When a graph showing an increase of unique Facebook visitors (March ’08 to December ’09) crossed my desktop at StepForth, one nagging question immediately entered my mind: Why doesn’t the number of Facebook visits have exponential growth? By definition, isn’t social media supposed to facilitate ‘Exponential’ — ‘Viral’ marketing, and shouldn’t the adoption of Facebook be a perfect illustration of this?
One conclusion: While Facebook may be ‘crushing’ the competition, market saturation may be close at hand (i.e. a growth plateau).
What do I mean?
Let’s assume that I just arrived from Outer Space and landed my spaceship on the grass of Golden Gate Park in San Francisco. Let’s also assume that the first Earthling that I encounter introduces me to Facebook.
Let’s suppose that after logging into Facebook for the very first time (and after I add this Earthling as my 1st friend), I decide to call 2 of my Outer Space buddies on my intergalactic telephone and tell them about Facebook. If my buddies create Facebook accounts, Facebook would see 3 new users contributing to the more than 450 million unique visitors per month. Now, let’s assume that my 2 Facebook friends from Outer Space each add 2 of their friends (that are not also a friend of mine or of each other*). Now, there are in total, 7 new aliens from Outer Space using Facebook. If this keeps happening (aliens from Outer Space keep telling their friends about Facebook and their friends keep joining) the graph that shows Facebook growth should have a sharp upward increase over time.
The graph released yesterday in this Silicon Alley Insider article (Figure 1) that crossed my desktop suggests that the adoption of Facebook the past few years is not exponential— that friends are not telling friends, who are not telling friends about Facebook. What gives here? Shouldn’t the world’s biggest social media website have the characteristic growth of ‘something’ social? For example, such as the number of feral rabbits at the University of Victoria? Rabbits are very “social”. (Rabbit populations are modeled by something called a Fibonacci Sequence and as the students on that campus know, exponents are definitely involved). Shouldn’t the growth of Facebook look like a pandemic spreading across the population, or a virus multiplying in our bodies? Where’s the expected ‘viral’ part of Facebook’s growth?
If we do a quick and dirty analysis of the plot in Figure 1, we can show that the growth is linear or very weakly exponential. According to the curve fits in this analysis (Figure 2), the Facebook growth curve is actually best fit to a linear regression line. Hence, nowadays, it appears that Facebook is faced with linear growth and is definitely not ‘viral’. This growth might eventually decrease towards zero (market saturation). What I expected to see when the Silicon Alley Insider article crossed my desktop is illustrated in the inset of Figure 2 which was taken from an article published in 2006 showing a year’s worth of Facebook growth.
This line of reasoning has led me to 4 possible conclusions:
- Facebook will soon be experiencing market saturation if they don’t reach some new markets,
- Facebook is not spread like a virus but is sold to new users in a more ‘individualized’ way (and I’m not sure exactly what this would be),
- The ‘exponential’ part of Facebook’s growth has fallen off the left-hand side of the graph (the exponential growth is past history).
- *We are running out of new friends who could be added to Facebook who are not friends of someone else already on facebook (this would result in an exponent that is approaching 1 from something higher than 1). Perhaps this relates to that 6th degree of separation idea?
Do you think Facebook is about to reach saturation? Please share your thoughts.
Dr. Zeman specializes in the analysis of biometric (eye-movement, muscle activity, brain activity) and behavioral data and does research and development in the areas of serious video game design, human-media interaction, and evaluation of therapeutics on our brain function and behavior. Additional information can be found on his website: http://www.abvsciences.com/expertise/