The Science Behind Viral Marketing

Many people don’t realize the advances in the science behind viral marketing.   Experts often known as “Viral Tuners” are applying a systematic data driven process to creating viral customer acquisition drivers.  By testing and optimizing the viral elements of a widget or websites they are often able to push the viral coefficient above 1.0.  This essentially means that each new user generates more than one additional user.  In other words, growth accelerates geometrically until saturation.

Even if you can’t get the viral coefficient above 1.0, you can still greatly benefit from virality.  The echo effect of a viral coefficient that falls below 1.0 is still important. For example, if a tracked user from a campaign costs $1 to acquire and you have a .75 viral coefficient, you actually acquire about 3.5 customers for that $1 and therefore your average customer acquisition cost goes down to less than 30 cents. If you can find a way to generate an ARPU (average revenue per user) of $.50 you have a strong business – even if you spend $1 to acquire that first user.  The higher your viral coefficient, the bigger the echo effect of every dollar spent.

Again, if you can get the viral coefficient above 1.0, the echo effect is infinite and your cost per acquired user approaches zero.  Any time you can acquire users for free, a viable business model becomes almost inevitable.  As long as you don’t have a significant marginal cost per user outside the acquisition cost, you simply need to micro-monetize these people and you’ll easily be able to grow a profitable business.

All of the above is pretty abstract, so I’ll give you an example of the only viral growth driver I’ve ever created. Ten years ago I launched a YouTube sized game widget called Trivia Blitz that appeared on over 35,000 websites. Each widget carried an “add this game to your website” link, helping it to spread virally between sites.  We acquired several million users with this widget, paying the referring site 50 cents per user (a small fraction of the average lifetime value each user generated in advertising revenue).  To achieve a high average lifetime value, it was important to transition users from the referring site to the engaging multiplayer games on the Uproar.com website – which was already among the stickiest on the entire web.

During these irrational dotcom bubble years, our competitors were wasting millions on crazy marketing campaigns like expensive branding campaigns on TV or multimillion dollar AOL deals.  By focusing our resources on Trivia Blitz and tightly tracked direct response marketing drivers we were able to take leadership of the game category, surpassing Microsoft, Sony, Yahoo, and many gaming startups backed by leading Silicon Valley VCs.  Trivia Blitz was such an efficient customer acquisition tool that it eventually helped position Uproar.com for a NASDAQ IPO.  Among public companies, Uproar was able to achieve the lowest customer acquisition costs for a free registered user (about 1/6 of Yahoo’s cost which was considered pretty good at the time).

While we didn’t optimize the virality of Trivia Blitz, it is still illustrative of the results that are possible if you can create a viral coefficient above 1.0.  I’m now focused on learning the skills of viral optimization, so I can take something that is slightly viral and fine tune it into a major viral marketing driver.

If you are interested in reading more about viral marketing, I highly recommend Andrew Chen’s blog.  I’ll blog about any other great resources that I find.

3 thoughts on “The Science Behind Viral Marketing

  1. Pingback: What Makes a Great Startup? « Startup Marketing Blog - By Sean Elis

  2. Hi,

    In your post, you wrote:

    “For example, if a tracked user from a campaign costs $1 to acquire and you have a .75 viral coefficient, you actually acquire about 3.5 customers for that $1…”

    How did you compute the number of acquired customers as 3.5? Doesn’t a viral coefficient of .75 indicates we are going to acquire 75 new customers for every new 100 customers?

    I’d really appreciate if you clarify this issue. I am a beginner in this field and I’m sorry if my question sounds stupid.

    Thanks and congratulations for this really great blog!

  3. Hi – you are right on the 75 new customers for every 100 new customers, but then you have to take the same ratio on this second set of 75 users and then again, etc… Eventually the chain dies. Hope that makes sense.