Customer Heterogeneity: FEB12018X Lecture 2 PDF

Customer Heterogeneity. All Customers Differ. Empirical Marketing (FEB12018X) Radek Karpienko 5 Sources of Customer Heterogeneity • Individual differences • Life experiences • Functional needs • Self -identity/image • Marketing activities 3 Source Description Examples Individual difference...

Customer Heterogeneity. All Customers Differ. Empirical Marketing (FEB12018X) Radek Karpienko 5 Sources of Customer Heterogeneity • Individual differences • Life experiences • Functional needs • Self -identity/image • Marketing activities 3 Source Description Examples Individual differences A person’s stable and consistent way of responding to the environment in a specific domain Favorite colors, Big 5 personality traits (openness, conscientiousness, extraversion, agreeableness, neuroticism) Life experiences An individual’s life experiences capture events and experiences unique to his or her life that have lasting impact on the value and preference he or she places on products and services, which in turn affects preferences independent of individual differences A child raised closer to the equator, in warmer climates, will typically have a higher preference for spicy foods, as a carryover of past periods when spices were used to preserve and help mask the taste of food more likely to spoil in warmer climates Functional needs An individual’s personal decision weightings across functional attributes based on his or her personal circumstances What price can they afford to pay (income), how long does the product need to last (quality, warranty), when will they use the product (battery powered, size), and are there any special usage features that they need (waterproof)? Self -identity/image Customers actively seek products that they feel will support or promote their desired self -image Motorcycle riders often wear leather (functional and image driven,) and Goths like the color black because of their desire to identify with the image of a specific user or social group Marketing activities Firms’ attempts to build linkages between their brands and prototypical identities or meanings BMW paid $25 million to have James Bond drive a BMW in the movie Skyfall , based on the belief that Bond’s image would be aspirational to many potential target customers (e.g., men aged from 30 to 50 years) Five Sources of Customer Heterogeneity Latent Customer Heterogeneity • Latent customer heterogeneity is defined as potential differences in desires that are unobserved and have not become manifest in purchase preferences or behaviors yet • Latent customer heterogeneity can stem from several constraints : • Legal constraints (government regulations, patents) • Economic constraints (prohibitive prices, due to the size of the market or the costs of providing) • Technological constraints (only way known to make something) • Innovative constraints (no firm has yet identified and satisfied the need) 6 Example: AT&T (US) • In 1984, AT&T lost its U.S. government – granted monopoly, so direct competition began • By 1991, the company lost 83% of its sales revenue • Deregulation of this market allowed for the entrance of many new competitors, determined to satisfy customer needs better • Western Electric (the subsidiary of AT&T) came to an end in 1995 7 Agenda Framework for Managing Customer Heterogeneity • Inputs to Managing Customer Heterogeneity Framework • Outputs of Managing Customer Heterogeneity Framework • Process for Managing Customer Heterogeneity Approaches for Managing Customer Heterogeneity • Evolution of Approaches for Managing Customer Heterogeneity • Segmenting, Targeting, and Positioning (STP) Approach • Customer -Centric Approach 8 Marketing Principle #1: All Customers Differ ➔ Managing Customer Heterogeneity Managing Customer Heterogeneity Approaches & Processes Segmenting, targeting, and positioning (STP) Perceptual/positional maps Customer -centric view Analyses Factor analysis Cluster analysis GE matrix Inputs (3Cs) Outputs (STP) All Potential Customers •Needs •Demographics •Size, growth, perceptions Your Company •Strengths and weaknesses •Opportunities and threats Your Competitors •Strengths and weaknesses •Opportunities and threats Industry Segmentation •Customer segments •Needs, demographics, and opportunity of each segment Target Segment •Detailed needs, demographics, and value of target segment(s) •Discriminant function •Relative perceptions Positioning Statement •Who (target segments) •What needs/benefits •Why (relative advantage & support) 9 Inputs to the Managing Customer Heterogeneity Framework 1. All potential c ustomers -- needs, desires, and preferences across customers in an industry, geographic region, market segment, or product category 2. Your c ompany -- an inventory of the company’s strengths, w eaknesses, o pportunities, and threats (SWOT analysis) 3. Your c ompetitors -- an inventory of your competitor’s strengths, w eaknesses, o pportunities, and threats (SWOT analysis) 10 SWOT and 3C Analyses SWOT appraises the strengths , weak -nesses , opportunities , and threats that affect a company’s success. The 3C analysis evaluates customers, competitors, and the company itself. DAT 2.3 Description • To assess strategic marketing decisions by identifying critical internal and external environmental factors that will contribute to the success or failure of the strategy. • A SWOT analysis assesses the internal and external nature of the business, looking at current and future situations. • The 3C analysis emphasizes the need to focus on these three perspectives to gain competitive advantages. When to Use It Inputs • External (Environmental) Factors : relevant legal structure, competitor’s core competencies and market share, changes in customer demographics • Internal (Company -level) Factors : core competencies, market share, competitive advantages How It Works 11 SWOT and 3C Analysis (cont.) DAT 2.3 Description Example The managers of a bakery wish to open a new store in a neighborhood across town. They perform a SWOT and 3C analysis of the e nvi ronment to assess the obstacles they may face. Outputs of Managing Customer Heterogeneity Framework 1. Industry Segmentation describes industry segments and includes, for each named segment, salient purchase preferences, demographic variables, and potential demand opportunities • How can the marketplace be described using homogenous groups? • What does each group of potential customers want? 2. Target Segmentation moves from the overall market landscape to the specific segment(s) of interest, such that it extends the first output by providing a very detailed description of each target segment. • What set of segments will the firm pursue? • How does the firm identify each group of target customers? 3. Positioning Statements encapsulate the three questions into one concise statement that firms use to direct their internal and external marketing activities: who should the firm target, what needs and benefits are being fulfilled, and why does this offering provide a relative advantage over competitive offerings 13 Example of Managing Customer Heterogeneity 1) Identify Customer Segments 2) Select Target Segments 3) Position Against Competitors Gym Socialites Fashion Trend Setters Gym Socialites Fashion Trend Setters Urban Athletes Elite Athletes Seasonal Gym Members Potential Customers Who: Members of high -end, coed gyms What: Good looking but highly functional athletic wear Why: Highest performance materials and design that looks good Who: Fashion -conscious sporting fans What: Athletic wear as clothing Why: Newest, coolest designs that stand out from the crowd 14 Process for Managing Customer Heterogeneity To convert the inputs into outputs, marketers conduct a series of process steps 1. Segmenting – To initiate the segmentation, managers need to identify the key purchase attributes , that is the needs and desires that a potential customer evaluates when making a purchase decision for this category 2. Targeting – The targeting process follows naturally from segmentation, to identify which segments the firm wants to sell to, based on the attractiveness of each segment and the firm’s competitive strength in each segment 3. Positioning – The separation between targeting and positioning is often blurry. Many of the factors used to evaluate competitive strengths to select a target segment also impact the difficulty of executing an effective positioning strategy for that segment 4. Building Customer Centricity – Building a customer -centric organization is different from executing an STP process, in that it requires a top -down, enduring commitment from senior leaders to institute a customer -centric philosophy across the firm’s entire organization 15 Sales Marketing Accounting R&D Operation Sales Marketing Accounting R&D Operation Sales Marketing Accounting R&D Operation Product Analog Semiconductors Digital Signal Processors Wireless Devices CEO Sales Marketing Accounting R&D Operation Sales Marketing Accounting R&D Operation Sales Marketing Accounting R&D Operation Customer Government Large Enterprise Consumer CEO Firms are Shifting Toward a Customer -Centric Structure “So rather than relying on a structure focused on the company’s discrete product lines , Intel's reorganization will bring together engineers, software writers, and marketers into five market -focused units : corporate computing, the digital home, mobile computing, health care, and channel products — PCs for small manufacturers. ” (BusinessWeek ). (Lee, Sridhar, Palmatier, and Henderson 2015) 16 Agenda Framework for Managing Customer Heterogeneity • Inputs to Managing Customer Heterogeneity Framework • Outputs of Managing Customer Heterogeneity Framework • Process for Managing Customer Heterogeneity Approaches for Managing Customer Heterogeneity • Evolution of Approaches for Managing Customer Heterogeneity • Segmenting, Targeting, and Positioning (STP) Approach • Customer -Centric Approach 17 Evolution of Approaches for Managing Customer Heterogeneity Mass marketing era , which utilizes mass media to appeal to an entire market with a single message Niche marketing era , which concentrates all marketing efforts on a small but specific and well - defined segment of the population One -to -one marketing era , which advocates tailoring of one or more aspects of the firm’s marketing mix to the individual customer Target Market Size Large Small Potential customers Niche segment Niche segment Niche segment Niche segment Media Phone/direct mail Printing and Manufacturing Modular manufacturing/digital printing Communication Internet/mobile Many cable channels (narrowcasting) Large batch manufacturing Few national channels 18 Evolution of Approaches for Managing Customer Heterogeneity • Mass marketing era used mass media to appeal to an entire market with a single message, is a marketing strategy in which a firm mostly ignores customer heterogeneity, with the assumption that reaching the largest audience possible will lead to the largest sales revenue • Niche marketing era focused marketing efforts on well -defined, narrow segments of consumers, and by specializing, this method seeks to give the firm a competitive advantage • One -to -one marketing era is marked by a shift towards one -to -one marketing, such that firms attempt to apply marketing strategies directly to specific consumers • Across all three eras, the underlying method for dealing with customer heterogeneity is the same: focus on smaller and smaller groups of customers, such that the needs of each group are more similar as they get subdivided into smaller units, until the focus reaches an individual customer 19 Segmenting, Targeting, and Positioning (STP) Approach In order to better match heterogeneous customer needs, firms focus their efforts on small “homogenous” customer groups • Segmenting : Dividing market into groups of similar customers (slice the pie into pieces) • Targeting : Selecting best customer group (picking the slice to eat) • Positioning : Improve your relative advantage in the minds of your targeted customers (also addresses Marketing Principle 3 by building SCA) 20 Segmenting Dividing the market into groups of customers where: • Customers within group have very similar needs • Customers across groups have different needs Needs : Needs and benefits desired by customers for your offering • Segment on needs/benefits not descriptors • Uses one of the “Cs” as input: customers Descriptors : Observable customer characteristics that help you find and classify customers (e.g., gender, age, income, size, education, etc.) 21 Segmenting Steps 1. =dentify and refine potential customers’ needs and descriptors (qualitative research) 2. Collect data from random assortment of potential customers on importance of needs to purchase decision 3. Use needs to segment the market into homogeneous customer groups 4. Name segments (communication tool) 22 Cluster Analysis Cluster analysis is a data -driven partitioning technique that can be used to identify and classify a large set of heterogeneous consumers or companies into a small number of homogeneous segments. DAT 2.2 Description • To demystify customer heterogeneity by under -standing preference commonalities across subsets of customers. • To discover how consumers naturally differ and cater to the unique needs of chosen target customer segments. When to Use It How it Works 23 Cluster analysis usually consists of two steps: segmenting and describing. To perform these two steps, we need to collect two kinds of variables: bases and descriptors. Bases, such as desired product features or pricing requirements, provide the foundations for segmenting consumer s a ccording to their differences. Descriptors, such as demographic and geographic information, serve to pro le and eventually target the derived s egm ent. 1. In the segmentation step , we identify underlying subsamples of customers that are homogeneous in their bases (e.g., ratings on product preferences) and markedly different from other subsamples. For example, customers in one cluster might have very high preferences for qual ity and do not mind paying a high price, but customers in another cluster may be very value conscious and refuse to pay high prices. 1. In the describing step , we use descriptor variables to explain how the subsamples differ and thereby can derive efficient targeting strategies, tai lor ed to each subsample. For example, customers in the quality cluster might be mostly men in their early forties, whereas those in th e p rice cluster are mostly women in their early twenties. Using both bases and descriptor variables, we can discover how customers differ, which custome rs to target, and what marketing program to use. R, STATA, SAS, and SPSS software packages are tools that can help conduct the segmenting step; and K -means and hierarchical clus tering are approaches to enable cluster analyses. After the cluster analysis is done, a review of the segmentation results should determine whether the derived clusters make i ntu itive sense. Evaluations of the validity of the segmentation results and corresponding targeting strategy should consider the following important criteri a: • Identifiability : Do the derived segments represent real segments of customers, and can they be labeled using descriptors? • Stability : Are the derived segments likely to change rapidly over time? • Responsiveness : Will each targeted segment respond to the planned marketing strategies? • Viability : Can the company achieve its desired financial objectives with the segmentation scheme? Cluster Analysis (cont.) DAT 2.2 Description Example 1 24 Imagine there are five customers, rated on their intention to purchase (1 –15 scale). A hierarchical clustering procedure, based on Ward’s minimum variance criteria to minimize the sum of the square of errors, starts by assuming each customer is its own cluster. However, combining customers 3 and 4 seems intuitive since they have similar purchase intentions and it results in limited loss of information (0.5 on the dendogram). Similarly, combining customers 1 and 2 results in limited loss of information (4.5). Thus, five customers could be combined into three segments (1,2), (3,4), and (5). If we then try to combine (3,4) and (5) as one customer, the loss of information (25.8) is prohibitive. Thus, we stop at three segments (1,2), (3,4), and (5) Cluster Analysis (cont.) DAT 2.2 Description Example 2 A company conducted an annual customer satisfaction survey for an advertised product, collecting perceptions of the product’s price, quality, and distribution (on a 5 -point scale). To improve customer satisfaction and design more efficient targeting strategies, the company conducted a partition -based clustering analysis of the data and thereby identified three segments: consumers who are dissatisfied on all three attributes (Segment 1), consumers who are highly satisfied on all three attributes (Segment 2), and consumers who are highly satisfied on quality and distribution but dissatisfied on price (Segment 3). The table gives the mean statistics for each segment. A 98.80 Hierarchical Clustering Procedure Gives Dendogram B C D E 25.18 5.00 0.50 • Numbers represent amount of variance explained (don ’t worry about actual numbers but rather change) • Pick number of clusters where variance explained is relatively large • 3 clusters look best since 4 clusters only gives “5” more units versus 99 and 25 • But, look at 4 clusters to see what is different (2 clusters) (3 clusters) (4 clusters) 26 15 min break 27 Targeting • A market needs to select segments to target based on certain selection criteria • Market attractiveness (size, growth rate, price sensitivity, etc.) • Competitive strength (captures the relative strength of a firm, versus competitors, at securing and maintaining market share in a given segment ) • Uses all three “Cs” as input: customer, company, and competitors • An ideal target segment should meet six criteria: 1. Based on customer needs (customer care) 2. Different than other segments (little crossover competition) 3. Differences match firm’s competences (firm can execute within resource constraints) 4. Sustainable (can keep customers) 5. Customers are identifiable (can find targeted customers) 6. Financially valuable (valuable in the long term) • The GE matrix is one analysis tool designed to helps managers visualize and select target segments 29 GE Matrix: Analysis Tool for Targeting Seasonal Gym Members Urban Athletes Elite Athlete Gym Socialites Fashion Trend Setters Market Attractiveness Firm’s Competitive Strength The size of each “bubble” indicates the size of the market segment. Gym Socialites or Fashion Trend Setters are the “best” segments for this firm as they are in the upper right corner of the matrix and are larger markets. Weak Strong Low High Worst Segments Best Segments 30 Positioning • Process of improving your relative advantage in the minds of your targeted customers • Changing both your actual (e.g., innovation) and perceived offering (e.g., branding, relationships) • Uses all three “Cs” as inputs: C ustomer, C ompany, C ompetitors • Nearly everything you do impacts your positioning • P lace (Channel): Samsung dropping Kmart • P rice: No discounts at Tiffany • P romotion: Tiger Woods at Nike, Starbucks • P roduct: Bose, Apple • Perceptual maps: analysis tool to aid in positioning decisions • Repositioning: process by which a firm shifts its target market 31 Working Man Punk Teens Baby Boomers All American Teenagers Perceptual Map: Analysis Tool for Positioning Edgy Traditional American Eagle Forever 21 H&M L.L. Bean Carhart Contemporary Conservative Abercrombie & Fitch Wet Seal Hot Topic Dickies Abercrombie & Fitch changed its positioning, to move from a traditional and conservative segment to a more contemporary, edgy segment by adjusting several elements of its marketing mix, including products, store designs and locations, price points, and marketing communications. Abercrombie and Fitch 32 Positioning Statement Must Address Three Key Questions 1. Who are the customers? 2. What is the set of needs that the product or service fulfills? 3. Why is this product/service the best option to satisfy your needs (relative to competition or substitute; support for why)? This statement is the roadmap for a plethora of implementation decisions involved in marketing a product (both inside and outside the company) 33 Evaluating a Positioning Statement JC Penney: For [Modern Spenders and Starting -outs in mid -income levels who shop for apparel, accessories, and home furnishings] we offer [private -label, supplier exclusive, and national brands] that [deliver greater value than that of our competitors] because of [our unique combination of quality, selection, fashion, service, price, and shopping experience]. 1. Who are the customers? 2. What is the set of needs that the product or service fulfills? 3. Why is this product/service the best option to satisfy your needs (relative to competition or substitute; support for why)? 34 Working Man Punk Teens Perceptual Map: How to extract dimensions? Edgy Traditional American Eagle Forever 21 H&M L.L. Bean Carhart Contemporary Conservative Abercrombie & Fitch Wet Seal Hot Topic Dickies Abercrombie & Fitch changed its positioning, to move from a traditional and conservative segment to a more contemporary, edgy segment by adjusting several elements of its marketing mix, including products, store designs and locations, price points, and marketing communications. 35 Factor Analysis • Factor analysis is a data reduction technique that can be used to identify a small number of latent factors that explain the major variation in a large number of observed variables • When to use it? • To condense a large pool of potential customer needs, wants, and preferences into a short set of similar characteristics • To reduce high correlation among predictors 36 Q1: The product is reasonably priced Q2: The quality of the product is good Factor Analysis Factor Analysis Factor analysis is a data reduction technique that can be used to identify a small number of latent “factors” that explain the variation in a large number of observed variables. DAT 2.1 Description • To condense a large pool of potential customer needs, wants, and preferences into a short set of similar characteristics. • To reduce high correlation among predictors. When to Use It How it Works We begin with a large number of measured variables (e.g., 30) of customer survey measures. The factor analysis algorithm synthesizes the large number of meas ured variables into smaller sets (e.g., 3 –4) of latent “factors” that capture the essence of the meaning in the larger number of measures. To c hoose the total number of factors to retain, we observe how many factors have an Eigenvalue greater than 1. The strength of the association between a measure variable and it s factor is called the “factor loading.” When a measured variable has a factor loading greater than 0.3, it is generally associated with a factor. We categorize the measured variable with a factor where it has the highest loading (e.g., if a measured variable has factor loadings of 0.01 and 0.8 with Factors 1 and 2, we would associate the measured variable with Factor 2). Finally, we interpret what each latent factor represents, based on conceptual commonality underlying the measured variables’ loading on the factor. Example The manager of an online website collected customer satisfaction data from a survey of 1,000 customers on eight aspects of th e c ompany’s focal product. The table shows the factor loadings of a few variables after conducting a factor analysis with three factors. Factor 1 is highly ass ociated with product diversity, specialty, and price; thus, it can be interpreted as the “product” factor. Factor 2 is associated with cash back and discounts, and is thus labeled the “promotion” factor. For Factor 3, the “service” factor, delivery service and customer service have the highest factor loadings. The facto rs can be used as data input for segmentation analyses. The figure shows the focal attributes associated with each factor. 38 © Palmatier, 2017, Marketing Strategy, Palgrave. ISBN: 9781137526236. Types Factor Analysis • Exploratory (EFA): Look for structure in the data (“let the data speak”) • Confirmatory (CFA): Confirm theoretically derived structure Interpretation of Factor Analysis • Factor 1: Quality related questions • Factor 2: Price -related questions Factor Loadings Factor 1 Factor 2 Q1 0.873 0.165 Q2 0.841 0.084 Q3 0.034 0.876 Q4 0.842 0.155 Q5 0.248 0.795 ? ? ? ? ? ? Arnold, M. J., & Reynolds, K. E. (2003). Hedonic shopping motivations. Journal of retailing ,79 (2), 77 -95. Factor Analysis • As we extract consecutive factors, they account for less and less variability • Look at EV > 1 or #factors above elbow point 0 0.5 1 1.5 2 2.5 3 1 2 3 4 5 EV Factor Analysis: Positioning maps in assignment 1 Perceptual Map: Other empirical approaches? Many recent marketing studies use user generated content (UGC ) to derive perceptual maps and brand positioning • Example: Derive market structure from online discussions ( Netzer et al. 2012) . • Idea: Car models and brands that co -occur in a posting share some semantic meaning. • Data collection is cheaper & changes can be picked up faster than in surveys Netzer et al. (2012) Perceptual Map: Other empirical approaches? Netzer et al. (2012) Perceptual Map: Other empirical approaches? Netzer et al. (2012) Perceptual Map: Other empirical approaches? Questions