This is part 2 of our 2 part review of the academic paper ‘Seeding Strategies for Viral Marketing: An Empirical Comparison’ by Hinz, Skiera, Barrot and Becker. You can view part 1 here.
The social seeding strategies study involved three distinct seeding experiments to test the impact of high-degree seeds (hubs), low-degree seeds (fringes) and high-betweenness (bridges). In this article I am going to explain each of these experiments and draw on some important lessons for businesses.
Each of these seeding categories had a particular hypothesis attached to it in the study:
- Hubs – hubs have the highest numbers of personal contacts and so the hypothesis is that they will maximise the reach of a campaign because they will literally reach more people (e.g. Rogers 1962).
- Fringes – fringes maximise the reach of a campaign in virtue of the fact that they supposedly have more influence over their peers as their relationships with their connections are more meaningful (e.g. Becker 1970).
- Bridges – bridges link otherwise unconnected networks which means they will maximise the reach of a campaign by allowing the information to filter into other networks (e.g. Kemper 1980)
Study 1: Controlled setup
120 students were recruited to take part in this experiment. Varying levels of extrinsic motivation was factored into the experiment to share ‘secret tokens’. Four different groups were targeted; hubs, fringes, bridges and random to see which group of people was the most responsive.
Results: Hubs maximized responses and achieved the most responses. Bridge based strategies also worked well.
Study 2: Realistic setup
In this experiment business students were chosen to participate and intrinsic motivation was factored in – to share an interesting video that was relevant to them. Four different strategies were employed; hubs, fringes, bridges and random.
Results: Confirmed the findings of study 1 – that hubs and bridges are more effective strategies.
Study 3: Real world referral programme
This was a study of data from a real referral programme for a mobile phone provider. The study aimed to identify what factors were the driving forces behind the social contagion process. A calculation to establish the degree-centrality of each customer was worked out based on individual level call data. Other factors were taken into consideration to explain the likelihood of referrals e.g. age, gender, tariff plan and service usage.
Results: Hubs are more likely to lead to referrals (quantity) and these referrals are more likely to be successful (i.e. converted into a new sale). Also, hubs are more likely to participate in a viral campaign. What was recognized in this study through an analysis of the data is that hubs don’t have a direct influence on the success of their referrals – in other words, hubs aren’t better at referring; their success is based on their more extensive reach.
Summary of main conclusions
High-degree and high-betweenness seeding both came out on top in these studies. Both methods out-performed low-degree and random seeding suggesting that seeding strategies will benefit dramatically from targeting hubs and bridges. The studies also revealed the importance of socio-metric measures which seemed to have a greater influence than loyalty or revenue measures. The study also discovered that social media platform providers have access to data that could be very useful for targeting purpose – but that this valuable data has not yet been used.
One of the most interesting points is that hubs are more effective seeds because of the quantity of their connections and because they are more likely to participate – not because they are in any way more persuasive than fringes.
Useful lessons from the studies
The results obviously go a long way to guiding any sensible seeding strategy but they are also useful in terms of application to a less formal approach to disseminating content online.
The value of sharing video content
Individuals very much rely on personal recommendations when it comes to which online platforms they use, what they buy online and what they watch online. The importance of sharing via social media can’t be underestimated and this importance has been recognised by marketers who have adjusted the style and substance of their online content to make people want to share it. This process can be refined further by sharing the content with the right people in the first place – to give it the best springboard possible in the realm of the viral.
Targeting content to relevant audiences
At the very least, businesses can take from the study the importance of thinking before sharing content. Talking specifically about online video, there would be no point investing in producing a video and then just putting it on YouTube. The video should be targeted directly at relevant businesses, organisations and individuals who are likely to be interested, and know others who will be interested. It is worth bearing in mind that highly connected people are often more likely to share because they like to be seen to have adopted an innovation first (Burt 1987).
Kevin Allocca talks about Tastemakers in the following video – another term for a ‘hub’.
Seeding video content makes financial sense
With viral marketing campaigns, audiences do most of the leg work. This makes viral marketing campaigns relatively inexpensive when compared to more traditional forms of advertising. As long as the marketer has an understanding of their audience, produces content that they will connect with and want to share and launches the content in the best way possible (i.e. a seeding strategy) a viral marketing campaign can prove more successful as well as less expensive than traditional methods.
There have been many attempts to identify a formula for what makes a video go viral and the answer seems to be that, strictly speaking, there isn’t one. However, these attempts have uncovered that there are a number of things that you can do to give your videos the best chance of going viral, and a social seeding strategy is one of them.