A Multi-dimensional Clustering on Fuzzy Metrics to Classify CPG Pricing and Price Promotion Strategies: The Case of Pasta in Italy

Luigi Palumbo, Tiziana Laureti and Ilaria Benedetti

Introduction

Price promotions, which are temporary price changes offered by a seller, play a crucial role in retail, especially in consumer packaged goods (CPG) (Gomez, Rao Sc McLaughlin, 2007; Besanko, Dube Sc Gupta, 2005; Gedenk, Neslin Sc Ailawadi, 2010). Their importance for CPG manufacturers is indicated by the share of total revenue allocated for trade promotions,1 which varies between 10% and 25% (Boston Consulting Group, 2012; Gartner, 2015) and by the long-term impact of price promotions on consumer habits (Mela, Gupta Sc Lehmann, 1997).

Discounted products are generally in the minority on store shelves compared to those at regular prices (McShane, Chen, Anderson Sc Simester, 2016), but they comprise a proportionally larger share of sales. According to different sources and definitions, the share of sales for goods on promotion over total grocery sales varies across countries. For example, in the UK the share of products on promotion accounted for 32% in 2009 (Nielsen, 2009a) and 51.5% in 2015 (IRI, 2016), whereas in the US it accounted for 42.8% in 2009 (Nielsen, 2009b) and 34.9% in 2015 (IRI, 2016). The share of turnover for products on price promotion in Australia was 40.6% in 2015 (IRI, 2016) and 40% in 2017 (Nielsen, 2018a), whereas in New Zealand, it accounted for 55.8% in 2015 (IRI, 2016) and almost 60% in 2017 (Nielsen, 2018b). Finally, in Italy, the share of sales for goods on price promotion accounted for 27.9% in 2015 (IRI, 2016).2

Empirical evidence regarding the profitability of price promotion is controversial. Several studies reveal that promotions not only influence retailers’ revenue (Jedidi, Mela Sc Gupta, 1999; Dreze Sc Bell, 2003) but also have a positive impact on key intangible assets, such as brand equity (Buil, de Chernatony Sc Martinez, 2012) and brand loyalty (Gedenk Sc Neslin, 1999). However, other studies demonstrate that price promotions tend to have little long-term effects (Srinivasan, Pauwels, Hanssens Sc Dikempe, 2004). A recent study by Nielsen shows that 67% of trade promotions in the US in

2014 had a negative return for the manufacturer (Nielsen, 2015a). In addition, the positive impact on sales sometimes comes with negative effects for brands (Zoellner & Schaefers, 2015; DelVecchio, Henard & Freling, 2006).

Yet, retailers use various types of price promotion to increase sales and undertake complex managerial decision-making processes to define them (Bogomolova, Szabo & Kennedy, 2017). To shed light on the pricing strategies most widely adopted in Italy, we use fuzzy metrics, which allow us to better identify and understand the different commercial strategies used by retailers and brands. Given its importance in Italy and its complex competitive landscape, we focus on pasta (Cacchiarelli & Sorrentino, 2019).

Retailers are likely to have different pricing strategies for different category-brand combinations. By focusing on a single product category, i.e. pasta (which in Italy is a so-called destination category), we can capture the differences in pricing and price promotion strategies across retailers and geographic areas.

Therefore, this chapter responds to the following research questions:

  • 1. What strategies are employed for pricing and price promotion in the Italian pasta retail market?
  • 2. How are the strategies for pricing and price promotion distributed across regions, retailers and brands?

Pasta is an ideal sector for analysis to arrive at answers to those questions, which are relevant for marketing practitioners in CPG manufacturing and retail. Indeed, Italy is the world’s leading pasta producer and the largest consumer of pasta (International Pasta Organisation, 2014).

Therefore, brand managers need to review the overall positioning of their products to consumers, spotting potential inconsistencies that may create confusion or strategies that are not consistent with their overall plan (Bolton, Shankar & Montoya, 2010). Retailers need to control and evaluate their pricing and promotional strategies vis-a-vis those of their competitors to ensure that they transmit the right message to consumers about which store to visit when they are shopping for this product.

The geographic aspect is particularly relevant, because the local competitive environment is considered a key determinant of the results of retailers’ strategy, and some evidence indicates that retailer strategies tend to be homogeneous in a given area (Ellickson Sc Misra, 2008).

In this chapter, we employ a unique dataset of prices and temporary price reductions3 for retailers across Italy and consider the type-brand-retailer- region combination as a statistical observation. Our empirical analysis is based on 738 items of weekly average price data from January 2011 to June

2015 segmented across 16 retailers, 18 brands and five regions.

In order to explore the various pricing and price promotional strategies used for pasta, we apply a fuzzy-set theory approach (Zadeh, 1965,1968,1977;

Dubois 8c Prade, 1980; Zimmermann, 2001), as properly designed membership functions (MF) enable us to achieve a better classification of the data, smoothing distortions caused by outliers while still including them in the analysis. Another advantage of the fuzzy-set theory approach is that it overcomes the limits from discrete classifications of data, preserving a higher degree of information for analysis.

We propose the use of multi-dimensional clustering that takes into account regular prices and temporary price reductions to derive the strategies implemented by retailers for different brands and classify them according to the characteristics revealed.

The remainder of this chapter is structured as follows. First, we review studies related to price promotion to determine the relevant dimensions to consider in our analysis. Then, we present the research methodology and describe the data used. Results are presented in the next section. Finally, we offer the managerial implications, limitations, suggestions for future research and conclusions.

Conceptual Framework

Price promotion is seen as the only tool that can drive an increase in sales in the short term, unlike advertising spending, which has long-term returns.4 It is also perceived as less risky for small brands because the return on promotional spending is more certain in terms of demand generation (Anderson 8c Fox, 2019; Bogomolova, Szabo 8c Kennedy, 2017). Price promotions are another tool for manufacturers to use in maintaining brand awareness, as they nearly always accompany communication efforts by retailers (Anderson 8c Fox, 2019; Brito 8c Hammond, 2007; Blattberg 8c Briesch, 2012). Manufacturers of private label products for retailers also devote funding for promotional activities to support their sales, perhaps because product volume declines dramatically when products are not promoted (Anderson 8c Fox, 2019).

As mentioned earlier, the effects on retailer revenue are generally positive, especially retailers in CPG, as a majority of them historically rely heavily on price promotion as a competitive weapon, i.e. using a strategy known as ‘HiLo’ (‘high price, low price’) by retailers who offer periodic, temporary price discounts (low prices) in contrast to the regular prices (high prices). Other retailers, including Walmart, effectively built their success on an EveryDay-Low-Price (EDLP) strategy, used by retailers that maintain a low constant price without any discounts.

Some evidence shows that HiLo strategies lead to a higher profit margin, and moving from a HiLo to EDLP strategy entails high switching costs (Ellickson, Misra 8c Nair, 2012). Retailers’ economic results are also highly influenced by the local context, as competitive activity and socio-economic background may effectively alter the payoff from a HiLo or EDLP strategy. Stores in the same chain in different locations might implement different strategies to better compete in the local market (Ellickson 8c Misra, 2008).

Moreover, promotions may be strongly habit-forming for consumers. For instance, when JC Penney tried to switch from a ‘coupon sales’ pricing strategy to an EDLP strategy, it experienced a dramatic decline in store visits and sales (Mourdoukoutas, 2017), and eventually it was forced to return to a HiLo strategy. When even prominent manufacturers such as Procter & Gamble (P&G) reduced spending on price promotions to move to an EDLP strategy, it faced severe backlash, as consumers perceived that P&G brands suddenly became more expensive (Slater, 2001). The EDLP strategy carried out in the 1990s cost P&G a sizeable share of the market as well, because competing brands maintained a HiLo strategy, effectively eroding P&G’s position in several categories. Eventually, P&G returned to a HiLo strategy, even though some studies show that the EDLP strategy improved its profits as a percentage of sales. It seems likely that the drop in sales volume was of greater concern and had to be addressed (Ailawadi, Lehmann & Neslin, 2001).

The attractiveness of a price promotion is generally related to the magnitude of the discount versus the regular price. However, a particularly deep discount is a concern for brand managers. A retailer might offer an extremely low price to increase store traffic, but this move could also ‘disrupt the marketplace, anger competing retailers, and damage brand equity' (Anderson & Fox, 2019), with a detrimental long-term effect on overall brand performance (DelVecchio, Henard & Freling, 2006). According to Nielsen (2015b), price promotion is a driver that 45% of consumers report as important in store switching. It is widely documented that shoppers generally buy a basket of goods that includes both items on sale and those at the regular price (Anderson & Fox, 2019). Therefore, to maximise shopper traffic to a store, the retailer can use price promotion as a magnet to increase overall sales. The frequency of price promotion, i.e. the average number of times a specific brand is promoted over a specific period, is another key decision for brand managers and retailers (Allender & Richards, 2012).

In the retail context, a ‘product category’ is defined as ‘a distinct manageable group of products that consumers perceive to be related and/or substitutable in meeting a consumer need' (Blattberg & Fox, 1995). Specific product categories that are particularly well-suited for attracting consumers to a store are called ‘destination categories’. Within those categories, the items that are most important for attracting shoppers are called ‘key value items’ (Briesch, Dillon & Fox, 2013). The considerations by retail managers in pricing and promoting those categories and items go beyond simple margin targets, as the main goal is to leverage those categories for attracting shoppers in the store and generate beneficial spillovers in other categories (the basket building effect) (Chevalier, Kashyap & Rossi, 2003).

Therefore, pricing and price promotion strategies are key elements in an overall sale strategy for CPG manufacturers and retailers. The research on this subject is vast and consolidated, as ascertained in several literature reviews over rime (Kuntner & Teichert, 2016; Blattberg & Neslin, 1989; Blattberg, Briesch & Fox, 1995; Neslin, 2002; Ailawadi, Beauchamp, Donthu, Gauri & Shankar, 2009; Grewal er al., 2010; Fassnacht & El Husseini, 2013).

Bolton and Shankar (2003) derived an empirical taxonomy of pricing and price promotion strategies for grocery retailers in the US across several product categories using a multi-dimensional approach. Their study is a departure from the traditional classification of strategies (Zwanka, 2017) focused on FliLo and EDLP, either as a binary classification (Johnson, 2003; Bailey, 2008; Lai & Rao, 1997) or as a continuum with several intermediate strategies between those two extremes (Hoch, Dreze & Purk, 1994; Gauri, Trivedi & Grewal, 2008; Ellickson 8c Misra, 2008).

By referring to the Previous literature on price promotions (Bolton & Shankar, 2003; Hoch, Dreze & Purk, 1994; Blattberg & Briesch, 2012) we identify the following strategic dimensions to be considered for analysing retailers’ price strategies for pasta in Italy: regular price level, regular price variability, price promotion level, price promotion variability and price promotion frequency. In the next section, the metrics used to capture these dimensions are reported.

Data and Methods

The dataset we used in this study has been obtained from ISMEA data, after having carried out analyses to delete inconsistencies and impute missing data.5 The dataset includes 738 observations on regular prices (RP) and temporary price reductions (TPR), divided into two different tables, for pasta stock keeping units (SKUs) at different retail chains divided into NUTS1 regions in Italy. In this case, the 738 observations combine the type, brand, retailer and region.6 Data are reported for 234 weeks, starting in January 2011, and provide information on RP and TPR for 18 pasta brands across 16 retailers in 5 NUTS1 regions. The data points total 153.182 in the RP table and 16.910 in the TPR table. RPs and TPRs are list prices for the retailers’ chain in each region where they have stores.

The dataset is particularly valuable, as it enables us to follow the brand positioning across multiple retailers and regions. The addition of the pasta format to define the observations increases the granularity of our analysis and enables us to provide more insights for brands with more SKUs on retail shelves. Pasta, as a destination category with a large number of manufacturers of very different sizes, provides an ideal landscape for exploring different commercial strategies.

We propose multi-dimensional clustering that takes into account regular prices and temporary price reductions to derive the strategies implemented by retailers on the different brands and classify them according to the characteristics revealed. We used the partitioning around medoids (PAM) clustering algorithm, as it provides a higher level of robustness for noise and outliers than other methods, such as к-means. Furthermore, as the medoids identified in the PAM procedure are in our dataset, we can also use them as reference points for our MF.

As a first step in our procedure for analysing the dataset and searching for answers to our research questions, we calculate the following set of synthetic metrics (Table 20.1) to capture the dimensions identified above:

  • 1. Average RP, calculated as the arithmetic mean of RP time series;
  • 2. RP variability, calculated as the relative standard deviation (RSD) of RP time series;
  • 3. TPR percentage, calculated as the difference between RP and TPR values divided by RP, and multiplied by 100;
  • 4. TPR variability, calculated as the RSD of the TPR percentage time series;
  • 5. TPR frequency, calculated as the number of TPR data points over the number of RP data points.

A graphical analysis of the five metrics shows the presence of outliers and skewed distributions that could potentially affect the analysis of the data, as shown in Figure 20.1 for RP variability and TPR frequency.

We use the PAM algorithm to cluster the observations into two sets, representing high and low clusters, based on each metric listed above. This operation divides the dataset into clusters characterised by the low (or high) value of the metric being evaluated, as a basis for the following empirical definition of pricing and price promotion strategies. For instance, we identify the observations characterised by a high RP, compared to the rest of our sample. This transformation, which classifies every observation as high or low, effectively removes potential bias caused by outliers in the distribution of each metric.

To overcome the limitations in the binary classification performed in the previous step, we use the fuzzy-set theory approach to smooth the transition from low to high classification for each metric, defining an MF that will associate each observation with a degree of belonging to the low classification m( and consequently a degree of belonging to the high classification 1 -mr

Several types of MF for fuzzification of a dataset have been introduced in the literature. Many of the simplest forms - for instance, triangular,

Table 20.1 Pricing and price promotion metrics

Dimension

Metric

Min.

Median

Max.

RSD

Regular price level

Average RP

0.460

1.780

3.392

28.04%

Regular price variability

RP variability

0.000

3.460

18.074

71.87%

Price promotion level

Average TPR%

0.000

18.489

55.556

80.53%

Price promotion variability

TPR% variability

0.000

0.197

1.124

100.25%

Price promotion frequency

TPR frequency

0.000

5.556

98.291

120.21%

Dispersion of pricing and price promotion metrics

Figure 20.1 Dispersion of pricing and price promotion metrics.

trapezoidal or Gaussian - rely on arbitrary values for definition of the MF, which makes them prone to subjective evaluations. Other more sophisticated forms - for example, those based on a Lorentz distribution and its derivatives - offer a more objective approach to fuzzification.

In this case, we need the MF to do the following:

  • • avoid excessive change in membership levels at the edges of the distribution, as some metrics have outliers and hard limits at zero;7
  • • avoid reliance on any external parameters for their definition.

Keeping those points in mind, we define the MF in a series of stages based only on information derived from the dataset. The medoids found in the previous PAM procedure are used as partitioning points in the fuzzification. For each metric, data points below the low medoid have full membership in the low cluster (mL = 1) and those above the high medoid have full membership in the high cluster (mL = 0). For data points between them, we define the degree of membership in the low cluster for the data point i as the proportion of data points between the two medoids with a value higher than i.

The MF results are the basis for our PAM clustering aimed at discovering the different commercial strategies in the market. We select the optimal number of clusters for this operation based on average silhouette width.

For control purposes, we also perform the same clustering on the value of the original metrics, selecting the optimal number of clusters according to average silhouette width and forcing the same number of clusters that we found optimal for clustering on membership values.

We then characterise each strategy individuated in our procedure according to dimensions, region, retailers and brands.

Results

Table 20.2 lists the results from the initial clustering of values in high and low classifications according to each dimension. We show that some dimensions have a fairly skewed distribution of values.

Our MF attenuates this skewness to a certain extent, resulting in a relative majority of observations classified with a membership value blending high and low classifications, as shown in Table 20.3 and Figures 20.2-20.6.

Moving on to the results of our study, in a comparison of the different clustering procedures aimed at identifying pricing and price promotion strategies, two results are clear:

  • • clustering on membership values captures a deeper level of variability across the dataset, identifying nine clusters as optimal divisions, rather than the four clusters that were supposed to be optimal in clustering on the original metrics;
  • • clusters calculated on membership values are more consistent with the business logic behind pricing and price promotion strategies.

Table 20.2 Clustering results on each dimension

Metric

Low Medoid

Low N

High Medoid

High N

Silhouette

Average RP

1.515

393

2.385

345

0.709

RP Variability

2.875

527

6.982

211

0.607

Average TPR%

0.000

337

28.495

401

0.708

TPR% Variability

0.000

361

0.387

377

0.588

TPR Frequency

1.282

460

21.264

278

0.667

Note: The silhouette index is related to PAM clustering with two clusters for each metric. This is not necessarily the maximum silhouette index achievable for each metric, as a different number of clusters might have better performance. However, for the purpose of classifying pricing and price promotion strategies, this binary clustering enables us to link our analysis with business practices.

Table 20.3 Membership function results on each dimension

Metric

M, = 1

MJ0:1[

M, =0

Average RP

197

368

173

RP variability

264

368

106

Average TPR%

178

360

200

TPR% variability

248

302

188

TPR frequency

228

368

142

Membership function for average RP low

Figure 20.2 Membership function for average RP low.

Membership function for RP variability low

Figure 20.3 Membership function for RP variability low.

For example, to support this second point, we report the mapping of the different clusters against the average RP. The clustering derived on the membership values (Figure 20.7) is much more consistent with business logic than the one calculated with the original metrics. In contrast, the ‘traditional’ approach does not show a clear pattern consistent with the business logic when the optimal number of clusters is applied based on the original metric values (Figure 20.8) as well as when using the same number of clusters found to be optimal by clustering the fuzzy membership values (Figure 20.9).

Table 20.4 lists the results from cluster analysis based on fuzzy membership values measured with our observations, reporting mean scores of each original metric.

A Membership function for TPR percentage low

Figure 20A Membership function for TPR percentage low.

Membership function for TPR percentage variability low

Figure 20.5 Membership function for TPR percentage variability low.

Using fuzzy clustering, we can derive our empirical taxonomy of the different pricing and price promotion strategies based on the peculiarities of each cluster. We label them: everyday new low price (EDNLP), everyday low price (EDLP), surprise promo (SP), stable premium promo (SPP), luxury (L), high variability luxury (HVL), stable value promo (SVP), high variability value promo (HVVP) and high variability premium promo (HVPP).

Table 20.5 shows the combinations of pricing dimensions for each pricing strategy, listing the characteristics of those clusters based on the business domain and the size of each cluster.

This clustering leads to a fairly balanced panel, with a wide variety of strategies. This finding is consistent with the overall business scenario for

Membership function for TPR frequency low

Figure 20.6 Membership function for TPR frequency low.

Average RP and clustering on fuzzy membership values

Figure 20.7 Average RP and clustering on fuzzy membership values.

pasta market in Italy, characterised by fierce competition and a large number of competitors,. The amount of high TPR frequency and TPR% value items in the value (clusters 7 and 8) and premium segments (clusters 4 and 9) are roughly equivalent, and they are probably frequently and deeply promoted to attract shoppers (Bolton & Shankar, 2003). In contrast, the number of prices that change infrequently - with both low RP and low TPR frequency - is remarkably higher in the value segments (cluster 2) than in the premium segments (cluster 5). A large share of observations has extremely high variability, probably a replacement for TPR events, in both the value (cluster 1) and premium segments (cluster 6). Finally, a large cluster shows a

Average RP and clustering on original metrics

Figure 20.8 Average RP and clustering on original metrics.

Average RP and clustering on original metrics values, forcing the same number of clusters identified with the clustering on fuzzy membership values

Figure 20.9 Average RP and clustering on original metrics values, forcing the same number of clusters identified with the clustering on fuzzy membership values.

characteristic of the category, with average pricing, relatively infrequent TPR events and extremely high TPR% variability (cluster 3). This might mean that some TPR events will be much more attractive to customers than others, but the latter will be much less appealing in term of monetary savings.

Some underlying macro groups emerge from this clustering, with some strategies that rely heavily on TPR (clusters 4, 7, 8 and 9) and others that instead decline to pull this promotional lever (clusters 1, 2, 3, 5 and 6). Another characteristic, this time based on RP positioning, shows that some

Table 20.4 Mean scores on the dimensions for each cluster

N

Average

RP

RP

Variability

Average

TPR%

TPR%

Variability

TPR

frequency

Min.

738

0.460

0.000

0.000

0.000

0.000

Median

738

1.780

3.460

18.489

0.197

5.556

Max.

738

3.392

18.074

55.556

1.124

98.291

Cluster

1

52

1.225

6.502

5.661

0.013

0.509

2

104

1.569

1.437

3.507

0.010

0.793

3

81

1.874

2.803

6.702

0.520

6.691

4

73

2.378

3.119

32.353

0.147

13.229

5

51

2.540

2.123

2.048

0.001

0.107

6

59

2.325

6.374

5.214

0.027

0.636

7

102

1.530

2.349

29.979

0.222

17.647

8

98

1.316

5.474

22.283

0.405

18.675

9

118

2.330

7.311

26.780

0.435

20.815

Summary of pricing and price promotion strategies. Note

Figure 20.10 Summary of pricing and price promotion strategies. Note: Everyday new low price = EDNLP, everyday low price = EDLP, surprise promo = SP, stable premium promo = SPP, luxury = L, high variability luxury = HVL, stable value promo = SVP, high variability value promo = HVP and high variability premium promo = HVPP.

clusters use premium price positioning (clusters 4, 5, 6 and 9), while others use value positioning (clusters 1, 2, 7 and 8), and the remaining group positions itself around an average market price (cluster 3).

Having classified all observations based on our empirically derived taxonomy, we analyse how those classifications are distributed across regions, retailers and brands. We first examine group characteristics in the NUTS 1 regions, standardising each metric over the metric’s overall average for easier comparison (Table 20.6).

Table 20.5 Characterisation of pricing and price promotion strategies

Cluster

N

Label

Description

TPR

reliance

RP

positioning

1

52

Every day new low price

Low RP, extremely high RP variability, extremely low TPR%, frequency and variability

Low

Value

2

104

Every day low price

Low RP and RP variability, extremely low TPR%, TPR frequency and TPR variability

Low

Value

3

81

Surprise

promo

Average RP, extremely high TPR variability, relatively low RP Variability, TPR% and TPR ' frequency

Extremely high TPR%, high TPR frequency, low TPR and RP variability, high RP

Low

Average

4

73

Stable

premium

promo

High

Premium

5

51

Luxury

High RP, Low RP

variability, extremely low TPR%, frequency and variability

Low

Premium

6

59

High

variability

luxury

High RP and RP variability, extremely low TPR%, frequency and variability

Low

Premium

7

102

Stable value promo

Extremely High TPR% and frequency, low TPR variability and RP, medium RP variability

High

Value

8

98

High

variability value promo

High TPR%, frequency and variability, High RP variability, low RP

High

Value

9

118

High

variability

premium

promo

High TPR%, frequency and variability, high RP and RP variability

High

Premium

Average regular prices are quite similar across regions. More evident differences appear in comparing TPR frequencies, with extremely high use of this promotional tool in Southern Italy and much less so by all the other regions. The Islands, in particular, stand out as having much lower TPR frequency and TPR percentage than other regions.

To compare the composition of strategies across regions, we plotted the percentage of observations in each region classified according to our taxonomy (Figure 20.11).

Table 20.6 Summary metrics by region

Region

Average

RP

RP

Variability

Average

TPR%

TPR%

Variability

TPR

Frequency

N

Centre

1.024

0.662

0.992

0.730

0.826

226

Islands

0.980

1.032

0.569

1.334

0.417

64

North-east

1.032

1.194

1.066

0.738

0.760

180

North-west

1.022

1.175

1.063

1.094

0.898

113

South

0.921

1.127

1.066

1.491

1.848

155

Pricing and price promotion strategies by region

Figure 20.11 Pricing and price promotion strategies by region.

Southern Italy has a prevalence of strategies that rely heavily on TPR, but they are nearly absent in the Islands. Another notable characteristic is that the Islands present much less strategies that rely on value or premium pricing and much more strategies that position the RP near the market average.

Table 20.7 Summary metrics by retailer

Retailer

N

Average

RP

RP

Variability

Average

TPR%

TPR%

Variability

TPR

Frequency

Auchan

86

0.927

1.269

1.302

1.105

0.782

AZ

11

0.687

1.075

1.160

1.034

2.160

SISA

79

0.974

0.816

1.650

0.844

0.789

Code CRAI Ovest

8

1.091

0.623

0.067

0.000

0.030

Conad-Iper

178

1.047

1.251

1.588

1.540

2.150

Conad del Tirreno

89

0.986

0.680

0.335

1.743

0.460

Coop Centrale Adriatica

46

0.835

1.092

0.039

0.110

0.010

Etruria-SMA

7

1.065

1.104

1.558

0.678

1.175

Gabrielli

22

0.934

1.065

1.271

1.196

1.761

11 Gigante

12

0.892

1.174

1.415

1.795

1.221

Megamark

8

0.780

0.936

0.644

0.476

1.025

Multicedi

14

0.875

1.126

1.162

1.840

2.582

Nealco

86

1.153

0.417

0.125

0.000

0.047

SMA

55

1.083

1.279

0.423

0.237

0.088

Unicomm

15

1.031

1.285

1.451

1.023

1.369

Unicoop Tirreno

22

0.967

0.918

1.592

0.910

1.127

We then analyse our results by retailer chain, with the standardised metrics listed in Table 20.7. Note that the distribution of stores by retailer chain is not regionally uniform.15

Figure 20.12 shows the proportion of strategies employed by each retailer. The retailer profiles vary widely, from those that take a more balanced approach, such as Auchan, to those that clearly chose sides in terms of price positioning and the use of TPR, such as II Gigante and Multicedi.

Finally, we analyse the strategies by brand, with the standardised metrics listed in Table 20.8.

We see large differences in size for each group. Those that are smaller might be private labels, which are distributed by only a single retailer, or a local brand with limited distribution, often in a single region. Pricing strategies are strongly heterogeneous among brands.

Figure 20.13, which illustrates the proportion of pricing and promotional price strategies by brand, shows many interesting facts about individual brand positioning. Some brands take a clear EDLP approach (i.e. Bianconi, Ceccato, Coop and Valdigrano), positioning themselves with value pricing and refraining from TPR. But other brands rely heavily on TPR for their commercial strategy, positioning their pricing as value (II Gigante, Sisa) or using a mixed strategy (Multicedi). However, the brands mentioned so far all have smaller representation in our analysis. It is reasonable to expect that brands with a larger distribution adopt a wider variety of strategies. The remaining brands have different levels of reliance on TPR, and the brands

Pricing and price promotion strategies by retailer

Figure 20.12 Pricing and price promotion strategies by retailer.

more represented in our dataset (Barilla, De Cecco, Divella and Voiello) show a slight prevalence in the use of TPR in marketing strategies.

Regarding the pricing level, most brands are fairly consistent, pursuing strategies with common characteristics. However, a few brands employ strategies with contradictory characteristics. This can affect the clarity of the brand image among consumers, if the same brand is positioned very differently across regions or retailers in the same region.

Conclusions

In this chapter, we empirically identify nine different pricing and price promotion strategies for pasta brands in Italy, characterised by different levels and mixtures of RP and TPR positioning: everyday new low price, everyday low price, surprise promo, stable premium promo, luxury, high variability

Table 20.8 Summary metrics by brand

Brand

N

Average

RP

RP

Variability

Average

TPR%

TPR%

Variability

TPR

Frequency

Agnesi

11

0.987

0.523

0.693

0.647

0.502

Amato

10

0.801

0.709

0.494

0.608

0.214

Barilla

206

0.863

0.778

0.905

1.039

1.062

Bianconi

1

0.386

1.064

0.000

0.000

0.000

Ceccato

5

0.438

1.992

0.000

0.000

0.000

Colavita

4

0.412

1.887

0.438

1.047

0.122

Coop

20

0.625

0.717

0.018

0.000

0.004

De Cecco

218

1.290

1.066

1.189

1.048

1.263

Divella

82

0.699

0.932

1.414

1.136

1.442

G/Passione

3

1.748

0.688

0.254

1.241

0.434

Garofalo

27

1.297

0.787

1.375

0.780

1.244

11 Gigante

4

0.586

1.138

1.300

1.804

0.377

Jolly

26

0.929

0.288

1.014

0.180

0.199

Multicedi

2

1.164

1.791

1.132

2.369

0.835

Pallante Reggia

17

0.468

1.556

0.750

1.058

1.111

Sisa

9

0.653

0.808

1.349

1.394

0.334

Valdigrano

5

0.316

0.044

0.000

0.000

0.000

Voiello

88

1.155

1.709

0.738

1.285

0.593

luxury, stable value promo, high variability value promo and high variability premium promo.

We also show that fuzzification of the metrics that underlie the clustering enable us to identify clusters that are more consistent with the business logic in the pricing and price promotions strategies.

We developed a novel MF approach, based on preliminary clustering of the individual metrics and leveraging their medoids to derive membership values that: (1) are not influenced by arbitrary or exogenous values, (2) minimise the effect of outliers and (3) preserve the maximum information from the original metrics.

We use a dataset with weekly reported prices which has been filtered to remove inconsistencies and impute missing data. However, one limitation of our research is that our dataset does not contain information about display and features, two aspects that previous literature considers important in augmenting TPR results. In addition, the data refer to large regions, which might be characterised by a high level of heterogeneity in pricing strategies across sub-regions.

Further research using scanner data might be able to provide additional insights into retailers’ and brands’ pricing and price promotion strategies. Moreover, our clustering approach treats all variables as equally relevant for classifying the strategies. Clustering on weighted variables, giving certain dimensions of the marketing mix more relevance, might offer further insights and different perspectives on pricing and price promotion strategies.

Pricing and price promotion strategies by brand

Figure 20.13 Pricing and price promotion strategies by brand.

Notes

  • 1 Retail price promotions are generally backed by trade price promotions offered by CPG manufacturers to retailers. The degree of price reduction passed from retailers to consumers may vary widely across brands, categories (Besanko, Dube St Gupta, 2005) and time (Meza 8t Sudhir, 2006).
  • 2 Different sources use slightly different definitions in calculating these values.
  • 3 TPR is only one of the possible promotional strategies in CPG. However, several studies show how TPR is used more widely than other promotional options (multi-buy, extra products free, etc.) (Gedenk, Neslin 8t Ailawadi, 2010; Bogomolova (unify the citations in a single parentheses), Dunn, Trinh, Taylor St Volpe, 2015). Therefore, our research concentrates on TPRs.
  • 4 This bias towards short-term returns is deeply criticised by academics and practitioners who advocate the superiority of advertising in building customers loyalty, as stated by Ken Roman of Ogilvy St Mather: ‘Promotions rent customers, advertising owns them’ (Pettie St Pettie, 2003). Recent studies, however, show that brand loyalty is at an all-time low, despite increases in advertising spending (Jedidi, Mela 6c Gupta, 1999; Nielsen, 2019).
  • 5 We thank ISMEA, in particular, Maria Nucera for giving us the dataset. We performed several elaborations to correct reporting inconsistencies in the dataset, such as: removing data on price promotions that are inconsistent with regular prices; reclassifying regular price points which follow a price promotion pattern; removing anomalous regular price points; and imputing missing regular price points.

In our elaboration, we identified as TPRs - in addition to those self-reported by retailers - a price reduction of at least 10% (Nielsen, 2020) over the regular price that lasts for less than five weeks and is followed by an increase in the shelf price. In a series of studies on the German food retail market, a TPR is defined as a reduction of at least 5% over the regular price that lasts for a maximum of four weeks and is followed by an increase in the shelf price. The regular price is observed in the four weeks preceding a sales promotion (Empen, Loy Sc Weiss, 2015; Herrmann, Moeser Sc Weber, 2005). Combining this information with the reference from Nielsen (2020) on the Italian market and the metrics derived from our data, we feel confident about the consistency of our TPR definition metrics.

  • 6 The SKU is defined by a combination of format and brand.
  • 7 We assume that variability at the edges of the distribution has less impact on differentiating two strategies than variability in the middle of the distribution.
  • 8 As a result, in some cases a higher number of observations may be attributed to retailers by reporting the RP and TPR of more brands and SKUs and/or across multiple regions. Some retail chains in the analysis are in only a single region, resulting in smaller representation.

References

Ailawadi, K.L., Beauchamp, J., Donthu, N., Gauri, D.K., Sc Shankar, V. (2009). Communication and Promotion Decisions in Retailing: A Review and Directions for Future Research. Journal of Retailing, 85(1), 42-55.

Ailawadi, K.L., Lehmann, D.R., Sc Neslin, S.A. (2001). Market Response to a Major Policy Change in the Marketing Mix: Learning from Procter Sc Gamble’s Value Pricing Strategy. Journal of Marketing, 65(1), 44-61.

Allender, W.J., Sc Richards, T. (2012). Brand Loyalty and Price Promotion Strategies: an Empirical Analysis. Journal of Retailing, 88(3), 323-342.

Anderson, E.T., Sc Fox, E.J. (2019). How Price Promotions Work: A Review of Practice. In J.-P. Dube, Sc P.E. Rossi, Handbook of the Economics of Marketing (1st ed., Vol. 1, pp. 497-552). Amsterdam: North-Holland.

Bailey, A.A. (2008). Evaluating Consumer Response to EDLPs. Journal of Retailing and Consumer Services, 15(3), 211-223.

Besanko, D., Dube, J.-P., Sc Gupta, S. (2005). Own-Brand and Cross-Brand Retail Pass-Through. Marketing Science, 24(1), 110-122.

Blattberg, R.C., Sc Briesch, R.A. (2012). Sales Promotions. In O. Ozer, Sc R. Phillips, The Oxford Handbook of Pricing Management (pp. 585-619). Oxford: Oxford University Press.

Blattberg, R.C., Briesch, R., Sc Fox, E.J. (1995). How Promotions Work. Marketing Science, 14(3), G1-G236.

Blattberg, R.C., Sc Fox, E.J. (1995). Category Management: A Series of Implementation Guides (Vols. I-V). Washington DC: Food Management Institute.

Blattberg, R.C., 8c Neslin, S.A. (1989). Sales Promotion: The Long and the Short of it. Marketing Letters, 1, 81-97.

Bogomolova, S., Dunn, S., Trinh, G., Taylor, J., Sc Volpe, R.J. (2015). Price Promotion Landscape in the US and UK: Depicting Retail Practice to Inform Future Research Agenda. Journal of Retailing and Consumer Services, 25, 1-11.

Bogomolova, S., Szabo, M., 8c Kennedy, R. (2017). Retailers’ and Manufacturers’ Price-Promotion Decisions: Intuitive or Evidence-Based? Journal of Business Research, 17, 189-200.

Bolton, R.N., 8c Shankar, V. (2003). An Empirically Derived Taxonomy of Retailer Pricing and Promotion Strategies. Journal of Retailing, 79(4), 213-224.

Bolton, R.N., Shankar, V., 8c Montoya, D.Y. (2010). Recent Trends and Emerging Practices in Retailer Pricing. In M. Krafft, 8c M.K. Mantrala, Retailing in the 21st Century (pp. 301-318). Berlin: Springer.

Boston Consulting Group (2012). Consumer Products Trade Spend Benchmark. Boston, MA: The Boston Consulting Group, Inc (BCG).

Briesch, R.A., Dillon, W.R., 8c Fox, E.J. (2013). Category Positioning and Store Choice: The Role of Destination Categories. Marketing Science, 32(3), 488-509.

Brito, P.Q., 8c Hammond, K. (2007). Strategic Versus Tactical Nature of Sales Promotions. Journal of Marketing Communications, 13(2), 131-148.

Buil, I., de Chernatony, L., 8c Martinez, E. (2012). Examining the Role of Advertising and Sales Promotions in Brand Equity Creation. Journal of Business Research, 66(1), 115-122.

Cacchiarelli, L., 8c Sorrentino, A. (2019). Pricing Strategies in the Italian Retail Sector: The Case of Pasta. Social Sciences, 8(4), 113.

Chevalier, J.A., Kashyap, A.K., 8c Rossi, P.E. (2003). Why Don’t Prices Rise During Periods of Peak Demand? Evidence from Scanner Data. The American Economic Review, 93(1), 15-37.

DelVecchio, D., Henard, D.H., 8c Freling, T.H. (2006). The Effect of Sales Promotion on Post-Promotion Brand Preference: A Meta-Analysis. Journal of Retailing, 82(3), 203-213.

Dreze, X., 8c Bell, D.R. (2003). Creating Win-Win Trade Promotions: Theory and Empirical Analysis of Scan-Back Trade Deals. Marketing Science, 22(1), 16-39.

Dubois, D., 8c Prade, H. (1980). Systems of Linear Fuzzy Constraints. Fuzzy Sets and Systems, 3(1), 37-48.

Ellickson, P.B., 8c Misra, S. (2008). Supermarket Pricing Strategies. Marketing Science, 27(5), 811-828.

Ellickson, P.B., Misra, S., 8c Nair, H.S. (2012). Repositioning Dynamics and Pricing Strategy. Journal of Marketing Research, 49(6), 750-772.

Empen, J., Loy, J.-P., 8c Weiss, C. (2015). Price Promotions and Brand Loyalty: Empirical Evidence for the German Ready-to-Eat Cereal Market. European Journal of Marketing, 49(5/6), 736-759.

Fassnacht, M., 8c El Husseini, S. (2013). EDLP Versus Hi-Lo Pricing Strategies in Retailing - A State of the Art Article. Journal of Business Economics, 83, 259-289.

Gartner. (2015). Market Guide for Trade Promotion Management and Optimization. Stamford, CT: Gartner, Inc.

Gauri, D.K., Trivedi, M., 8c Grewal, D. (2008). Understanding the Determinants of Retail Strategy: An Empirical Analysis. Journal of Retailing, 84(3), 256-267.

Gedenk, K., 6c Neslin, S.A. (1999). The Role of Retail Promotion in Determining Future Brand Loyalty: Its Effect on Purchase Event Feedback. Journal of Retailing, 75(4), 433-459.'

Gedenk, K., Neslin, S.A., 6c Ailawadi, K.L. (2010). Sales Promotion. In M. Krafft, 6c M.K. Mantrala, Retailing in the 21st Century (2nd ed., pp. 393-408). Berlin: Springer.

Gomez, M.J., Rao, V.R., 6c McLaughlin, E.W. (2007, August). Empirical Analysis of Budget and Allocation of Trade Promotions in the U.S. Supermarket Industry. Journal of Marketing Research, 44(3), 410-424.

Grewal, D., Janakiraman, R., Kalyanam, K., Kannan, P., Ratchford, B., Song, R., 6c Tolerico, S. (2010). Strategic Online and Offline Retail Pricing: A Review and Research Agenda. Journal of Interactive Marketing, 24(2), 138-154.

Herrmann, R., Moeser, A., 6c Weber, S.A. (2005). Price Rigidity in the German Grocery-Retailing Sector: Scanner-Data Evidence on Magnitude and Causes. Journal of Agricultural & Food Industrial Organization, 3(1), 1-35.

Hoch, S.J., Dreze, X., 6c Purk, M.E. (1994). EDLP, Hi-Lo, and Margin Arithmetic. Journal of Marketing, 58(4), 16-27.

International Pasta Organisation. (2014). The World Pasta Industry Status Report 2013. Roma, Italy: International Pasta Organization (IPO).

IRI. (2016,11). Price and Promotion in Western Economies. Retrieved 02.23. 2020, from Iriworldwide.com: https://www.iriworldwide.com/IRI/media/IRI-Clients/ International/Price-Promotion-in-Western-Economies-a-Pause-in-Promotion-Es calation.pdf

Jedidi, K., Mela, C.F., 6c Gupta, S. (1999). Managing Advertising and Promotion for Long-Run Profitability. Marketing Science, 18(1), 1-22.

Johnson, L.K. (2003). Dueling Pricing Strategies. MIT Sloan Management Review, 44(3), 10-11.

Kuntner, T., 6c Teichert, T. (2016). The Scope of Price Promotion Research: An Informetric Study. Journal of Business Research, 69(8), 2687-2696.

Lai, R., 6C Rao, R. (1997). Supermarket Competition: The Case of Every Day Low Pricing. Marketing Science, 16(1), 60-80.

McShane, B.B., Chen, C., Anderson, E.T., 6c Simester, D.I. (2016). Decision Stages and Asymmetries in Regular Retail Price Pass-Through. Marketing Science, 35(4),619-639.

Mela, C.F., Gupta, S., 6c Lehmann, D.R. (1997). The Long-Term Impact of Promotion and Advertising on Consumer Brand Choice. Journal of Marketing Research, 34(2), 248-261.

Meza, S., 6C Sudhir, K. (2006). Pass-Through Timing. Quantitative Marketing and Economics, 4(4), 351-382.

Mourdoukoutas, P. (2017, 2 24). A Strategic Mistake That Still Haunts JC Penney. Retrieved January 11, 2020, from Forbes: https://www.forbes.com/sites/panosm ourdoukoutas/2017/02/24/a-strategic-mistake-that-still-haunts-jc-penney/#49 68b5041bcf

Neslin, S.A. (2002). Sales Promotions. In B.A. Weitz, 6c R. Wensley, Handbook of Marketing (pp. 310-338). London: SAGE Publications Ltd.

Nielsen. (2009a, 06 17). Promotions Spur Growth In UK Grocery Sales. Retrieved 02 23, 2020, from Nielsen.com: https://www.nielsen.com/us/en/insights/article /2009/promotions-spur-growth-in-uk-grocery-sales/

Nielsen. (2009b, 10 05). Almost Half of U.S. Supermarket Purchases are Sold on Promotion. Retrieved 02 23, 2020, from Nielsen.com: https://www.nielsen.com /us/en/insights/article/2009/almost-half-of-u-s-supermarket-purchases-are-sold -on-promotion/

Nielsen. (2015a, 02). The Path to Efficient Trade Promotions. Retrieved 02 23, 2020, from Nielsen.com: https://www.nielsen.eom/wp-content/uploads/sites/3 /2019/04/the-path-to-efficient-trade-promotions-feb-2015.pdf

Nielsen. (2015b, 06 25). Consumer Loyalty Is Not Much Deeper Than Our Pockets. Retrieved 03 05, 2020, from Nielsen.com: https://www.nielsen.com/us/en/insig hts/article/2015/consumer-loyalty-is-not-much-deeper-than-our-pockets/

Nielsen. (2018a, 08 03). Are We Really Getting Value From Our Promotions? Retrieved 02 23, 2020, from Nielsen.com: https://www.nielsen.com/au/en/insig hts/article/2018/are-we-really-getting-value-from-our-promotions/

Nielsen. (2018b, 05 15). The S470M Opportunity: Changing The Game With Price and Promotions. Retrieved 02 23, 2020, from Nielsen.com: https://www.nielsen. com/nz/en/insights/article/2018/the-470m-opportunity-changing-the-game-with -price-and-promotions/

Nielsen. (2019, 08 07). Global Ad Spending Is Up, But You Can’t Buy Loyalty in CPG. Retrieved 03 05, 2020, from Nielsen.com: https://www.nielsen.com/us/en/ insights/article/2019/global-ad-spending-is-up-but-you-cant-buy-loyalty-in-cpg/

Nielsen. (2020). The Ultimate Guide to Nielsen Retail Measurement Services. Retrieved 03 05, 2020, from Nielsen.com: https://www.nielsen.com/us/en/clien t-learning/retail-measurement-services/analyzing-promotion/

Pettie, S., & Pettie, K. (2003). Sales Promotion. In M.J. Baker, The Marketing Book (5th ed., pp. 458-484). Oxford: Butterworth Heinemann.

Slater, J. (2001). Is Couponing an Effective Promotional Strategy? An Examination of the Procter &C Gamble Zero-Coupon Test. Journal of Marketing Communications, 7(1), 3-9.

Srinivasan, S., Pauwels, K., Hanssens, D., & Dikempe, M. (2004). Do Promotions Benefit, Manufacturers, Retailers, or Both? Management Science, 50(5), 617-629.

Zadeh, L.A. (1965). Fuzzy Sets. Information and Control, 8(3), 338-353.

Zadeh, L.A. (1968). Probability Measures of Fuzzy Events. Journal of Mathematical Analysis and Applications, 23, 421-427.

Zadeh, L.A. (1977). Fuzzy Sets and Their Application to Pattern Classification and Clustering Analysis. Classification and Clustering. Proceedings of an Advanced Seminar Conducted by the Mathematics Research Center, the University of Wisconsin-Madison, May 3-5, 1976 (pp. 251-299). New York: Academic Press.

Zimmermann, H.-J. (2001). Fuzzy Set Theory - and Its Applications (4th ed.). New York: Springer Sience+Business Media.

Zoellner, F., &C Schaefers, T. (2015). Do Price Promotions Help or Hurt Premium- product Brands? The Impact of Different Price-promotion Types on Sales and Brand Perception. Journal of Advertising Research, 55(3), 270-283.

Zwanka, R.J. (2017). Everyday Low Pricing: A Private Brand Growth Strategy in Traditional Food Retailers. Journal of Food Products Marketing, 24(4), 373-391.

 
Source
< Prev   CONTENTS   Source