Think-ThroughViral Coefficient and Viral Growth in Viral Marketing - Think-Through

Viral Coefficient and Viral Growth in Viral Marketing

Ever wondered how to calculate the viral coefficient and what impact it has on the viral growth of users/customers?

With viral being one of the most effective and staggering (yet far from trivial from an implementation standpoint) ways to grow, it’s probably worth giving a look at how to calculate the viral coefficient, and what’s the viral growth path for a product that is inherently viral.

The concept is pretty simple: use your customers to spread your product to other potential customers, leveraging the strong endorsement that comes from the knowledge of the person referring the product (or site, or service).

Of course, the least the frictions to such a process, the faster the product will spread around. The higher the quality and usefulness of the product (or, more in general, its intrinsic value to the potential customer), the more likely the customer will be to promote it. The concept of viral marketing is, in its own essentiality, pretty simple.

How to calculate the viral coefficient? Assume you promote a product to 100 people, and assume each one of these 100 people have 10 ‘friends’, and refer your product to all of them. Assume the uptake rate is 10%, then the viral coefficient is:

Viral Coefficient = # referred * uptake ratio

In our example, the viral coefficient is equal to 10 * 10% = 1

1 isn’t a great viral coefficient. You are really looking for a viral coefficient greater than 1.2/1.3. Why? Because the higher the viral coefficient, the faster the resulting viral growth, and therefore the higher the number of customers at any given point in time. With any viral coefficient below 1, the growth will quickly plateau preventing the viral growth to actually kick in.

The chart below shows several viral growth path ad varying viral coefficients:

Viral Coefficients and Viral Growth

See that for a viral coefficient below 1.3 we can’t really name it a viral growth curve, more like a linear one.

To calculate the user base at any given point in time the formula is:

user base time T = (user base time T-1) + (new users time T-1) * (viral coefficient)

Therefore, at any given point in time the user population is equal to the population a period before + the latest new users multiplied by the viral coefficient.

A viral coefficient of 2 means that any new user brings in 2 additional new users, which will bring in 4 new users, which will bring in 8, which will bring in 16, then 32, than 64, then 128…

The table below gives an idea (click for bigger table):

viral growth table at varying viral coefficient

Viral speed, which is the amount of time every step takes, is fundamental. If it takes a decade to refer, then even with a coefficient of 2 you won’t go far away. If it takes a day, then every day you have a doubling of the user population, and in 33 days you reached 4+ billion people. Of course, at some point the viral coefficient (and therefore the viral growth) slows down and reaches a plateau, the reason being that we are all interconnected, and the more it spreads, the higher the chances our contacts have already been reached by the product, and expressed their interest (or lake thereof).

What internet has done is having sped up viral speed, and make extremely more simple to refer. It just takes a click to refer to 10, 50, 100 friends, and it’s done in a second. Well, not that simple, really, as competition in viral distribution has grown massively, and users are bombarded by offers, making it more difficult to catch their attention and interest, and to trigger a real viral growth. Many things come into play here, and I’ll cover them in another post.

Post to Twitter Post to Digg Post to Facebook Post to Google Buzz Post to Reddit Post to StumbleUpon Post to Technorati


  • 1. Andy replies at 23rd June 2011, 8:56 pm :


    I realize this article is a little old. You give a great explanation of viral co-efficient as well as some great visuals. I am a little confused on the term “user base time” and the calculations revolving around it. If you don’t mind could you maybe provide a working example so i can get a better idea? Thanks a lot for helping a stranger.

  • 2. Travis replies at 12th August 2011, 4:13 am :

    Great Article! We are actually building a 5 Year revenue model that needs to account for virality. I understand everything you mentioned above except for how to calculate the new users at any given time. (The second part of formula.) I think if I could see the excel file that generated your graph then I could understand that final part. Is there anyway I could get a hold of that? Thanks!

  • 3. Advice from Mentor Meliss&hellip replies at 6th February 2012, 7:50 am :

    […] analyze key values, including Customer Acquisition Cost (CAC), Average Revenue Per User (ARPU), Viral Coefficient. Constantly try to improve those numbers and think about how the design of your product or business […]

  • 4. 歡迎來到「平行行&hellip replies at 9th February 2012, 3:34 am :

    […] 而如果你是網路創業者,這件事情又更簡單,請問問你自己,你的「病毒參數」是多少?如果沒有答案,那就快去找答案吧。 […]

  • 5. Viral Coefficient: Bringi&hellip replies at 16th July 2012, 5:59 pm :

    […] a dramatic difference in the number of views. To learn more please look at a few places below. -… - […]

  • 6. Kieran Hanrahan replies at 19th July 2012, 2:29 pm :


    This is a very simplistic model and does not take account of a host of issues that affect a product or service with viral potential.

  • 7. Persuasive E-Commerce Des&hellip replies at 7th August 2012, 1:18 pm :

    […] is why it is so valuable to ‘go viral’. Recommendations by friends are seen as very valuable, and people return the favor by […]

Leave a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <blockquote cite=""> <code> <em> <strong>