|
Article Excerpt
Even in mature markets, managers are expected to improve their brands' performance year after year. When successful, they can expect to continue executing on an established marketing strategy. However, when the results are disappointing, a change or turnaround strategy may be called for to help performance get back on track. In such cases, performance diagnostics are needed to identify turnarounds and to quantify the role of marketing policy shifts in this process. This paper proposes a framework for such a diagnosis and applies several methods to provide converging evidence for two main findings. First, contrary to prevailing beliefs, the performance of brands in mature markets is not always stable. Instead, brands systematically improve or deteriorate their performance outlook in clearly identifiable time windows that are relatively short compared to windows of stability. Second, these shifts in performance regimes are associated with the brand's marketing actions and policy shifts, as opposed to competitive marketing. Promotion-oriented marketing policy shifts are particularly potent In improving a brand's performance outlook.
Key words: performance improvement; turnaround strategy; marketing mix; advertising; promotion
**********
A trend is a trend, is a trend, but the question is, will it bend? Will it alter its course, through some unforeseen force, and come to a premature end?
--Sir Alec Cairncross, Chief Economic advisor to the British government
1. Introduction
Year after year, marketing managers strive to improve the sales and profit performance of their brands. When products or markets are young, most of that sales growth comes from market expansion, which can produce positive sales trends for many years and for several competitors. As an example, all Japanese automobile brands gained sales and share in their emerging North American and European export markets in the 1970s and 1980s (Hanssens and Johansson 1991). However, in mature markets there are limits to expansion, e.g., consumer awareness and distribution may have reached a maximum, prices are in steady state, and competitive reaction to any new marketing initiative is fierce. Such mature product categories are typically viewed as equilibrium markets (Ehrenberg 1988). It is not surprising that in such markets, observed changes in market share are only temporary, and over the long run, market-share positions do not change (Dekimpe and Hanssens 1995, Nijs et al. 2001, Pauwels et al. 2002).
However, the mere fact that product markets have matured does not relieve managers of the pressure to grow their brands' performance. In particular, declining brand performance is regarded as an immediate reason for marketing intervention and even top management shake-up (Miller 1991). Moreover, management's fundamental "quest for more" (Hunt 2000) drives marketing investments, which, if effective, can create an upward trend in brand performance. On the other hand, demand saturation and competitive reaction pose limits to such performance growth (Bass et al. 1984). As a result, brand performance is subject to two opposing influences: mean reversion and change. Neither can last for a long time in mature markets: Prolonged periods of either flat or declining performance are incongruent with managerial objectives, and prolonged periods of growth are incongruent with market realities. Therefore, we may expect sales (1) performance in mature categories to go through successive regimes or windows of performance decline, stability, and growth.
Among these regimes, performance decline receives the most managerial and public attention because of its negative implications for investors and employees. Reversing a decline is considered more difficult than maintaining stability and often requires the use of a "turnaround strategy." For example, everyday low pricing may be gradually replaced by a strategy of high-low pricing, few advertising campaigns by many campaigns, and low levels of point-of-purchase activity by high levels of feature and display.
The empirical investigation of marketing turnaround strategies and their effects is mostly anecdotal in nature. For example, Advertising Age reported on the sales decline of the Budget Gourmet brand of ready-made food and attributed the turnaround to a highly effective advertising campaign (Bender 2001). However, we have no scientific evidence that the brand's performance improvement was actually due to the advertising campaign versus the pricing strategy change and increased point-of-purchase activity that occurred over this period. To the best of our knowledge, the only formal research on the impact of marketing policy changes on performance was conducted on Procter & Gamble's shift from promotion-intensive to advertising-intensive marketing support (Ailawadi et al. 2001). That research focused on a single identifiable regime shift in the data and provided no formal metrics for diagnosing gradual performance turnarounds over time.
In order to diagnose turnaround strategies, we need to first identify periods of poor performance in a brand's history. In particular, we must identify the beginning and the end of the decline. Secondly, we must isolate the causes associated with the turnaround. Such causes could be economic down- and upturns that affect the entire category, a single marketing action, or a sustained marketing policy change initiated by the brand, or competitive marketing activity.
Current market-response research does not yet offer a framework to either identify performance regime changes or to isolate their causes. Instead, recent papers have classified performance and marketing spending as evolving or stationary over the full data period (Dekimpe and Hanssens 1999). By far the most common scenario is business as usual, representing stationary performance and marketing in mature markets (Nijs et al. 2001, Pauwels et al. 2002). For the purpose of such classification, researchers study the full data period available, and perform their tests after allowing for seasonality and a deterministic trend (e.g., Srinivasan et al. 2004). Important changes in this full period, such as brand entry or channel addition, may be identified as structural breaks (e.g., Deleersnyder et al. 2002, Pauwels and Srinivasan 2004), with the market considered in equilibrium for the long periods in-between the breaks.
However, even in the absence of identifiable structural breaks, markets may not be stable at all times. Full-sample analysis may mask more subtle performance changes over time, i.e., smaller time windows in which performance is stable, improving, or declining. In other words, what appears to be a long period of stability in market performance to the researcher may in fact be a succession of time windows in which different players face different circumstances of growing, stable, and declining performance. Thus, the first objective of our paper is to propose a method for identifying performance regimes over time, along with transition points between them.
As argued earlier, some of these performance regimes (e.g., decline) are inconsistent with managerial objectives. In such cases, managers may go beyond single marketing actions and make course direction changes to the marketing mix to reverse an unfavorable path for the brand (Schendel et al. 1976). Therefore, the second objective of our paper is to relate changes in performance regimes to changes in marketing actions and marketing regimes. In so doing, we expand the scope of marketing-mix modeling: Whereas previous models were designed to measure the effects of single actions (such as a price change or an ad campaign) on current and sometimes future sales performance, we also analyze the regime-shifting effects of strategic change in marketing, such as a policy shift from low to high promotional intensity.
These two research objectives motivate us to (1) identify performance regimes and their transitions, and (2) investigate whether marketing actions may lead to improved performance regimes. We begin by classifying brand performance from a strategic perspective, and we formulate hypotheses on how different performance regimes are created over time, how these impact marketing decisions, and how these decisions, in turn, change business performance. Next we discuss three alternative methods to diagnose performance regimes and analyze marketing's power to affect them. We describe an extensive marketing database in the frozen-food category and use the three methods to provide converging evidence for our hypotheses. We conclude by highlighting managerial insights and avenues for future research.
2. Framework and Hypotheses
Reacting to a second-quarter operating loss of $1 billion, DaimlerChrysler's CEO stated "Admittedly, we have a setback in the third year [after implementing the Chrysler turnaround plan] but if you look at the trend we are moving in the right direction" (Financial Times 2003, p. 15). The quote illustrates how managers interpret their companies' performance in terms of trends and trend changes. Formally, brand performance regimes can be classified by their managerial desirability, based on two dimensions: the performance trend sign (up, insignificant, or down), and the change in this trend (accelerating or decelerating). (2) Table 1 combines these dimensions in six performance regimes, with accelerating growth (#1) and deteriorating decline (#6) the best-case and worst-case scenario, respectively.
As argued earlier, we do not expect brand performance to stay in any of these regimes for long time periods. In mature markets, sustained trends over long periods are unrealistic because they imply predetermined patterns that are independent of managerial and competitive marketing interventions (Lambkin and Day 1989). Second, at least the deteriorating decline scenario (#6) is unacceptable to managers. Their marketing actions aimed at performance improvement have the potential to turn the negative trend around (scenario #5), leading up to stable or even growing performance (Salmon 1988). Likewise, accelerating growth performance (#1) is unlikely to resist the gravitational forces of competitive reaction (Bass et al. 1984) and consumer habit formation (Ehrenberg 1988) for a long time. Therefore, we expect brand performance series to go through successive regimes of trend signs and trend changes. The question now becomes how often each regime occurs and whether marketing can affect regime shifts.
In the strategic change literature, punctuated equilibrium is the dominant paradigm for explaining regime shifts (Mullins et al. 1995). This paradigm holds that most successful organizations evolve through long periods of relative stability that are punctuated by occasional periods of upheaval. Punctuated-equilibrium theory argues that these revolutionary change or transition periods are typically short compared to the equilibrium periods.
We propose that this punctuated-equilibrium principle holds for market performance and marketing policy as well, for two reasons. First, buying behavior typically follows a stable pattern that is adequately captured by a zero-order stochastic process (Bass et al. 1984, Ehrenberg 1988). In mature categories, only strong customer motivation to revisit habitual buying patterns will change buying behavior and thus brand performance. Examples of such strong motivations are reactions to dramatic price reductions or creative product extensions (Simon 1997). However, such growth periods are not likely to last for extended time periods, because of consumer saturation and competitive reaction. On the flip side, periods of deteriorating decline will be especially short-lived because managers are pressed to take action to get out of such a clearly unfavorable regime. Because of this managerial action, we should observe periods of decline less often than periods of growth, which do not raise such strong concerns.
An example of subtly changing performance regimes in the frozen food market is the Budget Gourmet brand in the early nineties (Bender 2000), as shown in Figure 1. In the summer of 1992, management argued that Budget Gourmet's sales had been deteriorating over the past year, and the survival of the brand became uncertain. At that point, a new division president dramatically changed marketing policy, particularly in pricing (30% reduction over a prolonged time period), point-of-purchase activity (a major increase in feature and display), and advertising (a new campaign). After a few months, management saw strong performance improvement, which lasted for several more months, and the marketing campaign won the Advertising Age "Star" award for turning the brand around. Interestingly, neither the performance turnaround nor advertising's role in it are obvious from a visual inspection of the data; they require further analysis. Therefore, we propose:
[FIGURE 1 OMITTED]
HYPOTHESIS 1 (H1). Regimes of trending performance are shorter than periods of stable performance.
HYPOTHESIS 2 (H2). Regimes of decline in brand performance are less common than regimes of growth, which are in turn less common than regimes of stable brand performance.
While managerially relevant, Table 1's diagnostics about performance regimes are not sufficient for marketing decision makers. They also need to know how their actions may yield more favorable performance regimes. Unfortunately, previous literature offers limited guidance on this issue. Only two marketing concepts, the product evolutionary cycle and hysteresis, provide theory and empirical evidence on the triggers of performance regime transitions. First, the product evolutionary cycle (PEC) proposes explicit links between market growth and marketing influences (Lambkin and Day 1989, Tellis and Crawford 1981). Empirical studies include the impact of advertising spending on cigarette markets (Holak and Tang 1990) and new products' struggles with incumbent products for retail space and market share (Bronnenberg et al. 2000, Uhlrich et al. 2001). As such, this research stream allows for more flexible market growth patterns than the traditional product life cycle. However, the above studies are focused on emerging, as opposed to mature, markets and they have...
|
