Social Networks: Winner Takes All?

As an early stage technology investor, I steered clear of any one pitching another ‘social network.’ Anyone who wanted to go up against Facebook just sounded crazy. A question that’s thrown around too often in startup feedback sessions is “Well what if Google tries to build it?” This is a meaningless question because you can use it with just about any startup idea, but in the case of social networks, Google did try to build it… And yet even Google+ is trailing behind, so clearly this is a winner takes all market, right?

This summer I worked at Google[x], Google’s innovation lab known for its self-driving cars, Google Glass, and most recently announced moonshot, Project Loon. While interning here, I was surprised to learn that of all the dozens of campuses, hundreds of buildings, thousands of new initiatives, CEO Larry Page, sat on the same floor with the Google Plus team. He put 1000+ employees behind this project, and he situated himself amongst them. Google has not accepted that Google+ may be a failed social experiment. And if you’re like me, you have never logged into Google+ since you (maybe) created an account…

But then I read this VentureBeat article over the summer titled “Google+ continues to dominate LinkedIn & Twitter, could catch up with Facebook ( that stated, “of people who use social networks to log in to other websites, almost 46% use Facebook, but Google+ is a strong second place with 34% of social logins… At this rate Google will surpass Facebook by May 2016 at which time its users will generate over 1,096 billion +1s per month, while Facebook users will generate just 849 billion shares per month.” Whether you’re tracking registered users, active users, social logins, or shares, Google+ seems to be a surprisingly serious second. So is there an opportunity for another social network?

There is much written about the features of Google+ versus Facebook so I won’t spend time describing that here, but what I found interesting was the framework I recently learned in class on how to evaluate whether a space is in fact “winner takes all.” The three criteria for a Winner Takes All market are: 1) strong network effects, 2) high homing costs, and 3) low demand for differentiated product. This framework offered a new lens in which to think about Google vs. Facebook vs. New Startup. We know there are definitely strong network effects at play (the addition of each new user adds value to all the users on the platform). Are there high homing costs? I think most people would find it burdensome to go to Google+ to post something, and track comments on that platform, and then log into Facebook to post the same thing, and follow the discussion that is happening within that walled garden. So yes in the case of Google vs. Facebook, the high homing costs are working against Google+. Is there a low demand for differentiated product? Arguably, no, customers could value different features in a social network. The most common difference referenced is the desire for choosing groups to share particular content with so that not every one sees the same posts. There’s a desire for privacy or more selective information sharing across different types of connections. I’m sure if I asked you to name something annoying that you wish you could change about Facebook, you could easily think of three or more things such as “remove advertisements, only show updates in news feed from people I actually care about, or how can I easily remove all those ‘friends’ I added a long time ago without thinking twice about how many loose connections I now have on Facebook?” So this last criterion is in Google’s favor, but 1/3 on your side is still an uphill battle. However, based on these criteria, if Google+ nails the product differentiation then they could still in theory dominate Facebook.

But what is interesting here is that social networks may be one of the few spaces where a new startup has a better shot of taking Facebook than Google does. And here’s why. Looking at the same three criteria, a new startup has the potential to win on both the 2nd and 3rd criteria.  Unlike Google, which will probably never add the ability to “also post to Facebook,” a startup can be agnostic and say, “Do you want to also post to Facebook and Google?” And thus now this user has low homing costs to using this startup social network because anything shared here can be easily posted across Google, Twitter, Facebook, you name it, allowing it to gain user traction.  Then if the startup can nail product differentiation, then hey, I guess it could win and take it all.

Note: In a way, Instagram could be an example of a startup that succeeded at this. It was a brilliant acquisition by Facebook, but Twitter should have bought them. Twitter would be trending very differently if it owned Instagram. Now that’s another blog post.







read more

Over the summer I had the opportunity to work for both a booming start up – Airbnb – fighting its German copycat – Wimdu – and the dark side of the force in Rocket Internet’s version of Zappos in Brazil: Dafiti.

 I was therefore able to observe first hand how well the Samwer brother’s cloning method (basically pouring marketing dollars into the business) was effective when 18 months after it’s launch, JP Morgan invested 45M$ in Dafiti’s e-commerce business1.

 Simultaneously, barely a year after raising a huge initial round of funding (90M$)2, rumours of Wimdu closing its London and Paris offices reached my ears (the information has not yet been confirmed by RI). I couldn’t help meditate about the reasons why a similar method worked in the case of Dafiti and not Wimdu.

On paper, Wimdu had everything  to become a long-lasting competitor to Airbnb:

·      Platform is exactly the same

·      Marketing message is identical to Airbnb’s (“Travel like a human” vs. “Travel like a local”)

·      Value proposition to both guests and hosts are extremely similar (free listings and 3% commission for the host, 6-12% commission for the guests on Airbnb vs. 15% on Wimdu)

·      Wimdu had local offices in most major markets before Airbnb (Paris, London)

·      Both Wimdu and Airbnb’s pockets were as deep (90M$ vs. 120M$ funding)

Therefore what can explain this first-round knock-out?

1. The nature of the market called for a clear winner

It seems as though Rocket Internet bet big on this one. Indeed, the peer-to-peer industry for rooms / apartments has all the criteria to be a “winner-takes-all” market:

·      High positive network effects: the value of being a part of the “community” increases with the number of users for both hosts (increased demand / revenue) and travellers (competition driven prices on a given location, wider range of destinations). In addition, due to the trust factor involved in peer-to-peer transactions, the incumbent (Airbnb in this case) benefits from the higher number of past transactions (reviews of hosts and guests) thus strengthening these network effects.

·      High complexity to manage a listing on several marketplaces: although travellers are only one click away from another marketplace, hosts have extremely high multi-homing costs from managing a listing on two platforms (i.e. complexity to sync calendars and track bookings, etc…).

It is therefore not surprising to see one of the two emerge as the clear winner but what triggered such a quick victory?

2. Love mattered more than size

The two companies chose very different strategies when allocating their resources. Wimdu spent millions on “recruiting” listings3 via their sales force and travellers via online marketing to make up for being the underdog. In the meanwhile, Airbnb focused on its customer service and its community of users.

Unfortunately for Wimdu, in this industry, the viral coefficient is probably higher than anywhere else:

·      Every new host is a potential future traveller and every happy traveller is a potential new host

·      Additionally, the highly social nature of travelling (who likes to travel alone? who can refrain from “telling all about their vacation” the day they get home?) increases the importance of generating positive word-of-mouth

By focusing on customer service (24/7 customer service, 17 languages served, 90% of calls answered in 90 seconds,), Airbnb grew its community organically (in France, Airbnb had 4,000 listings before even having an office and a French website!). On the other hand, Wimdu let multiple incidents occur without reacting4=;

Additionally, Airbnb understood early on that with such a huge part of their business relying on “trust”, reaching out to their community of hosts was a key component of their success. By organizing offline events where hosts could “put a face on Airbnb” and enabling them to interact between each other, they created an offline community that powered their growth (in France, the number of listings doubled in the 4 months following the first offline meetups).  

Ironically, Airbnb could almost have been called a Zappos culture copycat for embracing Tony Hsieh’s perspective on customer service.

For all those romantics out there, it is a great story to think that in this day and age, “doing business like a human” can still overcome the aggressive strength of marketing dollars.  The question is: will Rocket Internet learn the lessons from this as they launch a Pinterest copycat (called “Pinspire”) in Asia5 ?





read more