Novel Quantitative Tools to Manage Viral Marketing Campaigns
Summary of Impact
Novel quantitative tools to predict, design, and manage viral marketing campaigns, developed by Prof Ralf van der Lans and his collaborators, have led directly to advances in business practice in creative marketing, benefiting an agency and its clients, ranging from multinationals to intergovernmental organizations. These advances, including a key performance indicator (KPI) not previously available and a predictive “seeding” model bringing new insights on social network relationships and influencers, have led to economic impact through improved monitoring and optimization of such campaigns for the agency implementing his tools. Meanwhile, wider impact on professional practice was achieved, with understanding among stakeholders in the marketing sector spread through the recognition of an overseas industry award in 2014; related trade and business media coverage of the award; and later media and conference interest. A chapter in a marketing handbook, downloaded over 2,300 times since 2017, has also broadened its impact among marketing practitioners.
Underpinning Research
Nowadays, viral marketing is among the fastest-growing forms of advertising, as evidenced by the Nielsen CMO Report (2018) on digital marketing spending. Such marketing, also known as social media campaigns, took off with the advent of Facebook, Twitter,Weibo, and other mass platforms in the first decade of the new millennium. Yet running such campaigns was initially more art than science, with little quantitative insight into how to maximize effectiveness and no analytical tools that could predict the reach of viral marketing campaigns, a crucial factor to determine return on investment for clients of marketing agencies. This was the setting for the research underpinning Prof van der Lans' quantitative tools, which started in 2006 in collaboration with the industry. Studies continued at HKUST Business School, which Prof van der Lans joined in 2010 as associate professor. Overall, the research can be divided into two phases:
(1)Development of a viral branching model to forecast reach and success of campaigns
In the first phase, Prof van der Lans led research that developed what he and his collaborators termed a “viral branching model” to forecast both reach and success of viral marketing campaigns. This was started before joining HKUST, with articles published in Marketing Science (2010) and The Connected Customer (2010).
The model was rooted in epidemiology, a field that explores the spread of actual viruses. However, in contrast to epidemics, where the spreading process is unobserved, Prof van der Lans developed a new branching model that can be applied to real data, collected during viral marketing campaigns. This line of research continued at HKUST [R1, R2]. The model takes into consideration the “viral index”, consisting of two parameters: 1) the average number of friends that an individual shares the campaign with, and 2) the probability that a contacted friend responds favorably to the shared message.
These parameters can be determined before, or estimated quickly after the launch of the campaign, and used to predict the final reach. The parameters are also useful for managing viral marketing campaigns, being determined by the design of the campaign. In addition, as the viral branching model can be implemented through Excel, it is easy to use in practice.
(2) Identifying influential individuals in social networks for “seeding” viral marketing
After joining HKUST, Prof van der Lans embarked on a second phase of research, in collaboration with Xi Chen, a PhD student at HKUST who is currently an assistant professor at Erasmus University Rotterdam. This phase focused on optimal seeding strategies, which are critical to initiate viral processes and campaign optimization [R3], [R4]. Other viral marketing and social network analyses have suggested people with many friends (or hubs) are most influential. Going beyond such traditional “recommendations”, Prof van der Lans and his co-authors found not all relationships to be equally important. They developed a seeding model that could ascertain the importance of each connection in a social network [R3]. The seeding tool was applied to two campaigns involving the promotion of microfinance programs in Indian villages and advertising campaigns on social media. The results showed improvements of up to 10% in the first application and over 90% in the second.
Continued collaboration with the industry partner focused on the information overload in social networks and the consequences for optimal seeding strategies. In an accepted paper at Management Science [R5], Prof van der Lans and his co-authors further show that people with many friends are less responsive as they face stronger competition for attention. This shows that targeting hubs, as traditionally suggested, leads to sub-optimal solutions as it ignores the fact that such individuals are less responsive and, therefore, less influential.
Details of the Impact
Forecasting and optimizing viral marketing campaigns are critical issues for practitioners to determine return on investment (ROI) and whether and how a campaign should be launched. Using the novel tools developed through Prof van der Lans’ research, companies and marketers are now able to 1) forecast the performance of a viral marketing campaign, even before a campaign is launched; and 2) boost optimization of viral marketing strategies (i.e. the content of the campaign and seeding strategies) to improve their services to clients in a wide variety of sectors. Such work has achieved impact through changing professional practice, and led to economic impact for the agency and its clients by optimizing ROI from their campaigns through their implementation of the new analytical tools. Wider awareness among international marketing practitioners has been generated through an industry award, media coverage, and dissemination through a book chapter and presentations at industry-related events and programs.
(1) Changing professional practice to boost business performance
Prof van der Lan's industry partner began to implement his viral marketing analytical tools before publication of the first paper in 2010. Since then, the industry partner has continued to evolve its business operations through on-going use of such tools and key performance indicators (KPIs) to monitor and forecast the performance of its creative campaigns, as well as improve them.
This has resulted in economic and corporate impacts by helping the company deliver greater return on investment for clients through more efficient practices. Such quantitative monitoring and forecasting had not been available to marketers before Prof van der Lans' research. As noted by the industry partner in a testimony: “Before our collaboration, the industry did not have analytical tools to predict the success of our marketing campaigns”.
Indicating the significance of the development of such methods for the company, the industry partner further noted that clients had been increasingly demanding objective metrics to predict and determine the success of campaigns, ahead of its collaboration with Prof van der Lans.
Showing the depth of impact on the agency’s operations that the research-based viral index had, the industry partner explained: “We now use the viral index for the following purposes: 1) objectively determine the performance of a campaign; 2) forecast the reach of a campaign, and 3) insights in how to improve the campaign. We monitor this index for all our viral marketing campaigns, and also report it to our client[s] as a key performance indicator.” The company has widened the reach of the impact across major brands, multinationals, and civil society organizations. Meanwhile, the company's staff have benefited from presentations provided by Prof van der Lans, enabling them to implement his recommendations.
In addition to campaign performance monitoring, the company has altered and improved its business practice through KPIs provided by Prof van der Lans. As the company wrote in its testimony: “Instead of focusing on the final reach of campaigns, we now focus [on] two important KPIs: 1) How active[ly] do people share our campaigns? and 2) How do people respond to campaigns?” These KPIs are based on on-going van der Lans’ research that considers competition for attention, which affects responsiveness, important for seeding strategies. From the KPIs, the reach of the enhanced strategies extend to different sectors of industry and the third sector through reports to its client base.
(2) Widening practitioner understanding of viral marketing internationally
The model and tools developed by Prof van der Lans have achieved impacts of broadening and deepening awareness of viral campaign practice and management through an industry award, coverage in a variety of media and publications, and conference presentations, attracting the attention of marketers and their organizations in different countries.
a) His 2014 article on viral marketing [R2] led to a jury of practitioners and industry leaders from the Dutch Center for Information-Based Marketing and Market Research (MOA) presenting Prof van der Lans and his co-author with the MOA Science Award for that year, with a news item appearing on the MOA website. MOA is an influential association of companies and institutions engaged in marketing research and intelligence that promotes, develops and stimulates the quality of market research. A further report on the awards by Adformatie, a Dutch multichannel platform – including a magazine and website that target all professionals dealing with brands and organizational behavior – noted that the research had been cited by the jury as “an innovative, very readable and practically applicable overview of recent insights in the field of viral marketing, in which the – methodologically advanced – work of the authors themselves play an important role”. (Original in Dutch: “een vernieuwend, zeer leesbaar en praktisch toepasbaar overzicht van de recente inzichten op het gebied van virale marketing, waarbij het – methodologisch geavanceerd – werk van de auteurs zelf een belangrijke rol spelt”.) The van der Lans award was also reported by Dutch trade magazine, Marketing Tribune, aimed at marketing decision-makers (e.g. marketing directors, managers) as well as marketing influencers (e.g. consultants). The online report attracted over 17,000 viewings, indicating wide-reaching recognition of Prof van der Lans' contribution to the industry.
b) A chapter by Prof van der Lans on integrating social networks in marketing decision models [R4] was included in the Handbook of Marketing Decision Models (second edition, 2017). This completely updated edition of the 2008 book was compiled for practitioners as well as academics. The chapter had been downloaded more than 2,300 times by the end of September 2019.
c) The global appeal of such research was highlighted when an article and video on Prof van der Lans’ seeding strategies [R3] appeared online in Dinero, a prominent business magazine in Colombia, in 2017. Highlighting Prof van der Lans’ digital marketing finding that proximity and intensity in social networks are more influential than numbers of contacts to its readers and viewers, the publication noted: “Understanding the dynamics of information dissemination and identifying influencers can be a valuable consideration for the design of strategic mechanisms, reaching the public, and fulfilling the goals of social media presence.” (original in Spanish: “Entender las dinámicas de la difusión de la información e identificar influenciadores puede ser una consideración valiosa para el diseño de mecanismos estratégicos, llegar al público y cumplir los propósitos de la presencia en redes.”
d) Further dissemination in presentations to a range of professional audiences has extended the reach of his research’s impact on practitioners’ understanding. In 2015, for example, Prof van der Lans presented on optimizing viral marketing campaigns in a keynote speech at the Enterprise Marketing and Globalization (EMG) Conference in Singapore before an audience of 50 professionals and academics. He also relayed his knowledge about viral marketing in the HKUST Executive Education Webinar Series in 2017, and executive programs in 2012 and 2016, with around 50 participants per program.
References to the Research
[R1] van der Lans, R, van Bruggen G, Eliashberg J, and Wierenga, B (2012). “Seeding a Message to Harvest Reach: Predicting and Optimizing the Spread of Electronic Word-of-Mouth”. GfK Marketing Intelligence Review, 4(1), p. 32-41.
[R2] van Bruggen, G, and van der Lans, R (2014). “Virale Marketing: Hoe Campagnes zich Online Verspreiden”. Jaarboek MOA/Ontwikkelingen in het Marktonderzoek, p. 9-24.
[R3] Chen, X, van der Lans, R, and Phan TQ (2017). “Uncovering the Importance of Relationship Characteristics in Social Networks: Implications for Seeding Strategies”. Journal of Marketing Research, 54(2), p. 187-201.
[R4] Chen, X, van der Lans, R, and Trusov, M (2017). “Integrating Social Networks into Marketing Decision Models”. Handbook of Marketing Decision Models, Second Edition, Springer, 505-529.
[R5] Gelper, S, van der Lans, R, and van Bruggen, G (forthcoming). “Competition for Attention in Online Social Networks: Implications for Seeding Strategies”. Management Science.