Network Effects

I was lucky enough to spend a bit of time this summer getting exposure to VC thanks to Neil Devani, a generous mentor and investor at Necessary Ventures. As I was going through a bunch of investor memos and fundraising decks that dove into some of the most brilliant startups today, I repeatedly came across a buzzword: Network Effects. I had absolutely no clue what it meant but its concept was being used to back bold, yet questionable statements such as exponential growth, impenetrable moats, and indispensable TAM. The concept was used so widely and vaguely that I decided to jump into the void and attempt to demystify its polluted status as a buzzword. As Neil mentioned, with every investment, there has to be a narrative of unlimited upside regardless of what stage you are in, and it turns out, network effects help achieve exactly that. I was able to learn about the concept of network effects thanks to an incredibly deep dive on the platform economy through Platform Revolution by Sangeet Paul Choudary, a world-renowned business-strategy expert and fortune 500 advisor. I’ll be diving into the content of the book and giving my take below.

There’s a pretty well-known concept called “economies of scale” that every student who takes a business/econ class learns about. Historically, the concept has referred to the supply side economy of scale. The definition we were all taught in school is for each additional unit of output, the cost of producing those units goes down. Here’s an example: the more cars a manufacturer is producing, the lower the unit price becomes. The first car was horribly expensive but every following car not that much. This is a competitive advantage for businesses that allows them to out-scale competition and dominate markets. What we have now, with platforms, or two-sided markets, is the demand side economy of scale. What does this mean? The more people who want to participate on your platform (as a consumer or supplier), the higher the value of that platform becomes. In other words, network effects are the incremental benefit gained by an existing user for each new user that joins the network. This phenomenon is the loose concept of what platform companies today like Uber, Airbnb, Google, Netflix, etc. use to achieve unprecedented scaling and valuations. But before we understand network effects, we need to understand what a platform is.

So what we historically have had are value chains, aka pipeline business in platform jargon. A company had suppliers, would add value to the product, and then sell it with profit to their customers. It’s a pretty linear process that goes through multiple stages of value creation starting from the supplier to the consumer. Platforms work differently. Platforms don't operate value chains but value networks. Platforms create value by enabling transactions between supply and demand in fragmented/disorganized markets. Think of Airbnb. They only provide the infrastructure for supply and demand to handle transactions and also enable both sides easy access to each other. The key to understanding network effects is to know that platforms can work with multiple parties and add value through market aggregation and market Intermediation. They don't sell products, they provide a platform that creates a market, which is then monetized by taxing transactions and market access for both demand and supply.

What makes network effects so special is that it has spawned a new method of value-creation from the interaction between consumers, suppliers, and products that has led to unprecedented valuation methods and transformed business strategy/valuation as it’s known. Network effects have given birth to metrics used in the investment world centered around product and market (cohort metrics such as retention and churn, CAC:LTV, multi-tenanting, etc.) rather than traditional financials and drive unprecedented amounts of investor interest as the platform companies’ success plays out its proof-of-concept. Giants and startups today are even aiming to become platform companies by baking their product/business strategy in order to get exposed to network effects.

Unfortunately, since the concept has become muddled by its various proclaimed adaptations, Here’s a breakdown of the established definitions. Network Effects can be split into four different types: Direct, Indirect, Two-sided, and Data. It is a non-binary concept; A company/product can have a combination of up to all four, or none of them.

Direct Network Effects - The value of a service simply goes up as the number of users goes up, but more specifically, the value of the network increases with each new user, for each other user there. This is also known as same-side effects and has been around for a long time already. Think about any closed-loop service, like telecommunication. The phone is only useful if other people also use the phone. Now if you add one more person to this network, both users can call each other and the value of the platform goes up. For each new user you add to this network, the possibility of calls between the users in the network shoots up and the phone becomes that much more valuable for each user. Additionally, each new user can add their own value to the network. For example, adding 911 emergency services as a user in this network makes the network dramatically more valuable to each user than adding another regular user. Direct network effects are also clearly present in social media platforms such as Twitter, Instagram, Reddit. At the most basic, Reddit is a platform for connecting users in ever-changing discussions. The more users that join, the more discussion topics that are posted, and the more responses those threads receive. As the users generate more content in the platform, the more valuable the platform becomes.

Indirect Network Effects -  increased usage of the product spawns the production of increasingly valuable complementary goods, which results in an increase in the value of the original product. A good example of systems that benefit from indirect network effects is Operating Systems (OS). While blackberry sold millions of copies with almost the same functions, Apple and Android users had access to thousands and millions of apps that were created by 3rd party developers. With each new user on an OS platform, there was an incentive for developers to produce apps, which in turn, increased the value of the OS (IOS, Android). More developers - more applications. More applications - more customers. More customers - more developers. And the loop continues. Now, blackberry is essentially irrelevant and the smartphone user market has exploded in size and value thanks to IOS and Android platforms utilizing indirect network effects.

Two-sided Network Effects -  Increases in usage by one set of users increases the value of a complementary product to another distinct set of users, and vice versa. Some great examples include intermediary platforms like Airbnb and Uber. Here’s a great graphic that explains uber’s network effects:

Uber, as shown, has a classic two-sided network effect. The more users on the platform, regardless of ride-hailers or drivers, the higher the value of the platform. Why? More drivers means more coverage/saturation. This gives for supply liquidity that results in faster pickups, less driver downtime, and lower prices. All of these benefits result in more users, which in return results in incentive for drivers to join the platform and meet this burgeoning demand. The coolest part about two-sided network effects (in my opinion) is that the interaction between both sides also has its own dynamics that can further incentivize participation and even business opportunity. The facilitation of transactions between customers and suppliers can be customized and manipulated on the platform. For example, users may want more premium options, resulting in the concept of new services such as Uber black and luxury Airbnbs, or even serving more affordable options such as Uber pool and shared/hostel Airbnbs. Ratings and status systems are in play with these platforms, with hosts/drivers getting their own ratings and titles (such as super hosts or 5-star drivers) that help attract more customers and incentivize hosts to stay on the platform and continue improving their services (the same concept applies with Uber drivers). When there may be friction between supply and demand, the platforms can step in by incentivizing certain sides and playing with pricing models in order to help maintain liquidity. A good example could be if the demand was overbearing supply in a location or if the supply was not in well-serviced geography, Uber and Airbnb could price-surge in order to incentivize supply to quickly meet that demand and direct servicing to where it is needed most. Two-sided networks not only provide value/opportunity to each side but the platform as a whole.

Data Network Effects - Data network effects occur when your product, generally powered by machine learning, becomes smarter as it gets more data from your users. In other words: the more users use your product, the more data they contribute; the more data they contribute, the smarter your product becomes (which can mean anything from core performance improvements to predictions, recommendations, personalization, etc.); the smarter your product is, the better it serves your users and the more likely they are to come back often and contribute more data – and so on and so forth. Over time, your business becomes deeply and increasingly entrenched, as nobody can serve users as well. Google is a great example of data network effects: the more people search, the more data they provide, enabling Google to constantly refine and improve its core performance, as well as personalize the user experience. Not only does this help Google develop itself as the most dominant and well-known product in its market, it enables Google to collect countless metrics and build datasets that are extremely valuable market commodities for monetization and product development. Datasets like these can be baked into more granular levels than just the core business, for example: recommendation engines that are now everywhere from Amazon (products you’ll want to buy) to Netflix (movies you’ll want to watch) to LinkedIn (people you’ll want to connect with), which all keep getting better with more users/data. The key here is that the process is automated within the product. Traditional businesses “learn” from data, which is typically done through human analytics and C-suite strategy shot calls, but the more you automate this process to refine itself, the more the presence of data network effects.

While all of these flavors of network effects are increasingly present in business models today, it’s also important to be able to fish out the models that aren’t actually network effects. Lots of great companies today are spending unreal amounts of money to get some sorts of network effects going within their products. It’s important to note that if you haven’t tipped into organic growth, there are no network effects -- and sniffing this out can be tough. Companies with a heavy marketing spend may show user growth, but it’s difficult to determine what the ratio of paid users (the ones you spend to acquire) to organic growth (the ones you don’t spend to acquire) is. Spending to acquire customers is by no means a bad thing, but it’s very important to look at attributes such as the timeline and targets for paid customer acquisition. You want to make sure that you aren’t spending too much for too long on a market that won’t help you ultimately feed into network effects. It’s incredibly important to understand how users will stay and be incentivized to continue contributing to the platform, and if it’s even the right market of users to be creating a platform with.

While network effects are great, it definitely comes with its vices. Oftentimes, businesses are prone to reverse network effects, where the value of each additional user might decrease the value of the platform if not managed properly. The connections, content, and clout generated by the users must be managed in some sort of way in order to ensure frictionless growth. If you suddenly have a million posts on your platform and a lack of content organization, lack of user regulation, subpar search engine, etc. you may fall victim to the traffic generated on your platform. In summer 2015, the Reddit community became increasingly unhappy with the censorship and questionable moderation of popular subreddits. This came to a head with Reddit execs and they fired Victoria Taylor, a manager that helped coordinate the sites extremely popular AMAs, as well as Ellen Pao, the CEO. In response, many users and subreddit administrators turned their accounts to private, effectively making huge swaths of the site's content unavailable. Dating sites and anonymous interaction platforms were plagued with fake profiles, bots, and creepy naked men. Managing monetization can also be an issue, oftentimes platforms may overwhelm its users with advertisements, solicited messages, and attempts to force users into giving more data. Promoted content can disrupt the amount of organic content and leave users unhappy with an oversaturation of monetization methods and lead them to leave the platform or even transfer over to another platform with better security, regulation, organization, etc.

On the opposite end of regulatory intervention, it’s incredibly important to understand what critical mass (revenue/size target at which fundamental changes can occur in a firm, and where self-sustaining market participation is possible) may look like for each network in order to understand scalability, regulation, go-to-market strategies, and sales/revenue models of platforms working with network effects. Thanks to the valuation of network effects being loosely adapted from Metcalfe’s Law (the value/effect of a network is proportional to the square of the number of nodes, or users, in the network (n^2)), institutional investors have prioritized market dominance in the name of greater delayed profitability or value of the business. Currently, VC dollars are taking billions in losses each year in order to scale market capture for unprofitable ventures in cutthroat markets, rapidly changing the way businesses are thinking about sustainability and liquidity. If you choose to monetize or regulate your platform too early or at the wrong timing, it could be fatal to your platform.

Network effects don’t simply happen. They are most often the result of carefully building and fueling a feedback loop. They can be extremely nuanced and conditional, especially since its a non-binary concept, but its presence is key to driving the success of almost all platform businesses today. Hopefully, this did a decent job of explaining what network effects are, and I’ll be diving into the application of it in business-strategy and giving a take on what it means for the modern VC scene in a future post.