Prediction and profitability in market segmentation typing tools

Abstract

A vital component in strategic segmentation is the typing tool. Little is known about their prediction performance. Even less is known how well they perform at the segment-level, in imbalanced situations, and how well they predict the smallest (minority) segment. We investigate using simulated and real-life data, how well typing tools perform overall and at the specific segment-level and we show the following. One, even when overall prediction accuracy is good, specific segments may be predicted poorly. Two, for valuable (minority) segments with high targeting costs misclassification can have a substantial impact on the profitability of the segmentation strategy. Poor prediction of a minority segment can happen in high and mildly imbalanced segments. Three, prediction of minority segments can vary substantially across different base classifiers and across imbalance correction methods. We find that performance can vary substantially across base classifiers and that support vector machines, overall, perform best. Four, the prediction of a (minority) segment can always be improved by using imbalance correction methods, and overall random under-sampling performs best.

Appendices

Appendix 1

Some selected market segmentation applications: Number of segments and imbalance

No.	Studies	Number of segments	Segment imbalance
1	Otoo, F.E., Kim, S. & Park, J. (2020). Motivation-based segmentation of Chinese travelers: The role of preferences, sociodemograhics and travel-related features. Journal of Vacation Marketing, 26, 4, 457–472	5	Balanced
2	Delley, M., & Brunner, T. A. (2020). A segmentation of Swiss fluid milk consumers and suggestions for target product concepts. Journal of Dairy Science, 103, 4, 3095–3106	3	1:3
3	Chakrabarti, A., Campbell, B. L., & Shonkwiler, V. (2019). Eliciting Consumer Preference and Willingness to Pay for Mushrooms: A Latent Class Approach. Journal of Food Distribution Research, 50, 1, 46–62	3	Balanced
4	Albayrak, T., Dursun, A. & Űnal, C. (2019). Do tourists have different motivations for online tyravel purchasing? A segmentation of the Russian market. Journal of Vacation Marketing, 25, 4, 432–443	3	1:3
5	Janardhanan, S. & Muthalagu, R. (2020). Market segmentation for profit maximization using Machine Learning. Journal of Physics: Conference Series, 1706	4	1:3
6	Jürkenbeck, K., Spiller, A. & Meyerding, G.H. (2019). Tomato attributes and consumer preferences – a consumer segmentation approach. British Food Journal, 122, 1, 328–344	6	1:2
7	Scheuffelen, S., Kemper, J. & Brettel, M. (2019). How do human attitudes and values predict online marketing responsiveness: Comparing consumer segmentation bases toward brand purchase and marketing response. Journal of Advertising Research, 142–157	6	Balanced
8	Elhai, J. D., & Contractor, A. A. (2018). Examining latent classes of smartphone users: Relations with psychopathology and problematic smartphone use. Computers in Human Behavior, 82, 159–166	2	Balanced
9	Liu, H.-Z., Zheng, Y., Rao, L.-L., Wang, F., Sun, Y., Huang, G.-H., Li, S., & Liang, Z.-Y. (2018). Not all gamblers are created equal: gambling preferences depend on individual personality traits. Journal of Risk Research, 21, 7, 885–898	3	1:2
10	Erdem, S. (2018). Who do UK consumers trust for information about nanotechnology? Food Policy, 77, 133–142	2	1:2
11	Griva, A., Bardaki, C., Pranmatari, K., Papakiriakopoulos, D. (2018). Retail business analytics: Customer visit segmentation using market basket data. Expert systems with Applications, 100, 1–16	10	1:3
12	Gengler, C.E., Mulvey, M.S. (2017). Planning pre-launch positioning: Segmentation via willingness-to-pay and means-end brand differentiators. Journal of Brand Management, 24, 230–249	4	1:6
13	Nadeem, W., Juntunen, M., & Juntunen, J. (2017). Consumer segments in social commerce: A latent class approach. Journal of Consumer Behaviour, 16, 3, 279–292	3	1:10
14	Pomarici, E., Lerro, M., Chrysochou, P., Vecchio, R., & Krystallis, A. (2017). One size does (obviously not) fit all: Using product attributes for wine market segmentation. Wine Economics and Policy, 6(2), 98–106	4	1:2
15	Bruwer, J., & Li, E. (2017). Domain-specific market segmentation using a latent-class mixture modelling approach and wine-related lifestyle (WRL) algorithm. European Journal of Marketing, 51, 9–10, 1552–1576	5	1:2
16	Palma, M. A., Ness, M. L., & Anderson, D. P. (2017). Fashionable food: a latent class analysis of social status in food purchases. Applied Economics, 49, 3, 238–250	4	1:6
17	Kolhede, E., & Gomez-Arias, J. T. (2017). Distinctions Between Frequent Performing Arts Consumers: Implications for Segmentation and Positioning. International Journal of Arts Management, 20, 1, 31–53	3	Balanced
18	Birkenmaier, J., & Fu, Q. (2016). Who Uses Alternative Financial Services? A Latent Class Analysis of Consumer Financial Knowledge and Behavior. Journal of Social Service Research, 42, 3, 412–424	5	1:4
19	Chan, C.C.H., Hwang, Y-R., Wu, H-C. (2016). Marketing segmentation using the particle swarm optimization algorithm. Journal of Ambient Intelligent Human Computing, 7, 855–863	4	1:4
20	Verain, M.C.D., Sijtsema, S.J., Antonides, G. (2016). Consumer segmentation based on food category attribute importance: The relationship with healthiness and sustainability perceptions. Food Quality and Preference, 48, 99–106	3	1:3
21	Adasme-Berios, C., Sanches, M., Mora, M. Schnettler, B.M, Lobos, G. & Diaz, J. (2016). Segmentation of consumer preferences for food safety label on vegetables: Consumer profiles in central and south-central Chile. British Food Journal, 118, 10, 2550–2566	3	1:3
22	Weijters, B., & Goedertier, F. (2016). Understanding today’s music acquisition mix: a latent class analysis of consumers’ combined use of music platforms. Marketing Letters, 27, 3, 603	4	1:4
23	Díaz, E., Martín-Consuegra, D., & Estelami, H. (2016). A persuasive-based latent class segmentation analysis of luxury brand websites. Electronic Commerce Research, 16, 3, 401	3	1:3
24	Segovia, M. S., & Palma, M. A. (2016). Buying your way into a healthier lifestyle: a latent class analysis of healthy food purchases. Applied Economics, 48, 21, 1965–1977	2	Balanced
25	Hosseini, M. & Shabani, M. (2015).New approach to customer segmentation based on changes in customer value. Journal of Marketing Analytics, 3, 3, 110–121	11	1:7
26	Bassi, F. (2015). Forecasting financial products acquisition via dynamic segmentation: an application to the Italian market. International Journal of Market Research, 57, 6, 909	5	1:20
27	Rundle-Thiele, S., Kubacki, K., Tcazzynski, A., Parkinson, J. (2014). Using two-step cluster analysis to identify homogeneous physical activity groups. Marketing Intelligence and Planning, 33, 4, 522–537	4	1:3
28	Alexandra, D., Miragaia, M., Maryins, M.A.B. (2015). Mix between satisfaction and attributes destination choice: A segmentation criterion to understand the ski resorts consumers. International journal of tourism research, 17, 313–324	6	1:3
29	Realini, C.E., Kallas, Z., Pérez-Juan, M., Gómez, J., Olleta, J.L., Beriain, M.J., Alberti, P., & Sañuda, C. (2014). Relative importance of cues underlying Spanish consumers’ beef choice and segmentation, and consumer liking of beef enriched with n-3 and CLA fatty acids. Food Quality & Preference, 33, 74–85	3	1:2
30	Campbell, C., Ferraro, C., Sands, S. (2014). Segmenting consumer reactions to social network marketing. European Journal of Marketing, 48, 3–4, 432–452	5	1:3
31	Liu, H.B., McCarthy, B., Chen, T. (2013). The Chinese wine market: a market segmentation study. Asian Pacific Journal of Marketing & Logistics, 26, 3, 450–471	3	1:10
32	Simunaniemi, A. ‐M., Nydahl, M., & Andersson, A. (2013). Cluster analysis of fruit and vegetable-related perceptions: An alternative approach of consumer segmentation. Journal of Human Nutrition and Dietetics, 26, 1, 38–47	2	1:2
33	Riefler, P., Diamantopoulos, A., & Siguaw, J. A. (2012). Cosmopolitan consumers as a target group for segmentation. Journal of International Business Studies, 143, 285–305	4	1:3
34	Liu, Y., Kiang, M. & Brusco, M. (2012). A unified framework for market segmentation and its applications. Expert Systems with Applications, 39, 10292–10302	3	1:3
35	Davis, S. (2012). Choosing the right baskets for your eggs: deriving actionable customer segments using supervised genetic algorithms. International Journal of Market Research, 54, 5, 689	4	Balanced
36	Ko, E., Taylor, C.A., Sung, H., Lee, J., Wagner, U., Navarro, Martin-Consuegra, D., & Wang, F. (2012). Global market segmentation usefulness in the sportwear industry, 65, 1565–1575	4	1:2
37	Barber, N., & Taylor, C. (2011). Equity benefits of smaller wine regions and life- style segmentation. Journal of Brand Management, 19, 2, 158–175	8	1:3
38	Kim, T. & Lee, H-Y. (2011). External validity of market segmentation methods: A study of buyers of prestige cosmetic brands. European Journal of Marketing, 45, 1–2, 153–169	3	1:2
39	Ponnam, A., Sahoo, D., & Balaij, M.S. (2011). Satisfaction-based segmentation: Application of the Kano model in Indian fast-food industry. Journal of Targeting, Measurement, and Analysis for Marketing, 19, 3/4, 195–205	4	1:2
40	Cui, G., & Wang, Y. (2010). Consumers’ SKU choices in an online supermarket: a latent class approach. Journal of Marketing Management, 26, 5–6, 495–514	3	NA
41	Liu, Y., Ram, S., Lusch, R.F. & Brusco, M. (2010). Multicriterion market segmentation: A new model, implementation, and evaluation. Marketing Science, 29, 5, 880–894	4	1:10
42	Gil-Saura, I., Ruiz-Molina, M-E. (2009). Customer segmentation based on commitment and ICT use. Industrial Management & Data, 109, 2, 206–223	4	1:3
43	Gabay, G., Paulus, K., Aarts, P., Beckley, J., Ashman, H., & Moskowitz, H. R. (2009). Cross-National Segmentation for Messaging about Cheese: Towards a New Approach to Consumer Understanding, Directed Development, and Targeted Marketing. Italian Journal of Food Science, 21, 4, 407–428	3	1:2
44	Shiftan, Y., Outwater, M/L., Zhou, Y. (2008). Transit market research using structural equation modeling and attitudinal segmentation. Transport Policy, 15, 186–195	8	1:17
45	Varela Mallou, J., Rial Boubeta, A., Braña Tobío, T., & Voces López, C. (2008). Application of latent class analysis to the investigation of customer loyalty in service companies. Methodology: European Journal of Research Methods for the Behavioral and Social Sciences, 4, 3, 87–96	4	1:25
46	Bruning, E.R., Hu, M.Y., Hao, W. (2007). Cross-national segmentation: An application in the NAFTA airline passenger market. European Journal of Marketing, 43, 11/12, 1498–1522	5	1:2
47	Bassi, F. (2007). Latent class factor models for market segmentation: an application to pharmaceuticals. Statistical Methods and Applications, 16, 279–287	3	1:4
48	Brey, E.T., So, S-I., Kim, D-Y., Morrison, A.M. (2007). Web-based permission marketing: Segmentation for the lodging industry. Tourism Management, 28, 1408–1416	3	1:3
49	Sajeev Varki, & Pradeep K. Chintagunta. (2004). The Augmented Latent Class Model: Incorporating Additional Heterogeneity in the Latent Class Model for Panel Data. Journal of Marketing Research, 41(2), 226	5	1:7
50	Sewitch, M.J., KLeffondre, K., Dobkin, P.L. (2004). Clustering patients according to healthcare perceptions: relationships to psychosocial characteristics and medication non-adherence. Journal of Psychosomatic Research, 56, 323–332	5	Balanced
51	Bhatnagar, A. & Ghose, S. (2004). A latent class segmentation analysis of e-shoppers. Journal of Business Research, 57, 758–767	3	1:9
52	Ter Hofstede, F., Steenkamp. J-B. E.M., & Wedel, M. (1999). International market segmentation based on consumer-product relations. Journal of Marketing Research, XXXVI, 1–17	4	1:7
53	Steenkamp, J-B. & Wedel, M. (1993). Fuzzy clusterwise regression in benefit segmentation: Application and investigation into its validity. Journal of Business Research, 26, 3, 237–249	3	1:2
54	Morwitz, V.G. & Schmittlein, D. (1992). Using segmentation to improve sales forecasts based on purchase intent: Which “intenders” actually buy. Journal of Marketing Research, XXIX, 391–405	5	1:9
55	Carroll, N.V., Gagon, J.P. (1983). Identifying consumer segments in health services markets: An application of conjoint and cluster analyses to the ambulatory care pharmacy market. Journal of Health Care Marketing, 3, 3, 22–34	5	1:10

Appendix 2

The SMOTE algorithm:

• Provide the algorithm with the number of minority segment samples, T, the number of synthetic minority segment samples to generate, N, and the number of nearest neighbors to find for each sample, k.

  1. For all T minority segment samples, compute the k nearest neighbors (or nearest samples) based on Euclidean distance.

  2. While N is greater than 0:

  2.a Select a sample i from the T minority segment samples. (If run out of samples from T we just loop over it again).

  2.b Randomly select a nearest neigbor for sample i, which we refer to as nn.

  2.c For all possible attributes:

  2.c. (1) Compute the difference between sample nn’s and sample i’s attribute value, store this value as diff.

  2.c (2) Randomly select a number between 0 and 1, store this value as gap.

  2.c (3) Set the synthetic datapoint attribute value to.

  sample[i][attribute] + gap*diff.

  2.d Store the new synthetic datapoint.

  2.e N = N − 1.