the following hypothesis:
H3b: In times of recession, R&D will be more effective for companies with high customer satisfaction than for companies with low customer satisfaction
2.3FIRM SIZE
Throughout the years, mixed findings have been documented on the relationship between firm size and firm performance. For example, Hall and Weiss (1967) find a positive relationship between firm size and profit, whereas Samuels and Smyth
Page 16 31 July 2012 (1968) find a negative relationship. Scott and Davis (2007) argue that, while firm size might have been a good proxy for profit in the industrial era, this is not appropriate anymore in the postindustrial era. Indeed, Dean, Brown and Bamford (1998) point out that small and large firms are fundamentally different and that they function better in different environments.
Large firms are associated with economies of scale, market power, and experience (Woo, 1987). They have large financial and human resources, a well- known reputation, and strong bargaining power (Dean, Brown and Bamford 1998). On the other hand, they are also characterized as being cumbersome and inefficient (Scott and Davis 2007). Small firms, in contrast, are associated with high levels of flexibility (Dean, Brown and Bamford 1998), fast decision speed (Chen and Hambrick 1995), risk seeking behavior (Hitt, Hoskisson and Harrison, 1991), and targeted innovation (Hamermesh, Anderson and Harris 1978).
Since we are not interested in the main effect of firm size, we do not state any formal hypothesis regarding the main effect of firm size on firm performance.
Instead, we will empirically test this relationship. In the next two subsections, we discuss the impact of firm size on advertising and R&D in times of recession.
2.3.1FIRM SIZE AND ADVERTISING IN TIMES OF RECESSION
There has always been much debate about whether there are true economies of scale in advertising. Chauvin and Hirschey (1993) find that advertising is more effective for large firms, but they also find that advertising is more effective for small firms opposed to medium-sized firms. Arndt and Simon (1983) claim that economies of scale in advertising can arise from two factors. First of all, large firms can be efficient through specialization, better administration and division of labor. Secondly, large firms can use their market power to, for example, negotiate better prices. Moreover, Spence (1980) argues that high levels of advertising expenditures create entry barriers, making it more difficult for new entrants to be successful. While there is some research supporting the claim of economies of scale in advertising (e.g. Brown 1978; Peles 1971), there also exist many studies which find no support for this effect (Seldon, Jewell and O’Brien
Page 17 31 July 2012 1999; Simon 1965; Simon 1969). Hence, the question remains whether there are true economies of scale in advertising or not.
Earlier, we saw that advertising can be effective in times of recession because of an increase in exposure, and because customers need to be reassured. The question is whether this will hold for both small and large firms. Srinivasan, Lilien and Sridhar (2011) argue that high market-share firms already have a high level of customer awareness and market penetration, and that when customers are reluctant to spend money, increasing advertising may increase costs without increasing sales. They also find empirical support for this inference. Although market share and firm size are not identical, they typically are correlated (Chen and Hambrick 1995). We, therefore, expect this assumption to hold for firm size as well.
Simon (1965) finds strong evidence that ‘the first ad is the most efficient, and additional repetitions do less and less work.’ In addition, Seldon, Jewell and O’Brien (1999) state that marginal audiences decrease when advertising expenditures increase. They claim that it is reasonable and rational to spend the first advertising dollars to media with the largest audiences, and that each following advertising dollar reaches a smaller audience. In times of a recession, advertising costs fall, resulting in an increase in exposure per advertising dollar (Bromiley, Navarro and Sottile 2008). As small companies have a smaller total advertising budget than large companies (Chauvin and Hirschey 1993), the increase in exposure per advertising dollar will reach a larger audience for small firms opposed to large firms, and will therefore be more effective for small firms.
In sum, we see that large firms already have a high level of awareness and recognition, making their advertising expenditures less effective when consumers are reluctant to spend money. In addition, because of decreasing returns, small companies will reach a larger audience through an increase in exposure per advertising dollar. We therefore state the following hypothesis:
H4a: In times of recession, advertising will be more effective for small companies than for large companies
Page 18 31 July 2012 2.3.2FIRM SIZE AND R&D IN TIMES OF RECESSION
Under normal economic conditions, R&D is more effective for large firms than for small firms (Chauvin and Hirschey 1993, Klette 1996). The question is whether this effect also holds in times of recession. We already saw that customers’ habits and preferences change, that customers want more value for money, and that recessions give rise to new opportunities. And it is precisely in these areas where small firms can take an advantage. Small firms are better able to remain knowledgeable about changing customer needs (Pearce and Michael 1997), and are more flexible (Dean, Brown and Bamford 1998). They will, therefore, be better equipped to take advantages of opportunities induced by a recession.
Grant (2010) states that competitive advantages emerge from either internal or external sources of change. A recession is an example of an external source of change. Grant (2010) claims that, to grasp the opportunities that come with external changes, two things are essential: the ability to anticipate changes, and the speed with which the firm is able to react on these changes. This corresponds with the already mentioned concept of proactive marketing (Srinivasan, Rangaswamy and Lilien 2005).
Chen and Hambrick (1995) show that large and small companies differ in their behavior. Large companies tend to be risk-averse, and lie under constant public scrutiny. Moreover, because of their size they may feel rich and powerful, which can breed inertia and complacency. Small companies, on the other hand, are constantly seeking for new opportunities and take more risks. Chen and Hambrick (1995) thus conclude that small firms have a greater motivation to act upon changes induced by the environment.
In addition, small and large firms differ in the speed with which they are able to react on changes in the environment. Because of their size, large firms are complex, bureaucratic and exhibit inertia (Scott and Davis 2007). Small firms, by contrast, are more simple structured and flexible (Dean, Brown and Bamford 1998). Small firms will therefore be able to act faster, and capitalize on opportunities induced by the recession more rapidly (Chen and Hambrick 1995).
Page 19 31 July 2012 Chen and Hambrick (1995) conducted their study under normal economic conditions. However, as recessions are turbulent and characterized by change, we expect that their findings will also hold in times of recession, and that, following their logic, small firms might even generate higher returns than large firms.
While there is evidence for economies of scale in R&D expenditures (Chauvin and Hirschey 1993), we assume that small firms can reap more benefits from R&D in a recession. In times of recession, customers’ preferences and habits change, and small firms are more knowledgeable about changing customers’ needs (Pearce and Michael 1997). Moreover, small firms will be more motivated to act upon changes in the environment, they can react faster on possibilities induced by a recession, and execute them quicker. We therefore state the following hypothesis:
H4b: In times of recession, R&D will be more effective for small companies than for large companies
2.4CONCLUSION
In this last section, we introduce the conceptual framework that displays the hypotheses proposed throughout this chapter (figure 2.3).
Figure 2.3 Conceptual framework
We want to emphasize that the main focus of our research lies on the interaction effects of the firm-level contingency variables customer satisfaction and firm size. Note that we use consumer confidence as an indicator of a recession; the
Page 20 31 July 2012 consumer confidence index is positively scaled, which means that a high level of consumer confidence means economic prosperity. As we expect a higher effectiveness of advertising and R&D expenditures in a recession, the arrows from consumer confidence to the effects of advertising and R&D on firm performance are negative.
Page 21 31 July 2012
3METHODOLOGY
We investigate the effects of advertising and R&D expenses on firm performance, while incorporating the economic situation and accounting for the moderating effects of customer satisfaction and firm size. We start this chapter with the specification of our model. We then continue with a description of the data and the variables. We end this chapter with a brief discussion of the statistical program used for modeling. The next chapter deals with the validation of the model and the results.
3.1MODEL SPECIFICATION
To test the hypotheses, we develop a multilevel model (1). It can be seen that the intercept (2), and the effects of advertising (3), and R&D (4) are allowed to vary per company, i.e. they are random. Their effects are explained by (interactions between) sales, customer satisfaction and consumer confidence, therefore, these variables (and their interactions) have fixed coefficients. This is in line with Boulding and Staelin (1995) and Srinivasan, Lillien and Sridhar (2011).
The model is tested for four dependent variables (p), namely: stock returns, EBITDA margin, net income and ROA.
Where,
and i = 0,1,2
(2) (1)
(3) (4)
(5)
(6)
Page 22 31 July 2012
Dependent variable p for firm j in period t;
Relative advertising budget of firm j in time period t;
Relative R&D budget of firm j in time period t;
American Customer Satisfaction Index (ACSI) of firm j divided by its corresponding industry mean in time period t;
Net sales of firm j in time period t;
Conference board's consumer confidence Index® at time t;
The mean of the dependent variable p;
, & The main effects for, respectively, Acsijt, Cct and Salesjt on dependent variable p;
& The two-way interaction effects of Cct * Salesjt and Cct * Acsijt on dependent variable p;
& The main effects for, respectively, advertising and R&D spending on dependent variable p;
, ,
The two-way interaction effects for, respectively, Acsijt, Cct and Salesjt
with advertising and R&D spending on dependent variable p;
The three-way interaction effects of Acsijt * Cct with advertising and R&D spending on dependent variable p;
& The three-way interaction effects of Cct * Salesjt with advertising and R&D spending on dependent variable p;
The level 1 residuals for company j in period t for dependent variable p.
It includes an autoregressive part and a part for random unobserved factors;
The level 2 residuals for i for company j for dependent variable p. i is the intercept, Advertising and R&D spending;
The autoregressive part of the residuals for dependent variable p. If no autoregressive structure is used, this is zero. Values can be found in table 4.1;
The moving average part for lag m of the residuals of dependent variable p. If no MA(q) structure is used, this part is zero. Values for Θ’s can be found in table 4.1;
Random unobserved factors for company j in period t for dependent variable p.
This multilevel method, with repeated measures, is suitable since it allows for the separation of cross-sectional and longitudinal effects (Skrondal and Rabe- Hasketh 2008) and handles missing values appropriately (Snijders and Bosker 1999). Furthermore, it allows us to include additional error terms for advertising and R&D expenditures (Pinheiro and Bates 2000; Snijders and Bosker 1999), thereby allowing companies to have different effects for their marketing variables. By allowing the intercept to vary per company, we allow the companies to have different averages of firm performance. The model is an extension of the model of Boulding and Staelin (1995), whose main purpose was to develop a reproducible model to test for R&D effects on firm performance. In
