Are Network Effects Overvalued?

Screen Shot 2014-12-17 at 1.23.27 PM

“In economics and business, a network effect (also called network externality or demand-side economies of scale) is the effect that one user of a good or service has on the value of that product to other people.”


Network effects have always been the utopian goal of any technology business. Unlike the typical case of diminishing economies of scale – where the cost of acquiring the marginal customer increases as you saturate the early majority of your market – network effects create increasing economies of scale. The incremental user will often pay for access to a bigger network –so margins actually can improve as user bases grow. Network effect “lore” has become the dominant paradigm for start-up building in Silicon Valley.

Everyone knows the story of Facebook vs. MySpace. MySpace had a much larger user base but Facebook focused on the “Social Graph” (and a much better product) and ultimately dominated the network. Paypal followed the precedent of Visa and Mastercard and acquired tons of merchants, believing in the network effects of more merchants leading to more users leading to more merchants leading to more revenue. Most recently, Uber has raised $1.4 billion of capital on the belief that it needs to grow its user base, provider base and served cities to have the largest two-sided marketplace around – on the assumption that the network effects this creates will ultimately create world dominance. eBay understood this very well – optimizing for maximizing the number and diversity of listings and buyers and ultimately dominating the auctions market. Microsoft did it by building the Windows platform and acquiring both developers and users on the platform – crushing Apple in the early years.

The recipe seems simple right?

  1. Find a possible two-sided market.
  2. Invest aggressively to create early supply and demand.
  3. Identify what it would take to have hundreds of millions of consumers, suppliers or other value-creators enter the marketplace.
  4. Raise insane amounts of money to be able to acquire those hundreds of millions of market participants.
  5. Let your network effects turn you into the monopoly that you deserve to be.

Interestingly – the desire to create “network effect” oriented businesses has particularly accelerated as the costs of distribution on the Internet have decreased. It now seems possible (particularly given that capital availability has increased) to actually acquire hundreds of millions of market participants in a way that seemed impossible previously.

But what if those network effects are just fundamentally overvalued? What if the same factors that make it possible to build a network effect, also make it possible to break a network effect?

The core assumption underlying network effects is that switching costs for market participants grow as the size of the network increases. It’s hard for a user to switch their windows laptop for a Mac, because all of the applications they love were written for Windows. One can’t get off the iPhone because all of your devices, music and videos were built for it. It’s hard to get off Facebook, because all of your friends are on it. But maybe those switching costs are fundamentally overvalued – and that actually, its just as easy to get large numbers of users to switch to a new service (even for a network-effect driven business) as it was to acquire them.

Ultimately, users had no trouble switching from Windows to Macs – when Macs became clearly better. And while iPhones had the early lead in the smartphone category – Android has 62% market share versus Apple iOS which has 38%. And empirically – if the Facebook network effects were that strong and switching costs were that high – why would Facebook have paid $20 billion for WhatsApp and $1 billion for Instagram? Why did anyone even start using WhatsApp or Instagram, if the network effects were really that strong? Sure, the use cases and core users of those services might have been different from Facebook but they are adjacent enough that plenty of people would have suggested that the network effects of Facebook would carry it into those use cases much more naturally than a new start-up could emerge.

Switching costs (even for well formed networks) are just fundamentally weaker in the Internet age. While networks of developers, users, friends or merchants might provide substantial initial value, they can ultimately be defeated by a much better product. Pre-Internet, it would have been impossible for an amazing new product to get sufficient distribution to defeat a large network-driven competitor. Today, that new product has substantial access to capital and several billion Internet users that it can reach without ever opening a second office.

Lets do a thought experiment on Uber – today’s infallible star. The theory is that Uber has already dominated the transportation industry, building an amazing network of drivers and users so that no one will ever beat them. But don’t bet against innovation. What if I developed the following strategy? I decided to build the world’s best car service for the city of San Francisco? I provided self-driving cars so that I didn’t need to worry about their network of drivers and had a dramatically reduced cost structure. I had amazing service in my cars (super nice food, drinks, seats). I made the GPS actually work to a level of precision that Uber didn’t. I made the app even slicker and created a loyalty program of my own. I raised enough capital so that I had a sufficient number of cars and an even shorter wait time than Uber. I donated some of my Company’s proceeds to San Francisco charities to foster local buy-in. I focused my marketing maniacally on the communities of people who take cars most often– just in San Francisco. Suppose it was just a much better product for the San Francisco market? What would the switching costs be for a user to download Raj’s Car Service app versus Uber? Would that user be unwilling to do so because Uber has hundreds of millions of users or a large base of drivers?

The ultimate kryptonite for a network-based business is having a better product in an adjacent, but bigger market. Lets take eBay and Amazon. In 2001, as the Internet bubble died a temporary death – Amazon was supposed to be dead and eBay was the clear winner. They were the ones with the network effects from auctions (and no inventory costs). Fast forward to 2014, and Amazon is worth twice as much as eBay. Amazon had the kryptonite. They had a better product (more selection, fixed prices, great fulfillment, great customer service), in an adjacent market (e-commerce) that was so much bigger than the auctions market. Kryptonite.

The quest to build competitive advantage, economies of scale and switching costs is a worthy goal for all of us in technology. But beware the next time someone pitches you an idea on “the Uber of pets.” Ask them why they are trying to build that network? And beware of the trap of investing massively in building the two-sided network – only to have it marginalized by someone with a better product.

Image from network by Simon Cockell licensed under CC by 2.0

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: