Blog Post

Not All Network Effects Are Created Equal

Part of the excitement driving LinkedIn’s IPO last week comes from investors associating social media with network effects. You know, the principle that the value of a network increases dramatically with the number of its participants. That’s the engine that drove Facebook and eBay (s ebay), not to mention the public telephone network. Markets with network effects tend to have explosive growth and can end up with winner-take-all market share.

But not all effects are equal, and assessing the valuations and competitive positions of social media companies depends on knowing which network effects are actually at work and how those could play out.

First, let’s look at a few of the different network effects currently in social media:

  • Core network effect utility: There’s a difference between economies of scale and the magic of adding connections to a network. Groupon is building a user base, a sales force and relationships with thousands of merchants, but until it uses its sales data to offer personalization, targeting and other marketing programs to merchants, it won’t achieve much beyond scale. Even then, once critical mass is achieved, additional connections don’t add as much value.
  • Viral growth: LinkedIn and Facebook initially grew their networks the old-fashioned way: Users invited other users to join. Viral pass-along is a key growth driver for social commerce and games, but now services and apps can hitch a rideon existing social networks, leveling what was once a steep playing field.
  • Business model that reinforces the effects: While there are minimal network effects for its search users, there are huge ones for its advertising network. Google’s $25 billion in extremely profitable search advertising depends on attracting advertisers to its dominant search audience and insuring a liquid marketplace via bidding and enforced relevance to create an unbeatable paid search business. Plus, Google lets developers using its services and APIs tap into that revenue stream with minimal effort.
  • Participant lock-in: Technology platforms create positive business opportunities for developers. But they can also achieve customer lock-in for their originator by making those same developers dependent on APIs. End users can be locked in, too, via familiarity (e.g., the QWERTY keyboard) and data storage (e.g., contact info, photos, message repositories) that raise switching costs for members.

As illustrated in the table above, here’s how network effects are shaping competition in selected social media markets:

  • Social graph: Though there are network effects aplenty, consumers tend to belong to multiple networks (Facebook, Twitter, Foursquare), meaning would-be data miners must target multiple data sources. And the industry is only just beginning to harness that collection of big data into reliable revenue streams.
  • Likes and log-in networks: Facebook was smart to hang Likes and Sign-ins off its Connect network, as each feature complements the others and assists in distribution; now LinkedIn and Google are trying to do the same. Bolting an ad network on top of those networks could provide missing revenue reinforcement.
  • Social commerce: As noted, most social commerce is more scalar than social. Without the additional services for consumers and merchants previously mentioned, single-market entry barriers and switching costs will remain low.

To read about more network effects and how those are shaping the current crop of social media companies, please see my latest Weekly Update at GigaOM Pro (subscription required).

Image courtesy of flickr user anselm

5 Responses to “Not All Network Effects Are Created Equal”

  1. I think one key will be O2O-Online to Offline-marketing.

    Groupon, Yelp,, etc. If sites can translate online activity to meaningful offline activity (in terms of business), then the network effects will be very real.

    It’s definitely a multipart equation.

    (see my previous comment about conversion, which I should’ve mentioned after posting this).

  2. Brilliant post! The social networks are sitting on data gold mine and have barely scratched the surface on the full potential of social data. Even though social usage data has the ability to revolutionize the entire customer experience from awareness to post-purchase, it’s tragic that most social sites are still relying on traditional ad models for revenue generation instead of creating new monetization/business models. I suspect it’s because the social sites have no clue what to do with all the data that’s passing through their systems every second.

    • Mia,

      I fully agree re: revenue models. I think it’s a case of tunnel vision: see what your “rabbits” are doing, and try to do it “better.” If other sites are pursuing the traditional ad model to generate revenue, it must be the best or only way, right?

      Twitter wasn’t profitable until it sold its data to Google and Microsoft, right? But that means it wasn’t doing anything with the data itself; it wasn’t making money off of it.

      There’s definitely value in data–see the sites that have been bought up for their metadata alone, like IMDB–but I can’t think of any unique monetization models.

      David, there was a good article (or I may have seen it on TV) about Groupon. Critics say that a simple Groupon rush does not result in repeat business, just a spike in traffic that may serve as advertisement for the business. That means that “conversion” of first time customers is still key.

      Facebook fan pages might show 100K+ “likes” but what does that actually translate to for a business? How do we know those users weren’t “trigger happy” and just hit the Like button because they saw friends like the same thing? And how do we know those users actually follow fan page activity, such that it results in new business?

      I am curious what the next great revenue model will be. Whoever comes up with it will be a true pioneer. Great article, David.

      • David Card

        From what I understand, most merchants indeed use Groupon for new customer acquisition. To sustain its growth, it and the other daily deals sites will have to plug into customer retention and loyalty, and offer other marketing-like services for their merchant partners.

        Matthew Ingram has written about how Likes drive traffic to news sites, so there’s some evidence that Likes indicate real recommendation value. But brands have to use the access a user’s Like gives them (in the form of feed updates) to encourage user return visits and further pass-along.