Breaking Down Network Effects
Understanding the acquisition, engagement and economic effect
I was reading Andrew Chen’s Cold Start Problem recently, and enjoyed his framing on breaking down network effects based on the part of the product / business it helps in.
This week, I’ll be going deeper into it and sharing some of my own thoughts on it.
First, what are Network Effects? The basic concept is that:
The value of a product or service to any one user goes up when the number of people who use that product or service goes up.
While there are many ways of categorizing or breaking down network effects, such as whether it is direct or indirect or how it comes about (data, social, etc.), I liked the way Andrew Chen framed it in the Cold Start Problem based on the user problem / journey it helps the product with.
In the book, Chen talks about how Network effects aren’t one singular effect, but rather made up of three distinct forces:
The Acquisition Effect which allows products to tap into their network to drive growth in users more cheaply as the network grows
The Engagement Effect which involves user engagement going up as the network grows
The Economic Effect which involves monetization rates going up as the network grows.
Let’s go deeper on each one.
📈 Acquisition Effect
Acquisition effect refers to the idea that as the size of a company’s network grows, it becomes easier for the company to tap into that network to acquire the marginal user.
The way to think about the acquisition effect is that for companies that are able to tap into their network, as the size of the network goes up, the customer acquisition cost to acquire users goes down (until a certain point).
The best way to measure the acquisition effect is usually by the percent of organically acquired users and the customer acquisition cost going down over time (within a network until a certain point).
Companies that are able to exploit the acquisition effect are able to leverage their existing user base to grow in two key ways:
Bandwagoning: Venture Firm NFX describes bandwagoning as the idea that social pressure to join a network causes people to feel they don’t want to be left out. Certain products and services are able to create this effect around specific niches, which results in it becoming easier to acquire users as the network grows in that niche, because all the other users feel FOMO and join.
Viral network-driven growth: Chen mentions that the acquisition effect primarily comes from product-driven viral growth, where the product is able to build in experiences which encourage users to organically or for a benefit share the product with others and thus grow its userbase. Now, some people believe that the viral effect is completely different from a network effect, but in this case, Chen defines the Acquisition effect to include a subset of the initiatives which may result in virality that leverage the existing network (e.g., a viral ad campaign won’t count).
Referral programs which incentivize the existing users in the network to invite others.
Some companies that have used this well include Uber and Wealthfron which typically had benefits for both sides. Note that some people believe this isn’t really a “network effect”.
Natural Sharing flows in product which work for non-users: Products which naturally involve sharing something / using with others that gracefully work for people not on the product / service yet are able to tap into their network to expand.
Examples of this include being able to share Dropbox files / folder links with non-users which increases the likelihood they may join and P2P apps such as Venmo, Cash App which allow sending money to someone not using the product yet (and thus encouraging them to join)
Leveraging bandwagoning in invite flows: Many products allow users to invite other users, but the best way to leverage that network is to then create FOMO in the invited user which encourages them to sign up.
As an example, once a user has been invited by their team on Slack but not joined, by emailing them updates such as “you missed X chats this week” with the chats described in the email, Slack leverages that team’s activity to create FOMO / bandwagoning in the invited user, which can deepen as the network already on Slack relevant to that user grows. Other collaborative products (Asana, etc) or messaging products have also implemented some version of this.
If you don’t yet receive Tanay's newsletter in your email inbox, please join the 3,000+ subscribers who do:
🤝 Engagement Effect
Engagement effect is the typical thing that people tend to think of when they hear the term network effects.
It refers to the idea a product / services become more valuable for each user such that their engagement with it grows as the size of the network on it grows.
The engagement effect is best measured by looking at usage metrics such as time spent / posts made / transactions completed (or whatever the key atomic metric of engagement in the product is) and retention rates of various cohorts, with ideally later cohorts having higher retention rates off the bat, and older cohorts showing an uptick in retention rates over time (as the network grows and makes the product more valuable).
Generally, engagement effects tend to happen pretty naturally in many network effect type products.
For example, most messaging apps have a direct network effect. As more users are on it, it becomes more valuable to all of them.
Similarly, most social networks and marketplace type products have a mix of direct and indirect network effects, where as there are more people on the platform, each post / listing has more potential reach and similarly for consumers, there are more posts / listings that they can go through.
However, there are a few things companies can do to better exploit engagement effects.
Understand when the engagement effect kicks in for a given user: Facebook very famously found that its ‘aha moment’ for a user was getting them to 7 friends within 10 days, which made it much more likely that they retained on a platform. If companies are able to understand what exactly is needed for that user to become part of / get the benefit of the existing network, then they can drive towards that outcome.
Develop new use cases as the network develops: Sometimes, products start with one use case, but as they grow in size and a network around them develops, they can then leverage that network to develop new use cases. These new cases are both a way of leveraging the existing distribution / scale the company has reached and a way of further strengthening the network effects of the business and growing the engagement effect.
As an example, Slack started as a collaboration product within companies. Once they had scaled to a large number of companies, they also launched shared channels which allows people to use slack across companies. These strengthen the network further and grow engagement even more.
Optimize the engagement loop: Almost all network products have some kind of engagement loop, which refers to the steps taken in the network for users to derive value. Companies can strengthen engagement effects by analyzing each step in the process and optimizing it, which results in the loops being more effect.
For example, Chen writes: “For a social or communication product, the loop often starts with a content creator posting or sending content. The content is then sent to everyone they are connected to, and depending on the size of the network, they get a nice stream of likes and comments back.” The way to optimize this loop would be:
to make it as easy as possible for users to post content (templates, formats, etc.)
to make sure enough users see the content (ranking, etc.)
to encourage and make it simple for users to interact with that content (lightweight reactions, suggestions for comments, etc.)
Leverage the data network effect: The data network effect, as defined by NFX, is when a product’s value grows as a result of more usage via the accretion of data. The data network effect can help strengthen the engagement effect if a product is able to learn from the usage of its network to improve the product for all users. Generally, this plays out in the form of better ranking and recommendations.
TikTok is a great example of this, which gets better at ranking and recommending content as more users spend more time on it. Not only does an individual user spending more time improve their own recommendations, but in aggregate each individual user is also helping improve recommendations for everybody else on the platform.
💰 Economic Effect
The economic effect refers to the ability of networked products to improve their per user monetization rates / profitability as the number of users grow.
It’s important to distinguish between benefits that a network may get as it grows purely due to economies of scale (which is true for almost all businesses) vs the benefits to it which accrue because of network effects. For example, many fixed costs or subsidies (cited in Cold Start Problem as an Economic Effect) can often be spread over a larger audience as the network grows, but those are more economies of scale than a real economic network effect.
Typically, a true economic effect is usually seen through average revenue per user going up (even adjusted for engagement) due to increased conversion rates largely due to the data network effect.
Network-driven products / companies can do a few things to take advantage of economic effects:
Leverage the data network effect they have to improve conversion rates: As the network grows, companies typically have more data which they can use for personalization, targeting and recommendations to improve the conversion along the key way the company monetizes.
For example, FB and Google among others, improve the quality of their targeting and personalization of ads as the user base grows because they have more data and usage to train their models on. This improves the CPM of ads meaning that even per a fixed number of ad impressions, users monetize at higher rates.
Similarly, marketplaces can improve their recommendation algorithms (consider Amazon’s products you may buy) which allows them to drive up average order value and improve revenue per user.
Design monetization such that the economic effect flows from the engagement effect: Companies can think carefully about their monetization practices and approaches and ensure that as the engagement effect kicks in, it also leads to the economic effect kicking in via improvement in their monetization. This involves having some form of a usage-based monetization or thought-through tiers where the paywalled features get more important as the network grows.
For example, any usage-based monetization model essentially directly benefits from higher engagement leading to better economics. For example, advertising-based businesses (more engagement means more ad impressions) and marketplaces which monetize via a take rate of all transactions.
Similarly, collaborative workplace products can be clever about their subscription tiers such that as the size of the network grows, it becomes more critical to be on higher tiers to function properly. Slack has done a good job with this, for example requiring a premium tier for searching through messages
That’s all for this week! If you enjoyed this piece, I highly recommend Andrew Chen’s book The Cold Start Problem which goes into this in much more depth and has many more different examples.
Thanks for reading! If you liked this post, give it a heart up above to help others find it or share it with your friends.
If you have any comments or thoughts, feel free to tweet at me.
If you’re not a subscriber, you can subscribe below. I write about things related to technology and business once a week on Mondays.