The 16 Types of Network Effects

One of the most important topics for startups, which doesn't not get enough attention in MENA, is the Network Effect. It is one of the strongest defensive methods for companies and represents 70% of the value of technology companies worldwide from the 1990s until now. That's why I decided to write extensively about it, with important references. The principles in this article can be a complete game-changer for founders and product managers who want to be in a different league.

Table of Content

• Physical (e.g. landline telephone)
• Protocol (e.g. Ethernet & Blockchain)
• Personal Utility (e.g. WhatsApp)
• Personal Networks (e.g. Facebook)
• Market Networks (e.g. AngelList)
• Hub-and-Spoke (e.g. TikTok & Medium)
• Marketplaces (e.g. Amazon)
• Delayed Value Marketplaces
• Asymptotic Marketplaces (e.g. Uber & Careem)
• Platforms (e.g. OS)
• Expertise (e.g. WordPress & Figma)
• Data (e.g. Google Maps)
• Tech Performance (e.g. BitTorrent)
• Language (e.g. Google)
• Belief (e.g. currencies)
• Bandwagon & Tribal (e.g. Apple & Harvard)

In the world of digital products, there are four families of defenses which are the ways that enhance the strength of your product or company against competitors or other alternatives. These four families are Network Effect, Scale, Embedding, and Brand. And yes, Scale is not the same as the Network Effect, as most people confuse them.

The Network Effect is a mechanism that occurs when each new user increases the value of the product/service for the other users in the network. Imagine the first person in the world who used the telephone. What is the value of the phone for him? There is no value because there are no other parties to talk to. But with the second user, there is value because he can talk to another person, and after a thousand users, there is more value because there are 999 other people to talk to, and so on. The most popular modern examples of this mechanism are Marketplaces like Amazon, where the network consists of sellers and buyers. With each new seller, the network gets value with product diversity, and each new buyer adds value by attracting more sellers.

What is a network anyways?

The network is made up of links and nodes, where the nodes represent the entities that are part of the network, and the links represent the relationships, transactions, and interactions between those nodes. These transactions can be money, information, communication, or anything else that can be transferred.

Types of Network Effects:

We can divide Network Effects into 16 types from strongest to weakest as follows:

Direct Network Effects:

1. Physical (e.g. landline telephone):

This refers to any tangible network, such as cables, electricity, roads, trains, water, and gas. This is the strongest type because you can build upon it to create other defenses, and competing with this type requires a large amount of capital. This type is so powerful that many service providers who have this effect can offer a very poor service and still dominate the market. Most networks of this type are public utilities and tend to be monopolized, and the most popular example of this type of Network Effect can be found in companies like KarmSolar.

2. Protocol (e.g. Ethernet & Blockchain):

In the field of communications, a protocol is the agreed-upon method and standard for communication. For example, the Ethernet protocol is responsible for communication between different digital devices in a network, and the blockchain protocol is responsible for tracking and verifying transactions. A protocol like SWIFT in the banking system regulates global banking transactions. This type of Effect is considered the second most powerful type, and other defenses like Embedding can be built upon it, which we will discuss later.

3. Personal Utility (e.g. WhatsApp):

This refers to any tool related to a person's identity that they use permanently for vital purposes and to communicate with other parties in the network. The absence of a person on the network can represent a significant obstacle to their daily transactions. The most common examples of this type are WhatsApp or SMS.

4. Personal Networks (e.g. Facebook):

These networks allow individuals to interact with others using their real identities in order to communicate with others. They are an extension of personal relationships in real life. Each new member adds value to the network as they may potentially connect with other members and also provide new content. The main difference between personal networks and personal tools is that tools are used for a specific purpose, while networks are less important and can be temporarily stopped. They are used for general communication. Examples of this type of network include Facebook, LinkedIn, and Twitter.

5. Market Networks (e.g. AngelList):

Market networks are a combination of personal networks and business transactions. They are networks of professionals who exchange complex services and often provide SaaS tools for members of the network to use. One of the most well-known examples of market networks is AngelList, which connects startups and investors. Market networks are similar to marketplaces, but they provide real identities for professionals to conduct their businesses and focus on creating long-term relationships between network members.

This type of network combines the power of personal networks, such as Facebook and LinkedIn, with a profitable and highly liquid business model, such as SaaS. Market networks are a relatively unexplored area in Egypt and the Middle East.

6. Hub-and-Spoke (e.g. TikTok & Medium):

This type occurs when the sub-networks provide their content to a central point, and then the central point redistributes the content to other nodes, relying on internal algorithms for distribution. This type creates an incentive for sub-networks to provide the highest possible quality of content so that their content can be redistributed. TikTok and Medium are popular examples of networks that apply this type.

2-Sided Network Effects:

7. Marketplaces (e.g. Amazon):

Marketplaces have two types of users, supply and demand or seller and buyer. Each party creates complementary value to the other party in the network. Every new seller adds value to the network through product diversity, and every new buyer adds value to the network by attracting more sellers. To break the network effect, a new competitor needs to offer higher value to both the seller and the buyer at the same time. This type derives most of its value from the network itself, not from the product.

One of the biggest drawbacks of marketplaces is Multi-Tenanting, which means that network parties can use multiple alternatives at the same time. The same seller can sell on more than one platform, and the same buyer can compare and buy from more than one platform. It is difficult to prevent new competitors, and the only solution is to keep offering higher value to network parties.

8. Delayed Value Marketplaces:

This is a different form of marketplaces that require a very large number of supply inputs before creating any real value for demand. For example, Talabat requires a large number of restaurants to join before there is any real value to the customer. But once this critical mass is reached, the network becomes very powerful.

9. Asymptotic Marketplaces (e.g. Uber & Careem):

This is a third type of marketplaces that differs from the traditional type of markets in the "value curve" or how fast the supply will increase alongside with demand and vice versa. This type is the opposite of the previous type in that the existence of supply creates a very high value for demand at the beginning, but at a certain point of increasing supply, the value of demand start to diminish. For example, Uber, up to a certain point, having more drivers means more value for passengers because it reduces waiting time. However, after a certain point of increasing drivers, the value does not increase for passengers to the same extent, i.e. the passenger finds a trip in 3 minutes instead of 10 minutes is significant, but finding a trip in 2 minutes instead of 3 minutes is not that significant.

10. Platforms (e.g. OS):

Platforms have two types of users like Marketplaces - supply and demand, or developer and user - but the difference is that the supply is part of the platform itself and not a separate product. The most popular examples of this type of platform are operating systems like Windows, Android, and iOS.

Other Types of Networks:

11. Expertise (e.g. WordPress & Figma):

These are tools and products that require certain expertise to use, and the more complex or difficult the tool, the more valuable the network. They are often used by professionals such as designers, developers, and data analysts. So more companies use the most widely used tools and more professionals learn and use the most in-demand tools. Joining more professionals and building more expertise in these tools adds value to the entire network. The most popular examples of these tools are Salesforce, Google Analytics, WordPress, and others.

12. Data (e.g. Google Maps):

This refers to increasing the network value with an increase in data and increasing data with more network usage. Each new user adds value because they feed the network with data, and as data accumulates in the network, each user takes more value from the data. An example of this type is Google Maps, which increases in value when it has more data from users.

13. Tech Performance (e.g. BitTorrent):

This refers to how increasing the network's users contributes to improving the quality and performance of the network itself, making it faster, easier, or cheaper. The most popular example is BitTorrent, which increases its speed in sharing files with more users on the network.

Social Network Effects:

14. Language (e.g. Google):

Increasing the number of speakers of a certain language creates a higher value for the language as a whole. A language spoken by a billion people is more valuable from a network perspective than a language spoken by a million people. This applies not only to languages but also to dialects and jargons. The question is how to use this in business. The effect of language can be used in naming products and companies. The most popular example is Google. When someone says "Google it" to another person, it will be a strange thing to use Bing instead. Here, the language itself becomes a barrier.

15. Belief (e.g. currencies):

Increasing the value of a belief is done by increasing the number of believers. The most popular example is gold, whose basic value is not only derived from its rarity or uses, but also from collective agreement and belief in its value. The same principle can be applied to Bitcoin, currencies in general, and the stock market.

16. Bandwagon & Tribal (e.g. Apple & Harvard):

This refers to the societal pressure an individual may feel if they do not join the network. This effect often occurs when people use a product in the early stages. The most popular example is Apple users. If you are a programmer or designer and do not use Apple, Apple users will make you feel like you do not belong to their group.

Tribal is similar to Bandwagon, and refers to increasing the value of a network by increasing highly cohesive groups. This can occur in private communities such as schools, universities, and companies. Companies can benefit from this type in two ways: either by creating similar groups of employees by providing a better working environment, where employees recommend other employees to the company, or by creating similar groups of customers, like Harley-Davidson and Apple whose customers are closer to club members.

Conclusion

In the end, we may need to ask whether knowing these types of networks is important or if it's just theoretical. Does it suit all types and sizes of companies? The answer is that not all knowledge can be ignored. Every size of company may need different types and parts of the network's impact throughout its lifecycle. Some types of networks are less costly than others, and some types can be designed and implemented in early stages of the product or company's life-cycle.

In the second part of this series, we will learn how to build these networks and understand their characteristics such as network density, directionality, regularity, asymmetry, homogeneity, and how to control the speed of spread and avoid customer leakage from platforms. We will also learn how to overcome the chicken-and-egg problem in multi-sided markets and gain more insight into the different mathematical laws governing network growth.

Note: The original reference for this and future parts of the series can be found under the name "The Network Effects Bible". You may notice some differences as the content is not translated but adapted to better suit our region.