Data Analysis Machine Learning and Applications Episode 2 Part 9 pdf

The discovery of association rules is a popular approach to detect cross-category purchase correlations hidden in large amounts of transaction data and extensive retail ments.. The ob-je

impact Breaking a promise and experiencing poor quality of repair work influence

solely satisfaction ratings (J36= −.23, p < 01 and J39 = −.32, p < 01), whereas CIs classified as showing no goodwill and restriction to basic service lowered customers

trust in the service provider (J28= −.14, p < 01 and J2 10 = −.14, p < 01) The incident category which should be primarily avoided is negative behaviors toward the customer, since it clearly has the most damaging impact on the customer-firm re-

lationship, due to its dual influence on trust (J2 11= −.27, p < 01) and satisfaction

(J3 11= −.26, p < 01) Interestingly, only one of the positive CI categories

(offer-ing additional service) impacts on satisfaction with the repair department (J33= 23,

p < 01) and none impacts on trust.

Fig 1 MIMIC model: CI categories and their impact on relationship measures, significant

path coefficients are depicted

5 Discussion

Even though several papers in the marketing literature have raised the questionwhether and which incidents are really critical for a customer-firm relationship (Ed-vardsson & Strandvik, 2000) ours is the first study to explicitly address this ques-tion In the present study, we conducted CI interviews without restricting valence

and number of incidents reported, and assessed their impact on measures of tionship quality Our results confirm that positive and negative incidents possess apartially asymmetric impact on satisfaction and trust Negative incidents have partic-ularly damaging effects on a relationship through their strong impact on trust (totalcausal effect: 0.58) These results are in stark contrast to Odekerken-Schröder et al.’s(2000) conclusion, that CIs do not play a significant role for developing trust Fur-ther the damage inflicted by negative incidents can hardly be “healed” with verypositive experiences, since the total causal effect of the number of positive incidents

rela-on trust is substantially smaller (0.12) Thus, management should clearly put sis on avoiding negative interaction experiences The employed MIMIC approachfollowed Gremler’s call (2004, p 79) to “determine which events are truly critical tothe long-term health of the customer-firm relationship” and revealed which specificincident categories have a particular strong impact on relationship health and should

empha-be avoided with priority, such as negative empha-behavior toward the customer The lected vivid verbatim stories from the customer’s perspective provide very concreteinformation for managers and can be easily communicated to train customer-contactpersonnel (Zeithaml & Bitner, 2003; Stauss & Hentschel, 1992) For further studies,

col-as pointed out by one of the reviewers, an alternative evaluation possibility would be

to measure the experienced severity of the experienced CI-categories instead of theirmere occurrence

References

BARON, R and KENNY, D (1986): The Moderator-Mediator Variable Distinction in Social

Psychological Research: Conceptual, Strategic and Statistical Considerations Journal of Personality and Social Psychology, 51 (6), 1173-1182.

BAUMEISTER, R F., BRATSLAVSKY, E., FUNKENAUER, C., and VOHS, K D (2001):

Bad is Stronger than good Review of General Psychology, 5 (4), 323-370.

BITNER, M J., BOOMS, B H., and TETREAULT, M S (1990): The Service Encounter

-Diagnosing Favorable and Unfavorable Incidents Journal of Marketing, 54(1), 71-84.

BOLLEN, K A (1989): Structural Equations with Latent Variables New York: Wiley.EDVARDSSON, B (1992) Service Breakdowns: A Study of Critical Incidents in an Airline

International Journal of Service Industry Management, 3(4), 17-29.

EDVARDSSON, B., and STRANDVIK, T (2000): Is a Critical Incident Critical for a

Cus-tomer Relationship? Managing Service Quality, 10(2), 82-91.

EICH E, MACAULAY D., and RYAN L (1994): Mood Dependent Memory for Events of the

Personal Past Journal of Experimental Psychology - General, 123 (2), 201-215.

FISKE, S (1980): Attention and Weight in Person Perception - the Impact of Negative and

Extreme Behaviour Journal of Personality and Social Psychology, 38 (6), 889-906 FORGAS, J P (1995): Mood and Judgment: The Affect Infusion Model (AIM) Psychological Bulletin, 117 (1), 39-66.

FORNELL, C., JOHNSON, M D., ANDERSON, E W., CHA, J., and BRYANT, B E (1996):

The American Customer Satisfaction Index: Nature, Purposes, and Findings Journal of Marketing, 60 (October), 7-18.

GEYSKENS, I., STEENKAMP, J-B E M., and KUMAR, N (1999): A Meta-Analysis of

Sat-isfaction in Marketing Channel Relationships Journal of Marketing Research, 36 (May),

223-238

GREMLER, D (2004): The Critical Incident Technique in Service Research Journal of vice Research, 7(1), 65-89.

Ser-JÖRESKOG, K and SÖRBOM, D (2001): LISREL 8: User’s Reference Guide Chicago:Scientific Software International

KAHNEMAN, D and TVERSKY, A (1979): Prospect Theory - Analysis of Decision under

Risk Econometrica, 47(2), 263-291.

MORGAN, R M and HUNT, S D (1994): The commitment-trust theory of relationship

marketing Journal of Marketing, 58(3), 20-38.

ODEKERKEN-SCHRÖDER, G., van BIRGELEN, M., LEMMINK, J., de RUYTER, K., andWETZELS, M (2000): Moments of Sorrow and Joy: An Empirical Assessment of theComplementary Value of Critical Incidents in Understanding Customer Service Evalua-

tions European Journal of Marketing, 34(1/2), 107-125.

ROOS, I (2002): Methods of Investigating Critical Incidents - A Comparative Review Journal

of Service Research, 4 (3), 193-204.

SINGH, J and SIRDESHMUKH, D (2000): Agency and Trust Mechanisms in Consumer

satisfaction and loyalty judgments Journal of the Academy of the Marketing Science, 28

(1), 150-167

STAUSS, B and HENTSCHEL, B (1992): Attribute-based versus Incident-based ment of Service Quality: Results of an Empirical Study in the German Car Service In-

Measure-dustry In: P Kunst and J Lemmink, (Eds.), Quality Management in Services (59-78).

STAUSS, B and WEINLICH, B (1997): Process-oriented Measurement of Service Quality

- Applying the Sequential Incident Technique European Journal of Marketing, 31 (1),

33-55

SZYMANSKI, D M and HENARD, D H (2001): Customer Satisfaction: A Meta-Analysis

of the Empirical Evidence Academy of Marketing Science, 29(1), 16-35.

TAYLOR, S (1991): Asymmetrical Effects of Positive and Negative Events: The

Mobilization-Minimization Hypothesis Psychological Bulletin, 110 (1), 67-85.

YBARRA, O and STEFAN, W G, (1999): Attributional Orientations and the Prediction of

Behavior: The Attribution-Prediction Bias Journal of Personality and Social Psychology,

76 (5), 718-728

ZEITHAML, V A and BITNER, M J (2003): Services Marketing: Integrating Customer Focus across the Firm (3 rded.) New York: McGraw-Hill

Target Marketing

Nicolas March and Thomas ReuttererInstitute for Retailing and Marketing, Vienna University of Economics and BusinessAdministration, Augasse 2–6, 1090 Vienna, Austria

march@troostwijk.de

thomas.reutterer@wu-wien.ac.at

Abstract The discovery of association rules is a popular approach to detect cross-category

purchase correlations hidden in large amounts of transaction data and extensive retail ments Traditionally, such item or category associations are studied on an ’average’ view of themarket and do not reflect heterogeneity across customers With the advent of loyalty programs,however, tracking each program member’s transactions has become facilitated, enabling re-tailers to customize their direct marketing efforts more effectively by utilizing cross-categorypurchase dependencies at a more disaggregate level In this paper, we present the buildingblocks of an analytical framework that allows retailers to derive customer segment-specificassociations among categories for subsequent target marketing The proposed procedure startswith a segmentation of customers based on their transaction histories using a constrained ver-

assort-sion of K-centroids clustering In a second step, associations are generated separately for each

segment Finally, methods for grouping and sorting the identified associations are provided.The approach is demonstrated with data from a grocery retailing loyalty program

In order to identify and to make use of possible cross-selling potentials, the posed approach builds on techniques for exploratory analysis of market basket data.Retail managers have been interested in better understanding the purchase interde-pendency structure among categories for quite a while One obvious reason is thatknowledge about correlated demand patterns across several product categories can

pro-be exploited to foster cross-buying effects using suitable marketing actions For

ex-ample, if customers often buy a particular product A together with article B, it could

be useful to promote A in order to boost sales volumes of B, and vice versa The

ob-jective of exploratory market basket analysis is to discover such unknown cross-itemcorrelations from a typically huge collection of purchase transaction data (so-calledmarket baskets) accruing at the retailer’s point-of-sale scanning devices (Berry andLinoff (2006)) Among others, algorithms for mining association rules are populartechniques to accomplish this task (cf., e.g Hahsler et al (2006)) However, suchassociation rules are typically derived for the entire data set of available retail trans-actions and thus reflect an ’average’ or aggregate view of the market only

In recent years, many retailers have tried to improve their CRM activities bylaunching loyalty programs, which provide their members with bar-coded plastic

or registered credit cards If customers use these cards during their payment process,they get a bonus, credits or other rewards As a side effect, these transactions becomepersonally identifiable by linking them back to the corresponding customers Thus,retailers are nowadays collecting series of market baskets that represent (more orless) complete buying histories of their primary clientele over time

2 A segment-specific view of cross-category associations

To exploit the potential benefits offered by such rich information on customers’ chasing behavior within advanced CRM programs, cross-category correlations need

pur-to be detected on a more disaggregate (or cuspur-tomer segment) level instead of anaggregate level Attempts towards this direction are made by Boztug and Reutterer(2007) or Reutterer et al (2006) The authors employ vector quantization techniques

to arrive at a set of ’generic’ (i.e., customer-unspecific) market basket classes withinternally more distinctive cross-category interdependencies In a second step theygenerate a segmentation of households based on a majority voting of each house-hold’s basket class assignments throughout the individual purchase history Thesesegments are proposed as a basis for designing customized target marketing actions

In contrast to these approaches, the procedure presented below adopts a novelcentroids-based clustering algorithm proposed by Leisch and Grün (2006), whichbypasses the majority voting step for segment formation This is achieved by a cross-category effects sensitive partitioning of the set of (non-anonymous) market basketdata, which imposes group constraints determined by the household labels associatedwith each of the market baskets Hence, during the iterative clustering process thesingle transactions are "forced" to keep linked with all the other transactions of aspecific household’s buying history This results in segments whose members can becharacterized by distinctive patterns of cross-category purchase interrelationships

To get a better feeling of the inter-category purchase correlations within the viously identified segments, association rules derived separately for each segmentand evaluated by calculating various measures of significance and interestingnesscan assist marketing managers for further decision making on targeted marketing ac-tions Although the within-segment cross-category associations are expected to differsignificantly from those generated for the unsegmented data set (because of the datacompression step employed prior to the analysis), low minimum thresholds of such

pre-measures typically still result in a huge number of potentially interesting tions To arrive at a clearer and managerially more traceable overview of the varioussegment-specific cross-category purchase correlations, we arrange them based on adistance concept suggested by Gupta et al (1999).

associa-The next section characterizes the building blocks of the employed methodology

in more detail Section 4 empirically illustrates the proposed approach using a action data set from a grocery retailing loyalty program and presents selected results.Section 5 closes the article with a summary and an outlook on future research

trans-3 Methodology

The conceptual framework of the proposed approach is depicted in Figure 1 and

con-sists of three basic steps: First, a modified K-centroids cluster algorithm partitions the entire transaction data set and defines K segments of households with an inter-

est in similar category combinations Secondly, the well-known APRIORI algorithm(Agrawal et al (1993)) searches within each segment for specific frequent itemsets,which are filtered by a suitable measure of interestingness Finally, the associationsare grouped via hierarchical clustering using a distance measure for associations

Fig 1 Conceptual framework of the proposed procedure

Step 1: Each transaction or market basket can be interpreted as a J-dimensional binary vector x n = [1,0] J with j = 1,2 J categories A value of one refers to the

presence and a zero to the absence of an item in the market basket Integrated into a

binary matrix X N, the rows correspond to transactions while each column represents

an item Let the set I pdescribe a group constraint indicating the buying history of

customer p = 1,2, P with {x i ∈ X N |i ∈ I p } The objective function for a modified K-centroids clustering respecting group constraints is (Leisch and Grün (2006)):

An iterative algorithm for solving Equation 1 requires calculation of the closest

centroid c (.) for each transaction x i according to the distance measure d (.) at each

iteration To cope with the usually sparse binary transaction data and to make thepartition cross-category effects sensitive, the Jaccard coefficient, which gives moreweight the co-occurrences of ones rather than common zeros, is used as an appropri-ate distance measure (cf Decker (2005)) Notice that in contrast to methods like the

K-means algorithm, instead of single transactions groups of market baskets as given

by I p (i.e., customer p’s complete buying history) need to be assigned to a minimum distant centroid This is warranted by a function f (x i) that determines the centroidclosest to the majority of the grouped transactions (cf Leisch and Grün (2006))

In order to achieve directly accessible and more intuitively interpretable results,

we can calculate cluster-wise means for updating the prototype system instead ofoptimized canonical binary centroids This results in an ’expectation-based’ cluster-ing solution (cf Leisch (2006)), whose centroids are equivalent to segment-specificchoice probabilities of the corresponding categories Notice that the segmentation

of households is determined such that each customer’s complete purchase historypoints exclusively to one segment Thus, in the present application context the set

of K centroids can be interpreted as prototypical market baskets that summarize the

most pronounced item combinations demanded by the respective segment membersthroughout their purchase history An illustrative example is provided in Table 1 ofthe subsequent empirical study

Step 2: The centroids derived in the segmentation step already provide some

indications on the general structure of the cross-item interdependencies within thehousehold segments To get a more thorough understanding, interesting categorycombinations (so called itemsets) can be further explored by the APRIORI algorithmusing a user defined support value For the entire data set, the support of an arbitrary

itemset A is denoted by supp (A) =| {x n ∈ X N | A ⊆ x n } | / | N | and defines the fraction

of transactions containing itemset A Notice that in the present context, however,

itemsets are generated at the level of previously constructed segments

The itemsets are called frequent if their support is above a user-defined old value, which implies their sufficient statistical importance for the analyst Togenerate a wide range of associations, rather low minimum support values are usu-ally preferred Because not all associations are equally meaningful, an additionalmeasure of interestingness is required to filter the itemsets for evaluation purposes

thresh-Since our focus is on itemsets, asymmetric measures like confidence or lift are less

useful (cf Hahsler (2006)) We advocate here the so-called all-confidence measureintroduced by Omiecinski (2003), which is the minimum confidence value for allrules that can be generated from the underlying itemset Formally it is denoted by

allcon f (A) = supp(A)/max B⊂A {supp(B)} for all frequent subsets B with B ⊂ A Step 3: Although the all-confidence measure can assist in reducing the number of

itemsets considerably, in practice it can still be difficult to handle several hundreds ofremaining associations For an easier recognition of characteristic inter-item corre-lations within each segment, the associations can be grouped based on the followingJaccard-like distance measure for itemsets (Gupta et al (1999)):

D (A,B) = 1 − | m(A ∪ B) |

| m(A) | + | m(B) | − | m(A ∪ B) | (2)

Expression m (.) denotes the set of transactions containing the itemset From

Equation 2 it should be evident that the distance between two itemsets tends to belower if the involved itemsets occur in many common transactions This propertyqualifies the measure to determine specific groups of itemsets that share some com-mon aspects of consumption behavior (cf Gupta et al (1999))

4 Empirical application

The following empirical study illustrates some of the results obtained from the dure described above We analyzed two samples of real-world transaction data, eachrealized by 3,000 members of a retailer’s loyalty program The customers made onaverage 26 shopping trips over an observational period of one year Each transactioncontains 268 binary variables, which represent the category range of the assortment

proce-To achieve managerially meaningful results, preliminary screening of the datasuggested the following adjustments of the raw data:

1 The purchase frequencies are clearly dominated by a small range of categories,such as fresh milk, vegetables or water (see Figure 2) Since these categoriesare bought several times by almost every customer during the year under in-vestigation, they provide relatively low information on the differentiated buyinghabits of the customers The opposite is supposed to be true for categories withintermediate or lower purchase frequencies Therefore, we decided to eliminatethe upper 52 categories (left side of the vertical line in Figure 2), which occur

in more than 10% of all transactions The resulting empty baskets are excludedfrom the analysis as well

0.0 0.1 0.2 0.3 0.4 0.5

Fig 2 Distribution of relative category purchase frequencies in decreasing order

2 To include households with sufficiently large buying histories, households withless than six store visits per year were eliminated In addition, the upper fivepercentage quantile of households, which use their customer cards extremelyoften, were deleted

To find a sufficiently stable cluster solution with a minimum within-sum of tances, the transactions made by the households from the first sample are split into

dis-three equal sub samples and clustered up to fifteen times each In each case, thebest solution is kept for the following sub sample to achieve stable results The con-verged set of centroids of the third sub sample is used for initialization of the secondsample Commonly used techniques for determination of the number of clusters rec-

ommended K= 11 clusters as a decent and well-manageable number of householdsegments Given these specifications, the partitioning of the second sample using theproposed cluster algorithm detects some segments, which are dominated by categorycombinations typically bought for specific consumption or usage purposes and othertypes of categorical similarities For example, Table 1 shows an extract of a centroidvector including the top six categories in terms of highest conditional purchase prob-abilities in a segment of households denoted as the "wine segment" A typical marketbasket arising from this segment is expected to contain red/rosé wines with a proba-bility of 32.3 %, white wines with a probability of 22.5 %, etc Hence, the labeling

"wine segment"

Equally, other segments may be characterized by categories like baby food/care

or organic products On the other hand, there is also a small number of segments withcategory interrelationships that cannot be easily explained However, such segmentsmight provide some interesting insights into the interests of households which are sofar unknown

Table 1 Six categories with highest purchase frequencies in the wine segment

No Category Purchase frequency

1 red / rosé wines 0.3229143

on the most interesting frequent itemsets, only the 150 itemsets with highest confidence values are considered for grouping according to step 3 of the procedure.Grouping the frequent itemsets intends to rearrange the order of the generated(segment-specific) associations and to focus the view of the decision maker on char-acteristic item correlations The distance matrix derived by Equation 2 is used asinput for hierarchical clustering according to the Ward algorithm Figure 3 shows thedendrogram for the 150 frequent itemsets within the wine cluster Again, it is not

all-straightforward to determine the correct number of groups g h Frequently proposedheuristics based on plotted heterogeneity measures does not help here Therefore, we

pass the distance matrix to the partition around medoid (PAM) algorithm of

Kauf-man and Rousseeuw (2005) for several g hvalues Using the maximum value of theaverage silhouette width for a sequence of partitions thirty groups of itemsets areproposed In Figure 3 the grey rectangles mark two exemplary chosen clusters of as-sociations The corresponding associations of the right hand group are summarized

in Table 2 and clearly indicate an interest of some of the wine households in hardalcoholic beverages

Fig 3 Dendrogram of 150 frequent itemsets mined from transactions of the wine segment

Table 2 Associations of hard alcoholic beverages within the wine segment

No association support all-confidence

2 {brandy, fruit brandy} 0.015 0.18

3 {fruit brandy, appetizers} 0.018 0.17

4 {brandy, appetizers} 0.016 0.15

5 {whisky, fruit brandy} 0.011 0.14

To examine whether the segment-specific associations differ from those ated within the whole data set, we have drawn and analyzed random samples withthe same amount of transactions as each of the segments The comparison of thefrequent itemsets mined in the random sample and those from the segment-specifictransactions shows that some segment-specific association groups clearly represent

gener-a unique chgener-argener-acteristic of their underlying household segment Of course, this is nottrue in any case For example, the association group marked by the grey rectangle

on the left-hand side in Figure 3 can be found in almost every random sample orsegment It denotes correlations between categories of hygiene products

5 Conclusion and future work

We presented an approach for identification of household segments with tive patterns and subgroups of cross-category associations, which differ from thosemined in the entire data set The proposed framework enables retailers to segmenttheir customers according to their past interest in specific item combinations Themined segment-specific associations provide a good basis for deriving more respon-sive recommendations or designing special offers through target marketing activities.Nevertheless, the stepwise procedure has it’s natural limitations imposed by thefact that later steps are dependent on the outcome of former stages A simultaneousapproach would disburden decision makers from determining various model param-eters (like support thresholds, number of segments) at each stage Another drawback

distinc-is the ad-hoc exclusion of very frequently purchased categories, which could be stituted in future applications by a data driven weighting scheme

sub-References

AGRAWAL, R., IMIELINSKI, T and SWAMI, A (1993): Mining association rules between

sets of items in large databases In: Proceedings of the ACM SIGMOD International Conference on Management of Data Washington D.C., 207–216.

BERRY, M and LINOFF G (2004): Data mining techniques Wiley, Indianapolis.

BOZTUG, Y and REUTTERER, T (2007): A combined approach for segment-specific

anal-ysis of market basket data In: European Journal of Operational Research, forthcoming DECKER, R (2005): Market Basket Analysis by Means of a Growing Neural Network The International Review of Retail, Distribution and Consumer Research, 15, 151–169.

GUPTA, G K., STREHL, A and GOSH, J (1999): Distance based clustering of association

rules In: Intelligent Engineering Systems Through Artificial Neural Networks ASME

Press, New York, 759–764

HAHSLER, M., HORNIK, K and REUTTERER, T (2006): Implications of ProbabilisticData Modeling for Mining Association Rules In: M Spiliopoulou, R Kruse, C Borgelt,

A Nürnberger, W Gaul (Eds.): From Data and Information Analysis to Knowledge gineering Springer, Heidelberg, 598–605.

En-KAUFMAN, L and ROUSSEEUW, P J (2005): Finding Groups in Data: An Introduction to Cluster Analysis., Wiley, New York.

LEISCH, F and GRÜN, B (2006): Extending standard cluster algorithms to allow for group

constraints In: A Rizzi, M Vichi (Eds.): Compstat 2006, Proceedings in Computational Statistics Physica-Verlag, Heidelberg, 885–892.

LEISCH, F (2006): A toolbox for k-centroids cluster analysis In: Computational Statistics and Data Analysis, 51(2), 526–544.

OMIECINSKI, E (2003): Alternative Interest Measures for Mining Associations in

Databases In: IEEE Transactions on Knowledge and Data Engineering, 15(1), 57–69.

PRINZIE, A and VAN DEN POEL, D (2005): Constrained optimization of data-mining

prob-lems to improve model performance A direct-marketing application In: Expert Systems with Applications, 29, 630–640.

REUTTERER, T., MILD, A., NATTER, M and TAUDES, A (2006): A dynamic

segmenta-tion approach for targeting and customizing direct marketing campaigns In: Journal of Interactive Marketing, 20(3), 43–57.

Ralf WagnerSVI Chair for International Direct Marketing

DMCC - Dialog Marketing Competence Center,

University of Kassel, Germany

rwagner@wirtschaft.uni-kassel.de

Abstract This paper introduces a finite-mixture version of the adjacent-category logit model

for the classification of companies with respect to their marketing practices The classification

results are compared to conventional K-means clustering, as established for clustering

market-ing practices in current publications Both, the results of this comparison as well as a canonicaldiscriminant analysis, emphasize the opportunity to offer fresh insights and to enrich empiricalresearch in this domain

1 Introduction

Although emerging markets and transition economies are attracting increasing tion in marketing, Pels and Brodie (2004) argue that conventional marketing knowl-edge is not valid for these markets per se Moreover, Burgess and Steenkamp (2006)claim that emerging markets offer unexploited research opportunities due to their sig-nificant departures from the assumptions of theories developed in the Western world,but call for more rigorous research in this domain But, the majority of studies con-cerned with marketing in transition economies are either qualitative descriptive orrestricted to simple cluster analysis This paper seizes the challenge by:

atten-• introducing a finite-mixture approach facilitating the fitting of a response modeland the clustering of observations simultaneously,

• investigating whether or not the Western-type distinction between marketing mixmanagement and relationship management holds for groups of companies fromRussia and Lithuania, and

• exploring the consistency of the marketing activities

The remainder of this paper is structured as follows The next section provides a

description of the research approach, which is embedded in the Contemporary keting Practices (CMP) Project In the third section, a finite-mixture approach is

Mar-introduced and criteria for determining the number of clusters in the data are cussed The data and the results of this study are outlined in section 4, and section 5concludes with a discussion of these results