Creating Value in Wikis and Blogs via Viral Marketing

Note: This case study is based on research which appears in
Measuring Value in Wikis and Blogs
Social Networks and Viral Marketing
Why Some Viral Marketing Doesn’t Work
Yes, You Can Predict Viral Marketing
Reading Virtual Minds Volume II: Experience and Expectation [[we hope it’ll be on Amazon soon!]]

The Client was a small New England university that was having difficulty attracting promising students and students with high academic standing.

The Problem was that the university had not kept up with changes in student habits and was not well equipped to teach students educational disciplines (such as the Socratic Method) for understanding and solving problems via the students’ new habits. The college did recognize that many of their students spent a great deal of time online and, although the college had its own internal and external web sites, student use of these tools was negligible.

The question which was asked had two components:

  1. Is there a way to increase visitor access, usage and appreciation of specific material on web sites?
  2. Is there a way this material can be designed to increase cognitive skills?

The Solution involved what NextStage calls site determinants.

This project was taken on before the concepts of wikis, blogs and viral marketing campaigns were lingua franca in the web world. That stated, NextStage’s proposal was to create a website with the following aspects:

  1. Content would be specifically designed to meet the non-conscious mindsets of the visitor (student) population
  2. There would be a free trial period followed by a monthly charge for site access
  3. Visitors would be given an alias and an email account on the system by which to contact each other
  4. Visitors would be able to post comments about the content and each others’ comments which would be available to all others visiting the site
  5. There would be no censoring of posted material
  6. Repeat visitors were given more and different content
  7. “Super” repeat visitors were given semi-administrative roles within the website
  8. All growth would be organic
  9. A core group of 20 students were given access and told they could direct others to the website only after they had created their own account (visitors who came to the site from other than a site-sponsored email account were not allowed access)

The site and the content contained therein was not affiliated with the college in any way other than being created and maintained by an associate professor. The core group of students were culled from one of the professor’s classes. Initial motivation for visiting the site was to critique design elements for extra credit. The website would be available for the duration of the course and three months thereafter, although visitors were not informed of this. The payment system (for monthly access) was designed to fail and not accept payments — hence terminate subscriptions with no financial liability to visitors — at the end of the experiment cycle.

The Solution (Goal Determinants)

The entire student body was some 2,000 individuals, about 70% of which were undergraduates. The success metric was determined ahead of time as getting some 600 regular visitors to the site. This is about 30% of the entire student population and about 43% of the undergraduate population. However, as part of the goal was to learn how to attract promising students it was recognized that not all of the end population would necessarily come from within the student body. The experiment was also limited in time-frame, although part of the research was also to determine if the site would be a visitor focus. From initial deployment to end of experiment was a total of six months.

The Result

Visitors and subscriptions increased over time

The figure above compares the Weekly Site Visitors (as determined by unique login ids) to paid Total Subscriptions over the fifty weeks the experiment was active. What is clearly obvious from the chart is that the addition of new, relevant content caused an increase in visitation. Worth noting is the large uptake in visits in week 28 (93 to 122 visits/week) compared to the smaller uptakes due to the addition of new content on week 10 (34 to 51 visits/week) and week 17 (63 to 71 visits/week). Content was determined based on TargetTrack analysis of visitors’ postings (blog- and wiki-style entries) to the site in the form of content commentary. TargetTrack analysis was made of visitor commentaries to determine various communication and cognition styles. Content was then selected which encouraged activity of these communication and cognition styles with the end goal of answering the two questions which precipitated the experiment:

  1. Is there a way to increase visitor access, usage and appreciation of specific material on web sites?
    The increase in traffic to the site demonstrates the increase in visitor access and usage of the site. An increase in blog- and wiki-style entries demonstrate an increase in activity (hence interest if not exactly “appreciation”) on the site.
  2. Is there a way this material can be designed to increase cognitive skills?
    A maturation of writing and communication skills is shown over time in visitor entries to the site. (An example anecdote is worth noting here; One of the assignments was to learn about the Socratic Method, the students had a difficult time completing the exercise. One student, in particular, left the class in tears. This same student, in a later conversation with the instructor, offered that she had taken what she’d learned and applied it to another class, achieving her best grade ever. This same student became one of the most prolific and insightful contributors to the experiment site)

The site content achieved these goals by promoting more commentary and causing visitors to suggest non-visitors go see the content and the visitors’ associated commentary. This element of the experiment was intended to increase traffic and viral marketing, both of which did occur as shown in the chart.

Note also that there is a steady increase in total subscriptions from weeks 6 through 39 with a sharp increase after. This is attributed to the main viral marketers — the student body — returning to campus after summer break and communicating the site to new students.

The goal of 600 regular visitors was realized in week 29 of the experiment.

Also noteworthy is the psychological characteristics of individuals who recruited new members to the site (ie, spread the virus most effectively). These characteristics included and are not limited to:

  • the most successful recruiters were the most active contributors
  • successful females recruiters engaged females more often than they engaged males in posting-exchanges by a ratio of 3 f-f postings to 1 f-m posting
  • off-site social recognition (via sports participation, bands, club affiliations and extra-curricular activities in general) contributed to the recruiter’s success by a factor of 2 to 1 (as measured by total social activity hours/week to successful recruitment of new visitors to the site)
  • regular contributors’ contributions were read more often than contributions from casual contributors
  • recognizable social hierarchies (dominance, gaming, fool/clown/crone archetypes) evolved with the time period of evolution dependent on traffic load

Two final notes to this section:

  1. the online success of individuals (as determined by recruitment and contributions) increased their likelihood of offline social recognition, willingness to take risks and be successful (as determined by follow up conversations, etc., with participants in the experiment) and
  2. the site is currently offline and still receives some 15 unique visits per day even though no marketing is being done.


This experiment — recognizably bridging from a small world model to a large world model — clearly demonstrates the power of viral marketing to promote a product which is intentionally designed to appeal to a specific audience. NextStage contends although can not currently validate that a larger seed audience (more than 450 unique individuals) and more highly refined product would cause an even greater uptake in the use of the such a site. [[We can validate this now.]]

Posted in , , , , , , ,

Leave a comment

Leave a Reply