innovation during an economic downturn
by
Melvin Waslander
University of Groningen
Faculty of Economics and Business
Msc Marketing
July, 2013
Address: Meerpaal 43
Postal Code, Place: 9732 AC Groningen Phone number: 06 - 48 48 59 16
E-mail address: m.l.waslander@rug.student.nl Student number: 2221675
This paper has attempted to uncover the effect of economic crises on a product's performance and has specifically related this effect to the Dutch car industry's innovation efforts. A set of multiple regression analyses has found sufficient support for the notion that new products are not only beneficial to a product's sales progression, these new products are also disproportionally more affected by competitors' new products than previously launched products. The impact by which new products are affected by competitor's innovative behavior is (in part) a function of the state of the economy, as downturns are associated with higher levels of new product introductions, suggesting an indirect effect of economic fluctuations. In a similar vein, the macro-economy's effect has been further decomposed demonstrating that at least part of its consequences are delivered through primary demand effects. A final finding of this paper is that high brand equity brands are more sensitive to economic fluctuations, as they lose sales at a faster rate during an economic downturn.
Research theme: Role of brand equity and the effects of innovation during economic downturns. Key terms: Industry-level competition, innovations, economic downturn, brand equity, automotive
industry, primary demand, indirect effects.
Groningen, July 2013
As it turned out to be, I managed to complete my Master's thesis. As might be expected, its completion truly feels as a relief, but in the greater scheme of things, it has also offered me great opportunity to use, understand, and deeply think about some of the means that were necessary to transform abstract, unordered data into tangible statements and conclusions.
First and foremost, I would like to thank Dr. Ir. M.J. Gijsenberg for his support and incredibly helpful contributions, as well as Dr. H. Risselada for additional directions on how to solve some of the issues I had faced.
1 INTRODUCTION...3
2 LITERATURE REVIEW...7
3 METHODOLOGY ...21
4 RESULTS...25
1 INTRODUCTION
Experts predict that the number of smartphone users worldwide will hit the 1.5 billion mark somewhere in 2016. Considering that the first predecessor of the currently-known smartphone already goes back to 1992(Wikipedia, 2013),the device's innovative trajectory has come a long way to find a seamless alignment with consumers' needs, now that smartphones have truly become a common good. Perhaps the most remarkable about the product's increasingly rapid success, is that it is in sharp contrast with the gloomy economic situation of the recent and ongoing years.
Reoccurring figures on economies across the globe reported by the media, show little to no signs of real recovery, with unemployment rates steadily going up, and numerous firms reporting
disappointing figures on their commercial performance.
This raises the question how the apparent success of one type of product or industry, relates to the dramatic fall of general economic activity which seems to hurt the majority of other
industries. It makes one wonder, why is there a number of industries in which almost every firm has tremendous trouble of getting by, whereas some firms operating in different industries seem
apparently unaffected by the economic turmoil, and demonstrate record-breaking sales figures and impressive financial performances(e.g., Nu.nl, 2013; Coevert, 2013).
Some claim that at least part of the answer lies with innovation, and notice how firms need to innovate rather sooner than later, in order to provide benefits to themselves, to their respective industry and eventually the economy as a whole(e.g., Greenhalgh, Rogers, 2012; Schumpeter, 1992).The only way out of a recession would therefore involve more investments on part by the firms in R&D, so that a larger portion of consumers will decide to buy their products again.
Although customers' needs play a central role in marketing, some argue that there are more than one route to success for a firm; for instance, O'Cass and Ngo have studied the circumstances under which taking less account of the needs of customers may actually have beneficial aspects to a firm's performance (O'Cass, Ngo, 2007). Taking the needs of consumers strictly into consideration, is referred to as a market orientation (Baker, Sinkula, 1999), and is related to the ability to
Indeed, one of the true advantages to innovation from the perspective of the consumer, is that it results in means (i.e., products, services) that have the potential to fulfill needs in unprecedented ways. Oftentimes, the needs that consumers are currently aware of, are but a tip of the iceberg. It are the latent needs – the needs consumers are not aware of – that offer a great deal of potential to products or services (Van Kleef, Van Trijp, Luning, 2005).Moreover,consumers are frequently not fully aware of the exact motivational drivers for preferring one product over the other(Simonson, 1993; Steenkamp, Van Trijp, 1997). For this reason, it is not surprising that there are numerous examples where consumers were truly surprised by new products that only demonstrated their use and value, after they were introduced to the market (Griffin, Hauser, 1993). The primary objective of new product development, remains to deliver a product that has high value to consumers, relative to alternative or existing products. This is mostly achieved by fulfilling consumers' (latent) needs (Slater, Narver, 2000).
As a consequence of the positive impact new products may have on consumers, firms that have a strong focus on offering new products, were found to perform well. Where some (e.g., Zhou, Yim, Tse, 2005)found a general positive relation between innovations and a firm's or product's performance, others (e.g., Sorescu, Spanjol, 2008) observed that especially “breakthrough” innovations are responsible for generating abnormal returns to a firm.
In the light of an economic crisis, it would be interesting to see to what extent economic conditions influence a firm's ability to remain innovative however. Since during an economic contraction, demand tends to drop faster than supply does, (Steenkamp, Fang, 2011)firms are forced to take adequate measures. Failing to do so, might soon lead them into financial problems, or even put a firm's existence in jeopardy. As a result, one of the most common reflexes for firms is to cut costs on activities that are considered to be secondary. The most notable example of activities that are among the first to be scrapped, are business processes linked to marketing (Deleersnyder, Dekimpe, Steenkamp, Leeflang, 2009; Mizik, Jacobson 2007). In a similar vein, firms are also tempted to slim down their budgets for R&D in the event of an economic contraction (Barlevy, 2007).
Not only firms are affected by deteriorating economic conditions; under the influence of economic contractions, consumers are also increasingly reluctant to spend money.
During recessions consumers might be directly affected because of a lower disposable income, for instance by becoming unemployed. Those consumers who might be little affected by tough
Expenditures are scrutinized and products and services that are not considered to be necessities are bought at lower frequencies. Also, since consumers generally have lower disposable incomes during economic contractions compared to economic expansions, people tend to become more price
sensitive (Estelami, Lehmann, Holden, 2001).This in turn, implies another source of pressure on firms' financial performance when the economy finds itself in a downward trend.
An aspect that could soften the sharp edges for firms a bit however, comes in the form of brand equity. People may appreciate a brand to such an extent that their feelings towards it, mimic that of a real-life relationship (Fournier, 1998). Brands for which this is the case, might suffer fewer negative consequences of an economic downturn. Since brands with high brand equity are
perceived to be different in a positive way, they are appreciated more than brands with low brand equity. Consumers are also willing to pay a higher price for brands with high brand equity
compared to low-equity brands (Chandon, Wansink, Laurent, 2000). Moreover, buyers of high-equity brands are also willing to invest more time or effort in order to acquire them, which is another indication of higher customer loyalty for high-equity brands relative to low-equity brands. Lassar et al. noted the link between higher brand equity and a higher level of confidence an
individual puts in a brand(Lassar, Mittal, Sharma, 1995). High brand equity brands are thus seen as more reliable and facilitate customer commitment. A possible benefit of stronger brands could therefore be that they help to mitigate some of the risks associated with buying products (Fischer, Völckner, Sattler, 2010). This is particularly relevant during economic downturns, because products with high brand equity might also be less affected by a slow economy, which is accompanied by greater uncertainty about people's financial future and the consequential, greater risk awareness for consumers.
Although the impact of economic crises on consumer behavior would seem noteworthy, little research has been done on the subject (Dutt, Padmanabhan, 2011). One explanation for that, would be that economic crises come and go, and their effects only last for a relatively short period of time, so that they hardly affect consumers (Friedman, 1957). In that sense, consumers would not only base their consumer decisions on their current income, but instead take into account the value of their future expected income. From this perspective, economic crises would merely have
temporal, relatively superficial effects.
An important aspect that is noteworthy about the current study, is that it will try to assesses the effects and effectiveness of R&D during economic downturns, about which little is still known (Steenkamp, Fang, 2011).The same holds for the effects of brand equity during economic
specifically relating this aspect to a product's performance.
Further insights are aimed to be provided by not only considering the effects for R&D on a single product in isolation, but in the context of a dynamic, competitive environment. Since both the perspective of firms will be investigated when it comes to innovative behavior, as well as the effects on consumers through the numbers on sales, this study is one of the few attempts to examine new product performance on an industry level, in the context of that competitive industry.
Because not all industries are equally affected by economic crises (Deleersnyder, Dekimpe, Sarvary, Parker, 2004),and furthermore, to limit the scope of research for this study, this study will primarily focus on durable products. Given that, the central theme of this paper is formulated as follows:
What are the positive and/or negative effects for firms of durable products to launch product innovations during an economic downturn, and what is the exact role of brand equity for a firm during an economic downturn?
In order to provide an answer to aforementioned central themes, this study will evolve around a number of research questions, namely:
– What is the effect of innovation on a durable product's sales?
– To what extent can innovation mitigate the negative effects of economic downturns? – How do the innovation efforts of one firm relate to its competitors in terms of sales? – How does the state of the economy affect a firm's innovative behaviour?
– What role does brand equity play in the performance of durable products during economic
crises?
The basic means to arrive at the final conclusions have been multiple regression analyses on data containing the sales figures and innovation measures for the Dutch automotive market.
2 LITERATURE REVIEW
State of the Economy
Definition of the business cycle. A business cycle refers to the fluctuations in economic
activity, where periods of economic expansion are both preceded and succeeded, by economic contractions as part of a continuous loop (Mitchell, 1927; Morley, Piger, 2012).Economic activity is expressed as GDP (Gross Domestic Product), which in turn refers to “the total market value of all
final goods and services produced over a given time span” (Kuznets, 1934).This would mean, that during times of economic contraction, the market value of all final goods and services follows a downward trend. As such, the term recession is a formalization of aforementioned, so that it is defined as two (or more) consecutive quarters during which the GDP demonstrates a negative growth rate (Moneta, 2005; Moersch, Pohl, 2011).
The views on what exactly causes the economy to expand and contract are numerous and are clearly beyond the scope of this study. Reflecting on the crucial role of credit that hailed the most recent economic crisis that started in 2008, two views are particularly interesting however. The one considers firms, the other view considers consumers as the primary driver of economic fluctuations or business cycles (Wunder, 2012).
Firm-driven business cycle. One view on business cycles, is that they are primarily
driven by firms who turn to larger portions of debt-financing as long as economic expansion continues (Wunder, 2012). This causes the value of assets to go up, because loans allow firms to acquire the assets that they could not have afforded otherwise. As a direct consequence, the
effective demand increases, the prices of assets go up and firms would therefore need more capital to successfully finance these assets.
From the perspective of the providers of capital (e.g., banks, financial institutions),
The final consequence is an economic contraction or a recession (Chen, Kontonikas, Montagnoli, 2012; Hyman, 1986),that will last until the economic situation stabilizes and a similar sequence of events starts over again.
Consumer-driven business cycle. A different view on the business cycle holds that business
cycles are primarily driven by consumer spending (Yerex, 2011).Credit would play a similar role in comparison with a firm-driven business cycle, except that consumers increasingly turn to credit for
consumptionpurposes. Once consumers are beginning to become unable to support the burden of
their debts (e.g., mortgages) they have no other option but to sell off their assets (e.g., houses). Here too, the number of defaulters pushes up the interest rates to compensate providers of capital for their loss of returns, due to debts that cannot be repaid. The higher interest rates will then serve to put a brake on the increase in demand for credit-funded assets-acquisition. Contemporaneously, the portion of consumers that has to sell their assets for liquidity purposes increases as well, which sets off rapidly dropping asset prices. The final outcome is an economic downturn (De Antoni, 2010).
Godley and Wray observed that most of the growth of aggregate demand in the late 1990s was accounted for by consumer spending (Godley, Wray, 2000). Moreover, Garnerfound evidence that by studying consumer debt measures, economic downturns could be predicted (Garner, 1996). As such, an increasing number of studies have found empirical evidence for the notion that it are in the first place the consumers, who actually have sufficient critical mass to tilt the economy towards
economic bloom or economic contraction(Wunder, 2012).
Consumer Behavior and Economic Downturns
Schipchandler noted that consumers tend to adjust their consumer behavior in response to an economic crisis, by becoming more reserved when it comes to their spending (Schipchandler, 1982). Some of the means by which they do so, includes consuming energy resources more
consciously. When consumers face deteriorating economic conditions, they also have a tendency to delay their purchases (Bayus, 1988), and owners of durable products make a more deliberate decision to extend the lives of their products, by having them repaired instead of replacing them with a new one.
A related change in the behavior of consumers, could therefore also be that they reallocate their expenditures, but only within categories (e.g., less money spent on one type of durable product, but more on another product that is just a different type of durable product) (Dutt, Padmanabhan, 2011).
Conspicuous buying. Quelch and Jocz state that some product categories are less affected
than others (Quelch, Jocz, 2009),and provide the example of necessities that are being less affected by crises. What constitutes a necessity however may considerably vary per country as well as per individual (Dutt, Padmanabhan, 2011).A striking example is that despite the economic crises, some luxury brands appear to perform better than ever. Moreover, these brands manage to successfully introduce products that are even more expensive than their former ones (Nunes, Dreze, Han, 2011). The luxury brands are likely benefiting from a phenomenon referred to as conspicuous consumption – “the extravagant spending on products intended chiefly to display wealth and thus signal status” (Veblen, 1899). The reason why conspicuous spending is especially relevant during economic downturns, could be explained by consumers' intrinsic motivation to hold up appearances during economic crises; certain people will have a desire to convey a sense of wealth and success, even though the negative effects of a slowing economy are truly felt by them (Bagwell, Bernheim, 1996). Because there is a strong social component involved in conspicuous consumption, the products involved are preferably public goods. These products can be easily displayed in public with a maximum effect of demonstrating one's wealth (Hoyer, MacInnis, Pieters, 2013; Nunes et al., 2011).
A different factor that influences consumers' response to an economic crisis, but one that is still a bit related to conspicuous consumption, is the level of materialism found in a consumer. Materialistic consumers value money and the possession of products as they see them as a way to achieve personal happiness and social progress (Moschis, Churchill, 1978). Since materialistic individuals are so concerned with the possession of products, they are also more reluctant to adjust their expenditure patterns during economic contractions (Richins, Rudmin, 1994). As a
consequence, consumers that are relatively more materialistic, are also personally more affected than less materialistic individuals, given the greater discrepancy they experience between a desired lifestyle and the lifestyle that they can afford (Ang, 2001).
Risky behaviour. A factor that plays an important role during economic crises, relates to how
individuals perceive risks and choose to act on them. Risk averse behavior is typically when someone is more likely to attain a smaller (but more likely) gain, if that would effectively reduce the likeliness of ending up with nothing (Shead, Hodgins, 2009).An example that is related to economic downturns, is when employees become more hesitant to apply for a different (perhaps better-paying) job, and would rather stick to their current job. Improving their performance at their current job and avoid getting laid off, is thus preferred over the prospect of a higher income,
because the uncertainty associated with the new job is perceived to be more relevant than the related potential benefits (Heyes, 2011; Kahh, Meyer, 1991).
Likewise, during an ongoing economic contraction, individuals are also expected to spend money more cautiously, in an effort to limit potential (financial) losses. Thus, less emphasis will be put on the purchasing of new products (and their potential benefits), due to the higher perceived risks, as a function of the general uncertainties related to poor economic situations.
This effect could be moderated by an individual's personal propensity for risk-seeking behavior (Price, Ridgeway, 1983; Schwartz, Hasnain, 2002; Weber, Milliman, 1997), since risk-averse individuals are even more likely to cut their expenditures during economic contractions than less risk-seeking individuals. People that are relatively risk-averse will also take the financial risks associated with buying products into account more, so that they will be interested more in products that are cheaper, including (lower-priced) do-it-yourself versions of a product (Ang, 2001).
An additional argument why people could become more risk-averse during recessions, could be found in the notion that economic downturns tend to lead to increased levels of stress (Hughes, Dennison, 2009). An implication of stressful conditions is that individuals respond to risks more instinctively and intuitively and link risk to a sense of danger, which they prefer to avoid (Slovic, Finucane, Peters, MacGregor, 2004).In fact, people with negative moods underestimate favorable outcomes and overestimate unfavorable outcomes (Nygren, Isen, Taylor, Dulin, 1996; Johnson, Tversky, 1983). This causes them to have an outlook on the future that is probably more grim than justified. This, in turn, could cause people to be even less willing to spend money on products, that are not absolute necessities (Ang, 2001),such as luxury goods and durable products.
One of the fundamental reasons why consumers purchase fewer durables in particular during economic contractions, is reflected by the negative changes in income level during economic
Deleersnyder et al. have also noted that sales for consumer durables do not necessarily have to fall more drastically than non-durables, although expenditures on the former, do seem to be affected at an earlier stage during the economic contractions and recover more slowly during economic expansions (Deleersynder et al., 2004). A number of additional reasons why durables are more affected by economic downturns, include that, consumers have made shopping for fast moving consumer goods into a habit. Consumers find it difficult to change their fixed routines and are thus reluctant to cut back on non-durables (Katona, 1975). As a result, individuals scrutinize
expenditures on durables more, since these are less associated with habitual behavior. Consequently, consumers will be more likely to make deliberate choices to purchase durable goods or not, so that durable goods are more likely to be postponed in case of an economic downturn (Cook, 1999). A final reason why durables are often more affected, is that durables are relatively expensive, making them less affordable (Katona, 1975), which implies that consumers would relatively often have to buy durable goods on credit (Petersen, Strongin, 1996). The finding that people become more reluctant to close loans during economic contractions (Ang, Leong, Kotler, 2000; Stock, Watson, 1999) therefore negatively influences the ability and willingness to buy durable products. Based on the described theoretic contributions, the following hypothesis is postulated:
H1 As a consequence of an economic downturn, on an industry level, consumers will purchase fewer durable products.
Marketing literature has noted two separate sources for changes in an individual product's sales, for instance, as a result of price promotions (e.g., Van Heerde, Gupta, Wittink, 2003): On the one hand, sales can increase as a result of a greater overall demand for a product (primary demand). This occurs when consumers purchase a certain type of product rather sooner than later, or decide to buy larger quantities of these products. On the other hand, a brand can also increase its sales at the expense of competitors (secondary demand), which is the case when consumers switch brands (Bell, Chiang, Padmanabhan, 1999).Primary demand increases have the benefit that they still allow sales growth, despite potential losses of market share (Nijs, Dekimpe, Steenkamp, Hanssens, 2001). Conversely, lower levels of primary demand will generally bear a negative effect on the sales of individual products. Based on this notion the following effects are being considered:
H3 The effect of an economic downturn on the sales of a single brand of durable products, is mediated by the number of units sold on an aggregate level.
Innovation and Economic Downturns
Pro-cyclical innovative behaviour. During economic downturns, when firms have to
compete over a scarcer pool of resources (provided to them in the form of revenue through their customers), firms tend to look for ways to cut costs and save financial resources. As a consequence, they tend to slim down budgets reserved for R&D projects and reconsider their innovation strategies (Barrett, Musso, Padhi, 2009). Saving resources by cutting budgets for R&D is an especially
convenient measure, since R&D can only generate cash on the long run, but will consume resources on the relative short term. As a consequence, managers are relatively eager to scrutinize the
financial implications associated with R&D first, in the face of necessary cutbacks (Cincera, Cozza, Tübke, Voigt, 2012).
Especially when economic downturns coincide with a financial crisis, firms have a harder time getting access to external capital (Barrett et al., 2009). This will also have firms trim down their budget available to R&D, so that firms would rather postpone any major investment in innovations until the conditions for attracting capital improve (Paunov, 2010). Therefore, a
substantial number of firms decides to delay new product development projects or even scrap such projects altogether. Archibugi and Filippetti found empirical evidence that the average amount spent on innovation by firms within the EU, has significantly decreased in response to the recent financial crisis of 2008 and continues to be the case in the years afterward (Archibugi, Filippetti, 2011).
Oppositely, during times of economic expansions, R&D budgets are expanded and innovative projects are initiated. Scholars (e.g., Shleiffer, 1986; Francois, Lloyd-Ellis, 2003) have found that firms have a tendency to act pro-cyclically and would rather invest during economic blooms and disinvest during economic downturns. With respect to innovation, new ideas and projects are initiated during an economic upswing, whereas they are shelved during economic bad weather.
Counter-cyclical innovative behaviour. Taking into account a firm's production capacity,
firms may be more encouraged to launch new products outside of economic blooms however; when demand is relatively lower, it is noted, the excess amount of production capacity and staff can be directed to R&D activities and production for new products with much more ease. The great benefit of doing so during economic contractions, is that there is less interference with the business
processes associated with meeting a firm’s current demand (Devinney, 1990). This causes that R&D activities can be performed more cost efficiently as well. Considering economic blooms, the
associated higher demand for existing products would make it necessary to hire extra personnel or acquire additional assets to complete R&D efforts. In sum, it is concluded that the costs associated with producing new products are generally lower under conditions of economic contraction.
Still, when the economy expands, the market has a greater absorption power when it comes to the buying of new products (Judd, 1985). Firms would be more appealed to this aspect of economic fluctuations and therefore more likely to launch new products during economic blooms; the expectation of a higher demand for a new product, simply allows a new product’s sales
prognoses to be brighter-looking. An extra reason why firms have special interest in this, is because they wish to get a maximum return on their R&D investment in the shortest time frame as possible (Shleifer, 1986). Launching new products during economic contractions, offers less prospect to arrive at that objective. As such, the effect of an economic contraction is summarized into the following hypothesis:
H4 As a consequence of an economic downturn, on an industry level, firms that produce durable products will engage in fewer R&D projects, which results in a smaller number of competitive new product launches for a single brand of durable products.
Innovation as a means of competition. Firms that decide to launch new products during an
During an economic downturn, innovating will typically help firms to counter the negative effects of the tight conditions, and help increase a firm’s market share at the expense of its competitors (Hartl, Herrmann, 2006). In a more general sense, firms tend to compete with one another by innovating, resulting in the launch of new products with better qualities or upgraded functionalities (Schumpeter, 1992). Greenhalgh and Rogers consider innovation to be a means of competition as well (Greenhalgh, Rogers, 2012), and found that innovation offers a premium to firms that do, albeit relatively short-lived. The way innovation is depicted here is that it allows firms to initially steal profit away from other firms, but sooner or later, other firms will replicate the innovator’s success by introducing versions of their own that are considered to be the equivalent to the original innovation. Although this clearly happens at the expense of the original innovator, it is considered to be beneficial to the industry on the long term, as it helps improving overall industry profitability (Greenhalgh, Rogers, 2012).
These accounts are in line with Schumpeter's notion of creative destruction (Schumpeter, 1939), as it is believed that firms that employ innovative projects, must have a prospect of reaping the fruits from their actions. Therefore, the likeliness of taking part in R&D projects, is assumed to be higher under conditions where there are only a relatively small number of competing firms (Peroni, Gomes Ferreira, 2012). It implies that once the number of competitors increases, firms will generally become more hesitant to launch new products.
Others dispute this view, and propose that innovation is a means to differentiate the offerings of a firm relative to those of competitors. It is therefore believed that higher levels of competition will provide incentives for firms to engage in more innovative behavior (Polder, Veldhuizen, 2012). In an attempt to merge both conflicting views, it was therefore suggested that competition may actually have two effects on innovation working at the same time: on the one hand it spurs firms to innovate so that they can escape competition, on the other hand high level of competition may hurt the profit potential. This observation led to the notion of an inverted U-shaped relationship between competition and innovation (Aghion, Bloom, Blundell, Griffith, Howitt, 2005).
Castellacci noted that firms in competitive industries could leverage their own innovative behavior more effectively compared to less competitive industries, because firms can benefit from the imitation of advanced technologies (introduced by others) to their own benefit by
Aforementioned indicates that a firm's sales are substantially influenced by the innovative behavior of other firms. Moreover, products that are recently introduced are very much affected by products that were launched over the same period of time; imitator products will take away from an earlier innovation's uniqueness and the former's differentiating qualities as competing product
introductions serve as direct substitutes. The observation that firms can leverage their own innovative efforts as a function of other firm's product launches, next suggests a learning effect where knowledge and capabilities are accumulated with each consecutive product introduction. Also, successive product introductions would be an effective measure to guard off imitators as they will have a hard time keeping up with a frequent stream of novel products. Extending on this, the following set of hypotheses is proposed:
H5 Higher levels of competitive new product launches have a negative effect on the sales of a single brand of durable products.
H6 The effect of an economic downturn on the sales of a single brand of durable products, is mediated by the number of competitive product launches.
H7 A higher number of new product launches associated with a single brand of durable products, has a positive effect on its sales.
H8 The negative effect of a higher number of competitive product launches on a single brand of durable products' sales is positively moderated (i.e., becomes stronger), when that brand is
associated with a higher number of own product launches.
H9 The negative (interaction) effect of new product launches by a single brand of durable products, with a higher number of competitive new product launches, is negatively moderated (i.e., becomes weaker) as that brand has previously launched a higher number of consecutive new products.
Consumer Behavior and Innovation
Whether someone purchases a new product or not, is next dependent on an individual's intrinsic innovativeness, affinity with a particular type of product and additional influences that are situational in character (Foxall, Bhate, 1991).
When consumers were previously unfamiliar with new types of products or alternative brands, consumers are relatively hesitant to purchase them, due to the higher perception of risks and uncertainties that are associated with a new product. In that sense, four types of risk (Ram, Sheth, 1989)are especially relevant for innovations, which include financial risk, the risk of getting physically hurt by using the product, functional riskand social risks that relate to whether others will disapprove the purchase of a new product. The number of risks involved with purchasing and consuming products imply that when – for one reason or another – the perceived risk for a new product is considered to be higher, the rate of diffusion suffers and the product will less likely be successful (Rogers, 1995; Sheth, 1981).
Variety seeking. Apart from the associated risks, new products also offer means of variety to consumers. To what extent someone would appreciate variety is the result of a person's optimum stimulation level (OSL), which refers to how much stimulation an individual seeks for, in his or her life (McReynolds, 1971). People who feel that they need relatively much stimulation in life, tend to have a resistance to habitual behavior and overly fixed patterns. New products offer people with a high optimum stimulation level a good opportunity to experience new things (Burns, 2007; Seetharaman, Chintagunta, 1998).According to this view, for some people the higher risks associated with innovations are but a small sacrifice to the novel sensations and experiences that new products have to offer (Zuckerman, 1979).
Product adoption. Rogers identified five types of consumers, each with a different
Although consumers indicate that they feel a need for products with high levels of sophistication that in turn facilitate a better match with their needs (Vander Schee, 2012), in actuality, consumers mostly purchase innovations that are less complex (Arts et al., 2011).
Consumers and innovations during economic downturns. Downturns may offer
opportunities to companies, as Pearce and Michael have observed that recessions cause the current customer-firm relationship to be scrutinized and evaluated (Pearce, Michael, 1997). This would imply that firms would have to re-establish their current relationship with their customers, and also that competing firms have better openings on capturing market share away from other companies. Compared to economic blooms, customers may therefore be more likely to change firms from
which they buy their products during economic downturns (Cabalerro, Hammour, 1994).
Zhou found indications that when it comes to new product launches, innovative strategies are relatively more successful than imitation strategies and that innovation strategies appear to do better under conditions of demand uncertainty (Zhou, 2006). Innovative strategies were also found to be more beneficial compared to imitation strategies under relatively competitive circumstances. Given that economic downturns are characterized by demand uncertainty and tough competition, innovative strategies may thus help to mitigate the negative effects of an economic contraction. Further indications why this could be true are provided by Geroski et al. who found a general positive relationship between the number of innovations and a firm's profitability (Geroski, Machin, Van Reenen, 1993). More so, innovative firms were less affected by the negative effects of
Finally, new products could generate higher customer loyalty during phases of economic
contraction compared to economic blooms. This is, in part, because products that are introduced during economic contractions have consumers to infer qualities of robustness on these firms – i.e., new products that have been made available despite of the bad economic weather must be
originating from a superior company (Pearce, Michael, 2006).These observations lead to the following hypotheses:
H10 The negative effect of an economic downturn on a single brand of durable products' sales is negatively moderated (i.e., becomes weaker), when a larger number of new product launches is associated with that brand.
H11 The positive (interaction) effect of new product launches by a single brand of durable products, with an economic downturn, is positively moderated (i.e., becomes stronger), when that brand has previously launched a higher number of consecutive new products.
Consumer Behavior and Brand Equity
Brand equity is defined as those differentiating outcomes and effects of brand-specific marketing efforts, that allow consumers to tell the brand apart from other brands (Keller, 1993). As a result of one brand's marketing efforts, a consumer may form positive (or negative) opinions about it. This will ultimately have an influence over an individual's behavior toward that brand. If positive
associations have been formed over the years through a series of advertising campaigns, a consumer might develop a favorable evaluation of a brand, and ultimately decide to purchase that particular brand (again). This might not have been the case, if it were not for that specific brand and the unique associations someone has formed about it, due to the marketing efforts that this brand has employed.
One of the implications for brands with higher brand equity is that consumers tend to be less price sensitive towards them, and that people are more willing to pay premium prices for high-equity brands. (Ailawadi, Lehmann, Neslin 2003; Keller, 1993).Therefore, in the situation where a brand ideally enjoys the benefits associated with high brand equity, this would result in higher revenues, lower costs and better profit prospects.
A way to quantify the significance of brand equity has been proposed by Simon and Sullivan. They roughly explain brand equity as the discounted value of future earnings that are associated with one particular brand, that are specifically related to that particular brand, and would not have been earned, if it were a different brand (Simon, Sullivan, 1993).
From the perspective of consumers, brands with higher brand equity are perceived to have superior qualities compared to other brands in a product category. As such, the beneficial aspects for a brand with high brand equity, are a direct result of a positive evaluation of a brand and could eventually result in a form of customer loyalty or an augmented level of involvement with the brand (Venkateswaran, Binith, Geetha, Ananthi, 2011).According to some (e.g., Fournier, 1998),the way an individual interacts with a brand could even take the form of a human relationship, so that emotions of affection are felt for a brand. As such, it is expected that high-equity brands are less affected by economic downturns so that the hypothesis is stated as:
H12 The negative effect of an economic downturn on a single brand of durable products' sales is negatively moderated (i.e., becomes weaker), when that brand has high brand equity, compared to low brand equity.
The proposed hypotheses of this paper and expected relationships are summarized and graphically depicted in figure 1 on the next page.
It should be noted that the directions (i.e. positive versus negative effect) of the interaction effects can be interpreted as such that a negative moderation causes the lower-order relationship to become weaker (e.g., negative main effect becomes less negative) or switch direction (e.g., negative main effect becomes a positive main effect). Positive moderation causes the lower-order
FIGURE 1 Conceptual model
* denotes the hypothesized mediation effects
-H1 H4
-+
-High versus low Brand Equity (W) Competitive new product
launches on industry level
(M1)
+
-H11 H9 H8 H12 Number of consecutive own new product launches
(Q)
+
-Number of products sold on industry level (M2)
H3*
Number of own new product launches (V)
+
H10-H5 H2 H6* H7
3 METHODOLOGY
Data Collection and Sample
In order to test the hypotheses, and for practical purposes, the scope of research assessing the effect on durable products, has been further confined to the automotive market in the Netherlands. The car industry presumably has several benefits, as it is a relatively important industry from an
international perspective and it has few idiosyncrasies compared to other industries. Therefore, the findings for the automotive industry are believed to be generalizable to other industries as well (Srinivasan, 2009).
The dataset has been composed from raw data that have been made available through the website of Dutch car magazine Autoweek (autoweek.nl), which has been proven to be a source of data for at least one previous study as well (Bakker, Van Lente, Engels, 2012). The used data are publically available and consist of yearly figures on car sales per manufacturer and car brand model. The data have next been aggregated and recoded to allow for statistical analyses.
Dependent Variable
The number of sales in the Netherlands for an individual car brand model was used as the
dependent variable, because it is probably the most meaningful and most straightforward measure that has been available to assess a brand's performance. Car sales have been recorded on a year-to-year basis so that the analyses also follow a year-to-yearly pattern, over the period 1983-2012 and were taken from the Autoweek's website.
Independent Variables
Economic downturns. Economic downturns and upturns have been operationalized by using
the fluctuations of GDP in one year relative to the previous year for the Netherlands (percentage of growth/decline relative to a previous year). The data were obtained via the website of the Dutch Central Agency for Statistics, also known as the CBS (cbs.nl). Moreover, to consider the effect from the perspective of economic downturns, as opposed to economic upturns, the GDP fluctuations were negated. This way, higher values reflect a more negative economic situation, whereas lower values indicate relative economic prosperity; somewhat similar to a dummy variable, but only without losing data on the relative magnitudes of the fluctuations.
Competitive new product launches. The number of competitive launches over the timespan
First of all, separate car brand models have been identified and recorded with the associated number of respective launches a car brand model was responsible for in a given year (1983-2012). In the event that the main source of data provided no record on a car brand model's respective year of launch onto the Dutch market, but was still shown to sell a minimum number of cars in one or more consecutive years, the year of launch was reconstructed based on additional sources (e.g.,
Wikipedia.org). Because the data did not allow a strict discrimination between what constitutes a really new version, an updated version or a minor update like a facelift (e.g., Herzenstein, Posavac, Brakus, 2007),the level of innovativeness of a car brand model has not been taken into account beyond whether the car brand model had been new to the market or not.
Total number of cars sold by the automotive industry. The total number of cars sold refers
to the total number of car units sold over a year in the Netherlands for any given year from 1983 to 2012.
Number of own car launches. The total number of car launches one specific car brand
model has been responsible for, is the result of the same procedure used with the variable for 'competitive launches'. The computation of the number of car brand launches in a given year is the result of the total number of product introductions related to a car brand model, expressed by an integer number ranging from zero (i.e., no associated car launches in the observed year) to one, two, or more associated product introductions.
Consecutive number of own car brand model launches. The consecutive number of car
brand launches estimates the total number of car brand model launches in the years prior to an observed year. As such, this variable will not only reflect whether a car brand model is new to the market (i.e., a zero expresses that a car brand model is considered to be new to the market), it is also a measure that takes into account how often the car model has been updated already.
Brand equity. Because it will be practically impossible to reconstruct the brand equity of a
brand say, 15 years ago, and the overall difficulty with assessing brand equity in a more qualitative way(e.g., Aaker, 1997),this study has chosen to assess brand equity in a more quantitative way and attempts to link brand equity to future commercial benefits(e.g., Simon, Sullivan, 1993).
This approach in particular, also takes into account that new car brand models that did not exist before, will still need some time to establish a level of brand equity. Time they will need in order to minimally qualify for one of the prerequisites to brand equity, namely brand awareness. Updates to already established car brand models by contrast, arguably have already managed toestablish some level of brand equity, so that the operationalization of brand equity should be able to capture that effect as well. Possible downsides to this approach will be discussed in the section “limitations”.
To account for the potentially misspecified effect on the variable of brand equity, a variable controlling for the number of years since a product's introduction/update, was included in the analyses.
Interaction Effects
The procedure to assess the interaction effects, has included mean-centering the original variables, followed by a calculation of the corresponding product terms for the variables involved,
respectively. Mean-centering helps reducing the negative effects as a result of multicollinearity, which is usually an issue when dealing with product terms consisting of two or more variables (e.g., Dunlap, Kemery, 1987).
Control Variables
Car types. Different car types are expected to be associated with different sales patterns; for
example, more luxurious cars such as cabriolets (Nayum, Klöckner, Prugsamatz, 2013) are expected to be relative niches in the market and therefore significantly deviate from more mainstream car types (e.g., hatchback) in terms of unit sales. To account for these differences, dummy variables were used to indicate whether a car brand model is associated with one car type or the other, or any random combination among those. The car type categories that have been identified include: hatchback, sedan, stationwagon, coupé, cabriolet, MPV, SUV and a remaining category consisting of minivans and cross-overs for example.
Years passed since most recent product introduction. As mentioned before, given the
The variable is referred to as the years that have passed since the most recent update. If the car has not been updated since its first introduction, (and only if this condition is met) the variable refers to the number of years since its initial launch.
Multiple Regression Analyses
The hypotheses were tested as part of two separate multiple regression procedures making use of the SPSS statistical software program. Testing the mediation effects was done by making use of the PROCESS-macro that was made available by Dr. A. Hayes. The PROCESS-macro makes use of a bootstrapping resample technique that generates 5000 bootstrapping samples (Preacher, Hayes, 2008) and provides an estimate to the 95% confidence interval of the mediated effects, testing their respective statistical significance.
The second procedure included a (hierarchical) moderation regression analysis to estimate the interaction effects with the advantage of assessing relative model fit, for each block of variables that was entered into the regression model. To prevent multicollinearity issues from diffusing the results (as much as possible), all independent variables (in addition to the continuous control variable) have been mean-centered (Aguinis, Gottfredson, 2010).The descriptives to the dependent and independent variables and control variables that were included in the models are summarized as part of table 1.
TABLE 1
Descriptives to independent variables and control variables
mean median st deviation minimum maximum
sales 2509.7 782 4216.9 1 57431 hatchbackₐ - - - 0 1 sedanₐ - - - 0 1 stationwagonₐ - - - 0 1 coupéₐ - - - 0 1 cabrioletₐ - - - 0 1 MPVₐ - - - 0 1 SUVₐ - - - 0 1
other types of carsₐ - - - 0 1
years since launch or most recent update 2.5 1 4.1 0 30
total number of competing car brand model launches 115 123 29.4 35 156
total amount of car units sold 494739.8 490407 47737.5 389131 611776
number of launches for focal car brand model current year .4 0 .8 0 7
cumulative own launches 3.7 2 4.6 0 34
brand equity 22672.2 3536 50850.2 0 584773
GDP fluctuations -1.8 -2 2 -4.7 3.7
4 RESULTS
Mediation Effects
A multiple regression model was estimated as part of Dr. A. Hayes' PROCESS-macro using SPSS (Preacher, Hayes, 2008), allowing the testing of the proposed mediation effects. The independent variables in the model managed to explain 48.7%of the variance (R²=.487, F(22, 5858) = 252.80, p < .01).
1st mediation effect – indirect effect of economic downturns on car sales via competitive
launches. To qualify for a statistically significant mediation effect, three conditions must be met
(Judd, Kenny, 1981; Baron, Kenny, 1986). The first condition requires that the independent variable
must have a significant effect on the mediator. The second condition states that the mediator should have a significant effect on the dependent variable, and the third condition describes how the direct effect of the independent variable onto the dependent variable should be significant, when the mediation effect is not being considered, nor controlling for the direct effect of the mediator(s).
As a part of this set of conditions to qualify for mediation, the effect of an economic downturn (independent variable) on car sales on an industry level (mediator) was found to be significant(B = -11226.49, t(5879) = -40.20, p < .01). Because the effect is both significant and negative, Hypothesis 1 was supported.
Next, the results indicate that the total number of cars sold (mediator), has a significant effect on the sales of an individual car brand model (dependent variable), so that the other condition for mediation was also met, as it states that the effect of the mediator on the dependent variable has to be significant (B = 0.00, t(5858) = 4.59, p < .01). Moreover, these findings support Hypothesis 2, as the aggregate number of car units sold is indeed positively associated with a single brand's sales.
The final condition states that the effect of the independent variable on the dependent variable – inconsiderate of the mediators and their respective effects – has to be significant.
The results show that, the effect of economic downturns on car brand model sales is significantly mediated by the industry level car sales (B = -48.52, CI -73.17 to -24.83), thus supporting Hypothesis 3.
2nd mediation effect – indirect effect of economic downturns on car sales via aggregate car
sales. The sequence of actions to the second mediation effect were performed in a similar way as
with the first mediation effect, so that the influence of an economic downturn (independent
variable) on the number of competitive launches (mediator), was found to be significant (B = 2.93, t(5879) = 15.44, p < .01), but positive instead of negative, thereby qualifying as a condition to mediation, but failing to support Hypothesis 4.
The direct effect of the number of competitive launches (mediator) on an individual car brand's sales (dependent variable) is significant as well (B = -34.01, t(5858) = -21.72, p < .01), and also has the expected direction of effect (i.e., negative), so that Hypothesis 5 was supported.
As mentioned with the first mediation effect, taking note of the consequences of an economic downturn on the number of units sold for an individual car brand model without controlling for the explicit effects of both mediators, indeed shows a significant direct effect (B = -226.42, t(5863) = -10.50, p < .01). As a result, it can be concluded that the mediated effect of industry level car sales on car brand sales given an economic downturn, is found to be significant (B = -99.98, CI -116.06 to -84.96), offering statistical support to Hypothesis 6.
In this case, as well as with the first mediator effect, the findings are the result of 5000 bootstrap resamples (Preacher, Hayes, 2008), offering confidence intervals for the mediated effect. When the confidence interval does not include zero (i.e., the stated estimates to the upper and lower bootstrap confidence interval, are either both positive or both negative), the mediated effect is said to be statistically significant.
FIGURE 2
Schematic representation of the mediation effects
*** p < .01
-+
+
-126.6070*** (-226.4163)*** 2.9393*** -34.0137*** -11226.490*** .0043***Control variables & interaction effects
Number of products sold on industry level (M2)
-Sales of own product (Y) Economic downturn (X)
Competitive new product launches on industry level
Moderation Effects
To test to what extent multicollinearity was an issue for this study, variance inflation factors (VIF) were tested, and the found VIF values that range from 1.2 to 2.8 for this study, are below the cut-off values (Hair, 2010). This indicates that multicollinearity is not expected to be a truly problematic issue (Neter, Kutner, Wasserman, Nachtsheim, 1996). Next, the bivariate correlations among variables do not present themselves to be a big issue either, as the correlations are generally moderate (see table 2).
TABLE 2
Correlations between the different variables
As a proposed way of testing moderation effects (Frazier et al., 2004)variables were hierarchically entered into the model (Aiken, West, 1991; Cohen, Cohen, West, Aiken, 2003),where the variables denoting the interaction terms were entered after the control variables and independent variables respectively(e.g., Dunlap, Kemery, 1987; Holmbeck, 1997).
As is shown in table 3, model 1 consisted of only the control variables that reflected the type(s) of car associated with a car brand model, and the number of years a car brand model has been on the market (or if applicable, when it had its most recent update) (adjusted R2 = 0.19, p < .
01). For the second model, the independent variables were included (adjusted R2= .45, p < .01), and
in model 3 the two-way interaction effects were introduced (adjusted R2 = .48, p < .01).
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 1. -2. .43 ** -3. .19 ** .14 ** -4. .36 ** .27 ** .51 ** -5. .03 * -.03 * .15 ** .17 ** -6. .20 ** .08 ** -.04 ** .02 .26 ** -7. -.07 ** -.25 ** -.26 ** -.18 ** -.11 ** -.14 ** -8. -.18 ** -.31 ** -.26 ** -.20 ** -.16 ** -.17 ** -.16 ** -9. -.13 ** -.19 ** -.18 ** -.14 ** -.10 ** -.11 ** -.10 ** -.08 ** -10. .22 ** .18 ** .15 ** .22 ** .09 ** .11 ** -.04 ** -.05 ** -.08 ** -11. -.30 ** -.27 ** -.28 ** -.21 ** -.05 ** .02 .15 ** .19 ** .12 ** -.08 ** -12. -.18 ** -.18 ** -.16 ** -.16 ** -.06 ** -.06 ** -.03 * .01 .46 ** -.32 ** .13 ** -13. .26 ** .31 ** .28 ** .39 ** .22 ** .19 ** -.08 ** -.06 ** -.14 ** .12 ** .12 ** -.15 ** -14. .54 ** .39 ** .19 ** .35 ** .10 ** .20 ** -.07 ** -.15 ** -.10 ** .15 ** -.03 * -.09 ** .69 ** -15. .07 ** .01 -.01 .02 .00 .00 .03 * -.01 -.02 .02 .06 ** -.04 ** .01 .01 -16. -.16 ** -.14 ** -.13 ** -.12 ** -.03 * .00 .04 ** .11 ** .08 ** -.07 ** .20 ** .07 ** .05 ** -.03 * -.46 **
-1. sales; 2. hatchback; 3. sedan; 4. stationwagon; 5. coupé; 6. cabriolet; 7. MPV; 8. SUV; 9. other types of cars; 10. number of own launches; 11. number of competitive launches; 12. years since launch or most recent update; 13. number of consecutive own launches; 14. brand equity; 15. total amount of car units sold; 16. economic downturns
Model 4, the final model, contains all effects including the three-way interaction effects (adjusted R2 = .49, p < .01).
TABLE 3
Significance of F-changes for the respective regression models
From the table below (table 4) can be gathered that a higher number of product launches associated with a car brand model is beneficial to the brand's eventual sales (B = 278.27, t(5858) = 4.50, p < .01), thereby supporting hypothesis 7.
TABLE 4
Parameter estimates of variables to the final regression model
sedan -139.17 -217.44 ** -278.91 *** -285.52 *** stationwagon 3205.01 *** 1717.25 *** 1625.91 *** 1615.70 *** coupé -1208.43 *** -723.00 *** -813.34 *** -808.16 *** cabriolet 2050.56 *** 1341.27 *** 1234.20 *** 1217.89 *** MPV -205.45 346.86 *** 253.66 ** 240.52 * SUV -1294.36 *** -60.23 -245.44 * -255.25 *
other types of cars -677.58 *** 20.20 -137.25 -146.38
years since launch or most recent update -109.74 *** -74.14 *** -82.05 *** -84.10 ***
number of competitive new car brand model launches -30.12 *** -33.46 *** -34.01 ***
total amount of car units sold (industry level) .00 *** .00 *** .00 ***
number of own launches 363.16 *** 319.87 *** 278.27 ***
number of consecutive own launches -189.06 *** -132.93 *** -128.63 ***
brand equity .05 *** .05 *** .05 ***
economic downturn -123.58 *** -125.59 *** -126.61 ***
economic downturn * own launches 32.22 25.52
economic downturn * consecutive own launches 25.21 *** 24.34 ***
ecnomic downturn * brand equity -.01 *** -.01 ***
competitive launches * own launches -9.97 *** -10.09 ***
competitive launches * consecutive own launches -4.42 *** -4.76 ***
own launches * consecutive own launches -10.82 -16.83 **
economic downturn * own launches * consecutive own launches 2.16
competitive launches * own launches * consecutive own launches 1.29 ***
* p < .1 ** p < .05 *** p < .001
included variables model adjusted R Square R Square Change F Change df1 df2 p-value F Change
control variables 1 .193 .194 176.853 8 5872 .000
independent variables 2 .454 .261 468.276 6 5866 .000
interaction effects 3 .484 .031 58.129 6 5860 .000
The interaction term of the number of competitive launches and the number of own launches did have a significant and negative effect on car sales (B = -10.09, t(5858) = -5.95, p < .01) which implies that Hypothesis 8 could be supported.
The three-way interaction effect between the numbers of competitive launches, own launches and consecutive launches under the flag of a car brand model was significant as well, in addition to having a positive parameter estimate, thus supporting Hypothesis 9 (B = 1.29,
t(5858) = 3.59, p < .01).
The results indeed indicate that the interaction term of economic downturns and a car brand model's own number of launches is positive, although it is not significant however (B = 25.52, t(5858) = .83, p > .05) meaning that Hypothesis 10 is not supported.
There was also insufficient evidence in favor of Hypothesis 11, since the three-way interaction effect between economic downturns, the number own launches and the number of consecutive launches was found to be non-significant (B = 2.16, t(5858) = .50, p > .05).
The final hypothesis, Hypothesis 12, was not supported. Although economic downturns and a car brand model with high brand equity represent a significant interaction effect on the sales of a car brand model, car brand models with high brand equity are in fact associated with a larger relative drop of sales during times of a stalling economy (B = -.01, t(5858) = -11.50, p < .01). This is contrary to the expected positive interaction effect, which would have suggested that high-equity brands could absorb some of the negative impact of economic contractions.
Robustness Checks
To assess the validity of the results, two alternative operationalizations of the state of the economy have been used to see to what extent the previous results could be replicated.
Consumer confidence shows very similar direct effects compared to the original model that used relative GDP fluctuations as an independent variable. As such all of the previously found effects were identical, except for the interaction term between competitive launches, own launches and consecutive launches, which consequently became non-significant. A model estimation including the US interest rates, principally shows similar significant effects with respect to the the previous findings as well. There was again only one exception, namely that the interaction term between the economic indicator (here: interest rates) and the number of own launches became significant whereas it was non-significant as part of the original model.
Discussion
Indirect effects concerning economic downturns. The prediction that the economic situation is of
negative influence to the number of cars sold on an aggregate level was met with sufficient statistical evidence. This result replicates the findings of previous studies that sales for durables tend to drop during economic downturns(Deleersnyder et al., 2004; Pearce, Michael, 2006).Causes for this effect are explained by fact that consumers simply can no longer afford durables(Katona, 1975)or the higher levels of perceived risk consumers experience when the economy takes a plunge (Kahh, Meyer, 1991; Shead, Hodgins, 2009).
This study next demonstrated how changes in the total amount of cars sold, (interpreted as a proxy for the primary category demand for cars), affect the number of cars sold on an individual car brand level. In other words, that portion of the variation in the sales of an individual car brand model that has been explained by fluctuations of the aggregate demand level. This study has thus managed to decompose this specific aspect of the direct effect of an economic downturn, which could be relevant since changes in primary demand and secondary demand, each have very different implications for effective marketing decisions (Arora, Allenby, Ginter, 1998).
Theoretically, the effect of an economic downturn/upturn on an individual brand could be fully mediated by the aggregated demand, but that would assume that the industry is perfectly stable and no significant changes in market share would take place. Needless to say, the competitive nature of a matured industry such as the automotive industry is highly unlikely to meet such
assumptions. Future research could further address more of the dynamics in an industry with respect to secondary category demand effects, and also what role economic fluctuations would play in that sense.
The appropriateness of decomposing the effect of the state of the economy is also
Despite the fact that the effect of GDP fluctuations on the total number of competitors' new product launches was highly significant, its effect was a different one than the one hypothesized. Although a large number of studies found pro-cyclical patterns for both investments in R&D(e.g., Cincera et al., 2012)and product launches (e.g., Shleiffer, 1986; Francois, Lloyd-Ellis, 2003),the outcomes of this particular research suggests that over the past 30 years, car manufacturers have generally
demonstrated counter-cyclical innovative behavior, as was proposed by Devinney(Devinney, 1990)
for example.
This study also supported the idea of a negative effect on the sales of an individual car brand model, caused by a greater number of new product launches by competitors (Greenhalgh, Rogers, 2012).This is especially interesting because the mediated effect of the number of competitive new product launches on individual car brand model sales, seems to aggravate the already negative effect of a slow economy. This latter finding might not have surfaced without decomposing the effect of an economic crisis, and also provides an answer to the question of what specific forces affect the performance of a single brand of a durable product. As such, the effect of the number of competitive new product launches has been a disguised one, as it easily could have been attributed to a direct consequence of macro-economic fluctuations.
Beneficial aspects of innovative behavior for focal brand. The finding that new product launches exert a positive influence over a car brand model car sales, might be explained by a new product's ability to satisfy consumers' needs in new ways, that former products cannot (Pearson, 1993; Gatignon, Xuereb, 1997).The notion that sales tend to increase in response to a firm's
FIGURE 3
Effect on sales as years pass by for established products, measured since latest update
(Please note that for figure 3 and figure 4, the interval to the number of years since the latest update, has been somewhat arbitrarily set to 0 – 8 years. This has been done for graphical purposes and due to the distorting effect of outliers beyond 8 years.)
A somewhat similar picture is shown for car brand models that were previously unknown, albeit with some differences; for car brand models that are new to the market and make their first appearance on the Dutch car market, the peak in sales usually follows one year after their
FIGURE 4
Effect on sales as years pass by for new-to-the-market car brand models, measured since launch
Interplay of own product launches with competitors' innovations. One can imagine that the
benefits associated with innovations for one brand model will apply to another as well, so that innovation serves as a double edged sword; newly introduced car brand models benefit the brand itself by boosting its own sales and accomplish this (likely) by hurting the market share of existing, competitive, products. Something which has been suggested by previous observations of other scholars already(e.g., Castellacci, 2011; Greenhalgh, Rogers, 2012).
The way this study was set up, the number of own launches principally follows the rationale of an extended dummy variable: the number of a brand's own launches indicates whether a car brand model has or has not been responsible for one or more product introductions in a particular year. In figure 5 the two slopes for a (relatively) low and high number of own launches, demonstrate that as the number of competitive launches in a year increases, the products that are the most
But as the intensity of competing car brand launches increases however, sales are expected to drop at a (significantly) faster rate than established car brand models, that did not recently introduce any revisions or updates to their former versions.
FIGURE 5
Interaction effect between a brand's own new product launches and competitive launches
This would suggest that in a way, new products form a segment on their own (Schumpeter, 1992), and that newer products compete with one another in a more direct way(Greenhalgh, Rogers, 2012; Castellacci, 2011).
The dataset that has been used, has considered new product launches to be those products that were introduced to the market within the time span of a year. When new product launches are more loosely defined as products that were 'recently introduced', and thus taking into consideration the amount of time that has passed since a product was first introduced (or alternatively, most recently updated), the effect is even more dramatic (see figure 6). Brands that have recently been updated – or launched, if the brand has no history of updates yet – are truly very negatively affected by a high level of competitive new product launches.
FIGURE 6
Interaction effect intensity of competitive launches and time a product has been on the market
The positive thee-way interaction effect of competitive product launches relative to a brand's own product launches and previous consecutive launches suggest that car brand models seem to benefit from a larger number of consecutive product launches, but only as part of the stated interaction effect; interestingly, the consecutive number of own launches, has a negative effect on sales controlling for all the other effects (B = -128.63, t(5858) = – 8.99, p < .01). The graph below (figure 7) depicts the combinations of (high versus low) conditions for a brand's new product launches in a given year and its previous launches, indicating how often a brand has been revised or updated. Whereas high concentrations of competitive new product launches are generally
detrimental to a car brand's own sales, really new products that have been released a few years back, and have not been updated since, actually appear to benefit from high concentrations of competitive new product launches.
FIGURE 7
It might be that this specific type of products strike a balance between newness and familiarity with a greater audience. This would give these products an advantage over the older products that are considered to be somewhat outdated (Burns, 2007; Seetharaman, Chintagunta, 1998), but also over even newer products that have not managed to be part of people's consideration set around the time of purchase due to consumers' relative unfamiliarity with them (Aaker, 1997). A notion that is supported by the idea that new products must be known first before they can be bought(Rogers, 1995), and that previously unknown car brand models are associated with higher risks(Rogers, 1995; Sheth, 1981).
