How long did it take Facebook to reach a “tipping point” after which growth became exponential? How about Wikipedia? Yelp? YouTube? According to a group of Wharton School of Business and University of Pennsylvania School of Engineering students, Facebook took 3.4 years. Wikipedia took 4.1 years. Yelp and YouTube got there in 2.1 years.
The primary question these students sought to understand is that for two-sided networks, what factors lead to the proverbial “tipping point” after which growth becomes exponential? It is an interesting question especially as many of the most prominent user-generated and community websites are or display characteristics of two-sided networks. It is also a question that many entrepreneurs or investors considering such platforms will be interested in understanding.
While many of their findings & observations are intuitive, we thought that the scope and methodology they undertook to understand the length of time to tipping point as well as their analysis would be of interest to ChubbyBrain community members.
Their analysis sought to uncover insights into the following three related questions:
- For user generated or community websites that demonstrate two-sided network characteristics, what factors lead to a “tipping point” after which growth becomes exponential?
- For websites focused on a similar market, what determines the success of one over the other(s)?
- When thinking specifically about the college student market and developing community or user-generated content websites targeted at them, what do students value and as a result, what should entrepreneurs be cognizant of?
Before diving into the findings and implications, a few very brief definitions/terms related to two-sided networks.
A two-sided network is a market where the value of an agent in one market is related to the number of actions taken by agents in another market. The term network effects implies that the value for one agent is affected by the number of actions of other agents. (source: Economics of Two-Sided Markets by Professor Lorin Hitt, The Wharton School)
Question #1 – How to reach the Tipping Point?
Company Focus: Analysis examined Yelp, FaceBook, YouTube, Wikipedia, and CitySearch.
Methdology: ‘Active users’ data and Google Insights index data for Facebook yield extremely similar graphs over time. This implies that Google Insights data can be used as a reasonable proxy for active user data, which was more difficult to find for other websites in question.
To determine the tipping point, the team examined the Google Insights data. To make the data linear, took the natural log of both dimensions. They found the equation for this line and developed an equation for the original line. To find the growth in the index value over time, they took the derivative and then determined that when the index growth hit approximately 0.05 per week, the websites started to take off and display exponential growth. This point was chosen as the “tipping point”. The progression of events which unfolded and the activities undertaken by these websites support these findings.
The high-level findings for each website in question are provided in the graphs and milestone diagrams given below. Each figure has a yellow dot indicating the “tipping point” according to the methodology already described as well as various milestones along the way.
As you will see, the time to reach the tipping point for each of these sites varies quite significantly as mentioned earlier. They also analyzed CitySearch which did not hit its tipping point as you will see below.
• Facebook = 3.4 years
• Wikipedia = 4.1 years
• Yelp = 2.1 years
• YouTube = 2.1 years
So given the data, how can internet-based two-sided networks reach the “tipping point” quickly and effectively? Unfortunately, there is no magic bullet as you might suspect. The student’s observations are given below.
- Providing incentives or access to third party developers can increase network effects and expand user base. Facebook has done this as had YouTube by making it extremely simple for users to embed their videos.
- When relevant, increasing global appeal can increase user base as was done by Wikipedia.
- In cases where targeted, deep content is more valuable than broad, superficial content, targeting users in popular metropolitan areas may be a strategic way to propel traffic and website use as evidenced by Yelp.
- Luck and finding a big wave to ride do play a role and so macro events beyond your control can greatly affect traffic. It’s up to 2-sided network websites to take advantage of this by finding ways to appeal to a network that might benefit from your traffic and to scale appropriately to meet this demand as YouTube did.
- And finally, complacency can kill. Another offering with a similar mission and improved offering may negatively affect an incumbent’s position fairly rapidly despite scale advantages that they may have as seen in the case of CitySearch vs Yelp.
To see the findings related to questions #2 and #3, please see the attached embedded presentation (51 pages).
Thanks to Juhi Heda (Wharton/Engr 2010), Ursina Beerli (Wharton 2009), Steven Gao (Wharton/Engr 2010) and Grant Wilson (Wharton 2009) for sharing this analysis with us. We’d also encourage you to check out Juhi’s ChubbyBrain reviewer profile where you can see her many well-written reviews of startups. Her profile widget is given below along with the presentation.