Article: Thursday, 5 October 2017
Well-connected social influencers can speed up the adoption of just about everything, from new flavours of yoghurt to vaccination programmes. Targeting these influential people who can then spread the word can make marketing efforts of companies, governments and NGOs much more effective. Researcher Dr Xi Chen of Rotterdam School of Management, Erasmus University (RSM) has now developed a better method to select which influencers to target. It looks beyond just the number of connections these people have and also takes into account relationship characteristics. With this approach, up to 90 percent more people can be reached.
Networks exist not simply because you know people, Chen says. They are defined by the type of interaction you have with them. You could have many online friends, but you usually end chatting and meeting with only a handful. But most models for selecting influencers are still based on the number of connections of these people and their online friends.
For this study, the researchers developed a model of a network that includes character and quality of relationships. The result is a ‘weighted network’, in which the importance of relationships is calculated from actual behavioural patterns within the network. This idea can be adapted to any type of network, online or offline, Chen says.
The researchers first tested their model on an agency that promotes micro-credits in rural India. One of the problems the agency ran into was that ‘official’ communication channels were mistrusted. People would be much more likely to try out micro-credit if they heard it from somebody they knew and trusted. But where to start to maximise the outcome of such a word-of-mouth campaign?
The researchers started by investigating how households interacted with each other in the separate villages. They discovered 12 significant types of relationships, two of which proved most important: ‘going to temple with others’ and ‘being a relative’. Applying this filter allowed them to assess the relationships between households, and ultimately determine which households should be the ‘seed’ – the first to receive information about the micro-financing programme. Tests on existing data showed that by selecting seeding households using this method, the campaign could have reached 10 per cent more households with the same effort.
In a follow-up study that observed discussions about Superbowl commercials in an online social network, the researchers calculated that up to 90 per cent more people could be reached by selecting the right influencers. In this case, two characteristics of online friendship, the ‘number of messages exchanged’ and ‘duration of the relationship’ predicted most accurately what types of relationships should be most valuable for determining social influence.
The analysis showed that the increase in the number of people reached was achieved with a maximum of just 50 seeds. Adding more seeds would lead to overlap in audiences and diffuse the marketing efforts, Chen notes. Interestingly, those social influencers were often not the first to mention the commercial. Being an early adopter is no guarantee for having social influence, they discovered.
This approach, of carefully selecting influencers, is especially valuable for brands that run their own online communities or forums, such as Adobe or Apple, Chen continues. Having detailed insights about their users and how they interact allows them to identify who to target when marketing new products or services for example, he concludes.
Rotterdam School of Management (RSM)
RSM offers Executive Education and Master programmes in various business areas for any stage of your career. For instance:
Corporate Communications & PR Manager
Science Communication and Media Officer