Social networks and their economics influencing consumer choicedr soc

Social Networks and their Economics Influencing Consumer Choice Daniel Birke Social Networks and their Economics Social Networks and their Economics Influencing Consumer Choice Daniel Birke Visiting Researcher, Aston Business School, Birmingham, and works in a leading international management consultancy in Germany This edition first published 2013 C 2013 John Wiley & Sons, Ltd Registered office John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, United Kingdom For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com The right of the author to be identified as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988 All rights reserved No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher Wiley also publishes its books in a variety of electronic formats Some content that appears in print may not be available in electronic books Designations used by companies to distinguish their products are often claimed as trademarks All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners The publisher is not associated with any product or vendor mentioned in this book Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose It is sold on the understanding that the publisher is not engaged in rendering professional services and neither the publisher nor the author shall be liable for damages arising herefrom If professional advice or other expert assistance is required, the services of a competent professional should be sought Library of Congress Cataloging-in-Publication Data Birke, Daniel Social networks and their economics : influencing consumer choice / Dr Daniel Birke pages cm Includes bibliographical references and index ISBN 978-1-118-45765-8 (cloth) Social networks–Economic aspects Consumer behavior I Title HM741.B57 2013 658.8 34–dc23 2013017209 A catalogue record for this book is available from the British Library ISBN: 978-1-118-45765-8 Set in 10/12pt Times by Aptara Inc., New Delhi, India 2013 Contents List of figures ix List of tables xi Preface xv Acknowledgements Consumer choice in social networks 1.1 Motivation 1.2 Using mobile telecommunications to illustrate the economics of social networks 1.3 Structure of the book 1.4 Why you should read this book References Research into social networks in economics, sociology and physics 2.1 Introduction 2.2 The economics of networks: Key findings from economics and marketing 2.2.1 Introduction 2.2.2 Definition of network effects 2.2.3 Direct network effects 2.2.4 Indirect network effects 2.2.5 Implications for company strategies 2.3 Social network analysis: Key findings from sociology 2.3.1 A short history 2.3.2 Network analysis basics 2.3.3 Design of social network studies 2.4 Key findings from physics research into complex networks 2.5 Empirical research on social networks and network effects 2.5.1 Introduction 2.5.2 Big data: Massive electronic social networks xvii 1 11 12 13 13 14 15 17 18 24 24 27 29 30 32 32 32 vi CONTENTS 2.5.3 Challenges when identifying causal relationships in social networks 2.5.4 Empirical strategies to identifying causal effects in social networks 2.5.5 Empirical challenges and advances in the economics of network literature 2.6 Summary References 33 35 38 40 40 Marketing in social networks: The iPhone 3.1 Executive summary 3.2 Case study 1: Social network and viral marketing 3.3 Case study 2: Social advertising on Facebook 3.4 Introduction to the empirical study 3.5 Product diffusion in social networks 3.6 Modelling diffusion in social networks 3.7 Model estimation 3.7.1 Description of the data used: Very large-scale mobile network 3.7.2 Description of the statistical method used: Survival analysis 3.8 Model results 3.8.1 Non-parametric tests 3.8.2 Variable definitions 3.8.3 Model results: Impact of the social network on iPhone adoption 3.8.4 iPhone virality over time 3.9 Discussion References 47 47 48 52 54 55 57 59 Switching and churn in social networks 4.1 Executive summary 4.2 Case study: Customer retention in social networks 4.3 Introduction to the empirical study 4.4 Key findings from the switching cost literature 4.5 Modelling concept 4.6 Description of the data used: Another large-scale mobile network 4.7 Model results 4.7.1 Non-parametric tests 4.7.2 Variable definitions 4.7.3 Model results: Impact of the social network on customer churn 4.7.4 Robustness tests 4.8 Discussion References 71 71 72 75 76 78 59 60 62 62 63 64 65 67 68 79 81 81 81 83 85 86 88 CONTENTS vii How social networks influence consumer choice of mobile phone carriers in the UK, Europe and Asia 5.1 Executive summary 5.2 Case study: Using homophily for social network marketing 5.2.1 Mobile phone carriers 5.2.2 Online retailers 5.2.3 Online social networks 5.3 Introduction to the empirical study 5.4 Methodology 5.4.1 Design of the social network survey 5.4.2 Description of the statistical approach used: Quadratic assignment procedure 5.5 Understanding the properties of the social networks 5.5.1 Descriptive social network statistics 5.5.2 Graphical analysis of a social network 5.6 The impact of friendship on operator choice 5.7 Robustness of results 5.7.1 Non-respondents 5.7.2 QAP and multicollinearity 5.7.3 Ethnicity 5.8 Are stronger relationships more influential? 5.9 Friendship networks and choice of handset brand 5.10 Multi-country case study of operator choice in social networks 5.10.1 Malaysia 5.10.2 The Netherlands 5.10.3 Italy 5.10.4 Cross-country comparison 5.11 Discussion References 100 102 102 106 108 112 112 114 116 117 120 122 123 124 127 132 133 134 Coordination of mobile operator choice within households 6.1 Executive summary 6.2 Case study: Social network marketing to communities 6.2.1 International communities 6.2.2 Families 6.3 Introduction to the empirical study 6.4 Data 6.5 Descriptive statistics 6.6 The model 6.7 Multinomial logit model 6.7.1 Model parameters 6.7.2 Base model 6.7.3 Relationship types within households 6.8 Multinomial probit model 6.8.1 Independence of irrelevant alternatives 137 138 138 139 140 142 143 144 146 148 148 149 152 153 153 91 92 92 93 95 95 96 98 98 viii CONTENTS 6.9 6.8.2 Multinomial probit motivation 6.8.3 Estimation results Discussion References How pricing strategy influences consumer behaviour in social networks 7.1 Executive summary 7.2 Case study: Pricing digital products with network effects 7.2.1 Facebook 7.2.2 LinkedIn 7.3 Introduction to the empirical study 7.4 The mobile telecommunications industry in the UK 7.5 The model: Price discrimination between on- and off-net calls 7.6 Estimation results: Adapting consumption choice to price signals 7.7 Discussion References 155 157 158 158 161 161 162 164 164 165 167 169 173 175 176 Conclusions 8.1 Main results 8.2 Implications of interdependent consumer choice 8.2.1 For marketing practitioners 8.2.2 For academic researchers 8.2.3 For regulatory policy 8.3 Looking ahead: How social network analysis is changing research and marketing practice References 177 177 178 178 179 180 Appendix A Success factors for viral marketing campaigns A.1 Proposition excellence A.2 Observability of the product or its use A.3 Design the campaign around a good understanding of the specific role of word-of-mouth in propagating your product A.4 Word-of-mouth for economic benefit A.5 Exploit storytelling and tap into the zeitgeist A.6 Exploit influential expert users A.7 Appeal to communities of interest A.8 Conclusion References 183 185 186 Appendix B 193 Index Student questionnaire 180 181 187 187 188 189 189 190 191 197 188 APPENDIX A much discussed and recommended (or indeed advised against) by their users and the information is frequently acted upon by their friends However, post-purchase rationalisation effects mean that your new customers are particularly likely to speak positively about your product if they have made a large emotional or financial investment in the purchase Clearly this does not apply to purchases like cinema viewings and book purchases Examples: One restaurant in Ireland gives a voucher for a 10% discount on future purchases with the bill Member-get-member campaigns aimed at new customers are seen in a variety of industries ranging from magazine subscriptions and gyms to mobile phone contracts Recommended action: Understand the pre-purchase information needs of your potential customers Compare this with the information that your company (and your competitors) provide Research the content of new customers’ discussions about your product (from social media or market research) Identify positive facts about your product that can be communicated to the customer during and immediately after the purchase, to enhance the positive word-of-mouth effects Choose short, simple, powerful facts and provide evidence Also identify any untrue or misleading information about your product within new customers’ conversations and strive to correct these by improving your product communications Finally, work to enhance the positive effects of post-purchase rationalisation by providing extra reasons for new customers to feel positive towards you or to share their experience with their friends Special deals for new customers and schemes for new customers to introduce a friend are examples A.5 Exploit storytelling and tap into the zeitgeist Dunbar (2010) proposes that a central purpose of chat, or gossip, is to build and maintain group relationships In other words, rather than chatting to exchange information, we often exchange information in order to chat This need of humans to use chat to build and maintain social structures is responsible for many of the word-of-mouth conversations that take place – as listening in to strangers’ conversations at a bus stop will attest Therefore, if one can give people a good reason to include your product in their conversations, then strong word-ofmouth transmission can result Clearly, the right reason will vary by market segments If one can connect one’s product with matters of current importance (known as the zeitgeist) then the likelihood of your product making it into those conversations is much increased Examples: The entertainment industry has understood this for a long time – when a famous (or formerly famous) singer checks into rehab or announces that he is engaged, this may be driven by the manager’s desire for media coverage and the water cooler/bus stop conversations that it engenders, rather than a desire by the artist for a new start in life Recommended actions: Brainstorm interesting stories about your product – particularly those connected with the zeitgeist Provide them through your sales and care APPENDIX A 189 channels, and through PR/ social medial if appropriate Avoid blanket distribution – a titbit of information is of no value if the recipient has heard it before Monitor social media to determine which stories are working and which are not Learn, adapt, and keep trying A.6 Exploit influential expert users Many products have a small group of expert users Most mechanical or electronic products will have expert users who are well respected in their community for their skill and knowledge Providing extra technical information to these people can turn them into advocates for your products as well as providing valuable expert feedback on product functionality These insiders will share the information with others – thereby enhancing their own social standing and spreading the word about your products Typically, these influencers cannot be counted on to be evangelists – their value lies more in their knowledge than their loyalty However, it is important to understand that these expert influencers will be there whether the company supports them or not Example: Rosen (2009) describes Microsoft’s ‘Most Valuable Professional’ programme, an excellent example of such a strategy The MVP programme grew from recognition of the role of independent experts in solving and raising technical issues with Microsoft products via CompuServe forums in 1993 There are now over 3500 MVPs participating in Microsoft’s programme to nurture these expert influencers Recommended action: Identify your expert users and develop a programme to provide them with expert support Your support/technical teams may be sick of hearing from these users who take up so much of their time – work with them to identify the influencers in the user community, and build an ambassador programme around them A.7 Appeal to communities of interest Word-of-mouth transmission is greater within specific groups than within the population as a whole By making your product appeal to golfers, beekeepers, new mums or Esperanto speakers you will achieve greater word-of-mouth transmission within those groups In particular, ethnic minorities offer great potential for word-of-mouth campaigns, because members of these groups often lack access to other information sources (e.g social networks containing wise and experienced locals, or media in the country’s home language) enjoyed by the rest of the population However, it is important that the offer is relevant to the community An offer of low-price air tickets to Nigeria will have significant social value within the Nigerian community in London, whereas low-price air tickets to France or the USA will of course be far less relevant 190 APPENDIX A One issue with communities of interest is that the tightness of the group tends to vary inversely to the size of the group – so, depending on the economics of your marketing, the more strongly connected groups may be too small to be worth targeting However, targeting small groups can be particularly appropriate for start-up companies or for innovative products Example: A German telecoms company ran a highly successful campaign aimed at ethnic minorities to promote a low-cost international call tariff An outbound calling campaign in the immigrant’s home language ensured the message was well seeded – as well as delivering a strong response within the target group, uptake outside the target group but within the ethnic groups was substantial due to the viral effect Recommended action: Understand which subgroups of society your product appeals to, or can be targeted at Remember that the product must be relevant to the needs of the group Invest time and money in thoroughly understanding the needs of the group and how your product can be applied to them Build relationships with the group Deliver the campaign and monitor the word-of-mouth conversations via social media, market research and (if appropriate) on-going contact with group members A.8 Conclusion Above are some of the key success factors for word-of-mouth marketing success However, word-of-mouth is at the whim of the consumer, and experienced word-ofmouth marketers know that there is no sure-fire formula for word-of-mouth marketing Table A.1 Is your product / campaign suitable for word-of-mouth marketing? Fully Does the proposition deliver on its promise at every level (purchase, set-up, use, payment, brand, etc.)? Is the product or its use highly observable? Is the specific role of word-of-mouth in propagating your product understood? Does the campaign exploit this? Is there an economic benefit to word-of-mouth for your customers? Can you exploit it? Can you create stories about your product or campaign that fit with the zeitgeist? Can influential expert users be identified and nurtured? Will the product be made to appeal to communities of interest? Overall, how much word-of-mouth activity you expect this campaign or product to generate? Partly Not APPENDIX A 191 campaigns success It is nevertheless true that by optimising the key factors, the level of word-of-mouth will be increased The checklist in Table A.1 may assist the marketer in assessing the likely success of word-of-mouth marketing References Dunbar, R (2010) How many friends does one person need?, Faber & Faber, London Financial Services Club blog (2010) Exclusive: Interview with First Direct, 28th Sept 2010, http://thefinanser.co.uk/fsclub/2010/09/exclusive-interview-with-first-direct.html (accessed 02 February2013) Iyengar, R., Van den Bulte, C and Choi, J (2011) Distinguishing among multiple mechanisms of social contagion: social learning versus normative legitimation in new product adoption MSI Report No 11-119, Marketing Science Institute, Cambridge, MA Rosen, E (2009) The Anatomy of Buzz Revisited: Real-Life Lessons in Word-of-Mouth Marketing, Crown Business Appendix B Student questionnaire Nottingham University Business School Questionnaire on Network Effects and Mobile Phones We would be very grateful if you can complete this questionnaire It will give you some useful insights into the practical relevance of network effects, and will also be valuable data for Daniel Birke (a second year PhD student) who is preparing a thesis on network effects Daniel will present an analysis of the data in a subsequent lecture What is your gender? Male Female What is your Nationality? Please specify: What brand is your mobile phone? Motorola Nokia Samsung Siemens Sony-Ericsson Other, please specify: _ Which mobile network operator you currently use? O2 Orange T-Mobile Virgin Vodafone Other, please specify: _ Social Networks and their Economics: Influencing Consumer Choice, First Edition Daniel Birke © 2013 John Wiley & Sons, Ltd Published 2013 by John Wiley & Sons, Ltd 194 APPENDIX B How you pay for your mobile phone calls? Pay-as-you-go (pre-paid) Fixed contract (post-paid) About how much you spend on using your mobile phone per month? Less than £10 £ 10–19 £ 20–29 £ 30–39 £ 40–49 Over £ 50 Don’t know Do you pay for the costs of your mobile yourself? Yes Partially No Please specify who pays: Why did you choose your current operator? Someone else chose the network for me Please specify: I chose it myself, because of: strongly agree agree neither nor Disagree strongly Don’t disagree know Quality of the network (network coverage, roaming possibilities etc.) Special offer Cost of calls, text messages in general It is cheaper, because my friends/family use the same network Cost of handset Handsets available from this operator More services available (games etc.) Good customer service What paid services you use? (Please tick any that apply) Data transmission (Notebook, PDA etc.) Ringtones Multimedia messages APPENDIX B 195 Subscription to information services (sports, stock news etc.) Games Internet surfing None of them 10 Have you changed your network since the beginning of your studies? No => Please go to Question 11 Yes => Please also answer Question 10a) + Question 10b) 10a) If you have changed your operator, which one was your previous operator? O2 Orange T-Mobile Virgin Vodafone Other, please specify: 10b) What was the reason for changing your operator? Not satisfied with old operator Please specify: My friends use another operator Special offer Other, please specify: _ 11 How long you phone the following groups with your mobile per week? Friends Family Partner Other people Less than 20 mins 21 - 40 mins 41 - 60 mins More than 60 mins 12 How often you text the following groups per week? (No of texts) Friends Family Partner Other people 13 Do you know which operator your friends/family/partner use? Know it Know it for some My friends My family members My partner 14 Do you have a land line (fixed line) telephone? Yes => Please go to Question 14a No => Please also answer Question 15 Don’t know it 196 APPENDIX B 14a) If you have a land line, what you mainly use it for? Internet Phone calls Other, please specify: _ Finally, we are interested in how your mobile phone choice is influenced by your friends and colleagues Therefore, we have prepared an alphabetical name list of course participants (see next page): 15 Please mark you own name by putting a ring around your name 16 Please tick the people that you call Please use one tick for people that you phone occasionally and two ticks for people that you call frequently 17 If you think of the people you call up or text to most frequently, how many of them are also taking part in Economics of Organisation B? Please indicate them by writing ‘Top 5’ after the name (additionally to the two ticks) Would you be willing to participate in a second stage of this survey? Yes No If yes, please enter email address here: _@nottingham.ac.uk To thank you for your participation, we will draw three prizes from all the e-mail addresses entered The 1st prize is a £25 voucher for a night out, the 2nd and 3rd prize will be a bottle of wine We will present the winners and the results of the questionnaire at one of the forthcoming lectures Index anti-ulcer drugs, 39f AOL-Time Warner, 23 Facebook, 37, 52f, 93, 95f, 164f, 181 fixed effects model, 109f, 119f, 151 big data, 32f, 180f Hawthorne factory, 25 homophily, 35, 38, 82f, 92ff calling records, 59f, 79f, 93f causality in social networks, 33f, 58, 97, 134 churn, 76f churn influence, 73f churn pressure, 73f company strategies, 18ff, 75, 93, 161ff competition for the market, 19 complex networks, 30ff consumption interdependences, 13, 17f critical mass, 14, 56 customer retention, 50, 72ff diffusion, 14, 21, 31f, 40, 47ff, 54ff, 65ff diffusion speed, 31, 65ff epidemiology, 56 direct network effects, 13, 15ff, 39, 76, 158, 162 domino network effects, 16 economics of networks, 12 empirical literature review, 32ff experiments, 35ff IBM PC, 21 indirect network effects, 13, 17ff, 76, 158, 162 influencers, 27, 34f, 49f, 73f, 189f installed base, 19, 22 learning effects, 17, 39f, 56, 78, 158 legal implications, 23, 180f LinkedIn, 95f, 164f Microsoft, 23 multinomial logit, 148f multinomial probit, 153f Netscape, 23 network graphs, 24f, 27ff, 57, 106ff, 117, 123, 125 network matrices, 28, 99ff, 105f, 155ff network surveys, 29f, 98ff, 193 network value, 74 option-value network effects, 16 Social Networks and their Economics: Influencing Consumer Choice, First Edition Daniel Birke © 2013 John Wiley & Sons, Ltd Published 2013 by John Wiley & Sons, Ltd 198 INDEX path-dependence, 20 pricing strategies, 14ff, 22, 38ff, 78f, 97f, 109, 124, 161ff structural demand and supply model, 39 survival analysis, 60f switching, see churn quadratic Assignment Procedure (QAP), 100f, 114ff QWERTY keyboard, 17 targeted marketing, 37, 92, 181 tariff-mediated network effects, 15, 22, 78, 92, 122, 123 133f technology lock-in, 20, 76 time-series analysis, 173f transitivity, 25f triads, 25, 141 scale-free networks, 31 small-world networks, 26, 31 social advertising, 50, 52f, 164f social network influence, 49f social network marketing, 27, 32ff, 37, 49f, 52f, 72ff, 92ff, 138f, 164f, 183ff social network pressure, 48f sociogram, 24 standardisation, 19f strength of weak ties, 26f, 95 uncertainty, 17 VHS vs beta, 19f viral marketing, 48ff, 53, 65, 67, 74f, 183ff virtual networks, 15 word-of-mouth, 34, 51, 67, 187f Chinese British A Unknown Figure 5.2 Predicting user characteristics in a social network Colours: Vodafone Virgin Orange Other operators Three T-Mobile O2 Shapes: British Other Europeans Other Asians Africans Chinese +The Americas Figure 5.3 UK 2005: Student class social network Social Networks and their Economics: Influencing Consumer Choice, First Edition Daniel Birke © 2013 John Wiley & Sons, Ltd Published 2013 by John Wiley & Sons, Ltd Figure 5.4 UK 2005: Full student class social network1 Reproduced by permission of Elsevier Figure 5.5 UK 2005: Nationality and ethnicity of students Reproduced by permission of Elsevier Please see Figure 5-3 for the legend explaining shapes and colours Colours: Vodafone Virgin Orange Other operator T-Mobile O2 Shapes: British Other Europeans Other Asians Africans Chinese + The Americas Figure 5.6 UK 2006: Student class social network Reproduced by permission of Elsevier Colours: DiGi Celcom Maxis Shapes: Chinese British Malay Indian Figure 5.7 Malaysia: Student class social network Reproduced by permission of Elsevier Colours: KPN Mobile Orange Telfort T-Mobile Vodafone Other Figure 5.8 The Netherlands: Student class social network Reproduced by permission of Elsevier Colours: TIM Vodafone Wind H3G Figure 5.9 Italy: Student class social network Reproduced by permission of Elsevier ... Social Networks and their Economics Social Networks and their Economics Influencing Consumer Choice Daniel Birke Visiting Researcher, Aston Business School, Birmingham, and works in... identifying causal relationships in social networks 2.5.4 Empirical strategies to identifying causal effects in social networks Social Networks and their Economics: Influencing Consumer Choice, First Edition... collect Social Networks and their Economics: Influencing Consumer Choice, First Edition Daniel Birke © 2013 John Wiley & Sons, Ltd Published 2013 by John Wiley & Sons, Ltd 2 SOCIAL NETWORKS AND THEIR