For every Facebook, Twitter, or Instagram, there’s a Friendster, Pownce, or Color Labs —
a promising early competitor that has now been relegated to the social network graveyard.
For every Facebook, Twitter, or Instagram, there’s a Friendster, Pownce, or Color Labs — a promising early competitor that has now been relegated to the social network graveyard.
The reason why one social network succeeds while a near-carbon copy fails is often a mystery. The general consensus is that there’s no “magic formula” to guarantee the success of a social network — it’s a combination of the right team, product, and market coming together at the right time.
To explore any hidden insights from social networks of the past, I analyzed 37 venture-backed consumer social apps in three categories:
- Successes — 100M+ users and growing, many have IPOed or raised $100M+ in funding. Includes Snapchat, Facebook, Twitter, YouTube, and others.
- Failures — shut down (sometimes post-acquisition) or dramatically slowed/negative growth. Includes Formspring, Yo, Vine, Yik Yak, and others.
- “In purgatory” — upwards of 1M users, but seems to have slowing or stagnant growth. Includes AfterSchool, AskFM, Whisper, HouseParty, and others.
For these companies, I collected data on the founding team, platform features, user demographics, user growth, and funding history. I also analyzed academic research on virality in social networks, as well as case studies on competition, “fad” apps, anonymous apps, and pivots.
The following are my top ten lessons for evaluating the probability of success of a social network at the early stages:
- When it comes to the founding team, don’t shy away from dropouts, but undergraduate school does matter. Successes were approximately 3x more likely than other companies to have founders who dropped out of college. With the highest rates of success occurring when all co-founders dropped out together. Successes were also approximately 3.5x more likely to have founders who started at an Ivy League school or Stanford.
2. Age matters in a founding team, and experience is likely overrated. Successes were more than twice as likely as other companies to have a founding team with an average age of 25 years or less. Failures were most likely to have teams that spun out of tech companies (about 1.4x more likely than successes), and almost no successes were founded by serial entrepreneurs or former venture capitalists.
3. A social network’s UI should be simple, and not distract from the core post/response functions. Academic research has shown that the strongest signals of user retention over the long term are a user’s number of posts (and the responses they receive) in the first few days. Therefore, extra features that distract from user interaction can be harmful. One example is Meerkat, which Newsweek described as “cluttered and confusing” as compared to Periscope.
4. A social network needs to support a range of content mediums. Networks that focus on only one type of content (for example, audio or text-only) may seem to provide unique value. However, more than 75% of successes in my sample support posting video, text, and photos. This gives users a greater chance of being able to interact with content at all times — for example, users can listen to audio while driving, and look at text or photos in line at the store.
5. The long-term outlook for anonymous networks is not positive. Many anonymous social networks are still in the “purgatory” phase, while others have fallen by the wayside. The Washington Post reported that of the top 25 anonymous apps in 2014, only three still exist (and are in decline). UC Santa Barbara research found that connections made in anonymous networks tend to be weak — for example, only 13% of Whisper users have acquaintances they interact with across posts. This results in high user churn.
“You can’t create a sense of lasting connection if your name is not attached…It’s the inherent problem with anonymity: The mainstream will never embrace it.”
— Secret co-founder David Byttow
6. Seed funding and rapid early growth aren’t indicative of future success. Each of the three categories (success, failure, purgatory) had approximately equivalent average seed rounds of ~$1.1M. A network’s speed in reaching 100k or 1M users was also similar across the categories, with eventual failures actually growing slightly faster at the early stages. Whether or not a network can break the 5M or 10M user mark is a much better indicator of success.
7. A pivot isn’t necessarily a bad sign — and neither is having a competitor acquired by a major player. Many successful social networks have been created from the ashes of another failed startup, if the failure allowed the team to glean some crucial behavioral insight. Twitter (Odeo), Pinterest (Tote), Instagram (Burbn), and Foursquare (Dodgeball) are a few examples.
Similarly, though having a competitor acquired by a company like Google may be daunting, large companies have historically struggled with managing acquired social networks. Vine (Twitter), Formspring (Twoo), Del.icio.us (Yahoo), and Jaiku (Google) are examples.
8. Direct messaging and friend recommendations are key features. All successes in my sample had one-on-one messaging capability (some, such as Tumblr or Pinterest, added this over time). Successes were 1.5x more likely than failures to provide friend recommendations, even if only through integrations with other networks. The ability to tag users was also correlated with success, while location awareness and gamification had less significance.
9. A company’s reaction to security and performance issues are crucial. Issues like network speed or cyberbullying can be fatal at an early stage. Hohenheim University research established that key success factors in the adoption phase are data security, website stability, and reaction time. Down to Lunch never recovered from claims it was affiliated with human trafficking. A positive example is Whisper, which hired a team of 100+ moderators and deletes approximately 20% of posts for inappropriate language. Secret, whose founder infamously shrugged off bullying claims, hasn’t fared as well.
10. Early press and celebrity endorsements are just noise. It’s tempting to believe that “buzz” (especially on Twitter) will propel a network to success, but this hasn’t historically been the case. Pownce (Kevin Rose) was one example of this, More recently, Peach and Yo launched to great fanfare, but were unable to sustain momentum. Tracking actual downloads and churn gives a more much accurate picture.
Though the success or failure of a network will still ultimately come down to the intangibles, the signals outlined above may serve as a helpful guide when deciding whether or not an app is worth downloading (or investing in).
Have you found any other indicators of social network success? Email me at email@example.com.
Disclaimer: Data on the 37 VC-backed companies was pieced together from a variety of sources, including company websites, press releases, news articles, blog posts, and team LinkedIn pages. This analysis represents a best estimate given the available data.