Anti-network effect: Scaling up as the biggest threat to social networks

Sanget Paul Chodali, Sangeet Paul Choudary, a blogger on the Next Web, published in Saturday, said that while the development of today's web platforms is benefiting from the effects of the network, if you focus on the number of users and the size of the network, It is possible to fall into the state of extremes, trapped in the "anti-network effect", thereby sheng extremely and decline.

The following is the full text of the article:

For startups eager for excess returns, the net effect is the magic weapon of the dream. On the platform with network effect, the more users, the greater the value to users. Companies such as Facebook, Twitter, LinkedIn, YouTube and Skype all benefit.

2012 is a memorable year, for the first time there is a platform to connect 1 billion of users together. But as the online network reaches an unprecedented scale, many people are questioning its continuity, even fearing that such networks are too large and worthless.

To explore the future of online network, we must understand the relationship between network effect and value, and the factors that lead to the reaction of network effect.

Network effect and value

Under the influence of network effect, there is a strong correlation between scale and enterprise value. The larger the scale, the greater the value created for the user, which in turn attracts other users and expands further. Thanks to this "strong stronger" characteristic, once the network effect appears, the network scale will expand rapidly.

The value of the network has three main sources: relationship, content and influence.

Relationship: Through the network, users can find and establish relationships with other users. As more and more users join the network, the value that each individual receives will increase. The use of Skype and WhatsApp will expand as user relationships grow, as do Match.com and LinkedIn.

Content: Users can exploit and consume the value created by other users in the network. As users increase, the size of the content will be expanded to bring more value to users. Content platforms such as YouTube, Flickr, and Quora, as well as market-like companies such as Airbnb and Etsy, are increasingly useful because of the increase in creators and content.

Influence: Some networks have powerful users who have influence in the network. This powerful user can be found through Twitter's number of fans, Foursquare and Yahoo's reputation platform for questions and answers. In a network like Twitter, the larger the network, the greater the influence of powerful users.

Based on these three sources of value, larger networks can offer greater value in the following ways:

1, increase the potential relationship between users.

2, expand the relevant amount of potential content.

3, for strong users to provide more potential fans, enhance their influence.

In most networks, there are more than one source of user value. For example, Facebook initially created value through relationships, but as news feeds came along, content became a core driver of value. Recently, the introduction of subscription features has also increased the influence of some Facebook users.

Why is the network effect counterproductive?

In this view, it seems that the larger the network, the value of creating for users is bound to be more.

However, as the scale of the network expands, the value may decrease, for the following reasons:

1, Relationship: As new users join the network community, the quality of interaction may be reduced, spam and noise will increase.

2, content: After the increase in content, the network may not be able to deal with, lax management, thereby reducing the degree of personalization of content.

3, Influence: The network may inadvertently bias to the old users, to provide them with more resources, and ignore the new users.

The network effect creates the cycle of "strong stronger" and promotes the rapid growth of the network. In the same way, the "anti-network effect" can also go in the opposite direction, leading to a massive loss of users, Friendster, MySpace and Orkut are excellent examples.

Anti-network effect: Relationship

Relationship-oriented networks create value through interpersonal relationships, such as dating web match.com and professional social networking LinkedIn.

If new users reduce the value of existing users, these networks will suffer from the reverse network effect. To prevent this, there must be some resistance, such as setting a threshold for entry, or making it harder for users to establish relationships with others.

In dating sites, as the size of the community expands, female users often complain that more and more people are out of the box and eventually leave. Cupidcurated and other websites have taken some steps to address the problem. For example, they would examine male members in a way that is similar to a single bar restriction.

LinkedIn has also created its own "hurdles": banning users from communicating with people who are too far off. This ensures that users do not receive abrupt information. With this approach, LinkedIn also promotes barrier-free communication services OpenMail as a premium product.

In contrast, anonymous video chat tool Chatroulette lacks such "hurdles". As a result, the Web site experienced an "anti-network effect" after the initial boom, as anonymous naked men entered the network, adding to the discord and forcing the sheer number of users to evacuate.

Many dating sites and social networks like Orkut have experienced similar declines after reaching a certain size, mostly because of the noise generated by bogus accounts.

In a word, it is necessary to create the right threshold of network entry in order to achieve benign expansion of the relationship network.

Anti-network effect: Content

In content networks such as YouTube and Flickr, the larger the network, the more content creators often have to provide users with more content to consume. Networks such as Facebook and Twitter are content networks, in addition to relational networks.

In order to encourage user participation and scale up as quickly as possible, most content networks have low thresholds. To ensure content relevance and value, this network requires strong content governance and personalization of content for the user experience.

If the content management system is ineffective, the reverse network effect will appear. The relevance of content to users should not be reduced as content increases.

Content governance

Content networks create a set of content governance mechanisms through manual audits, computer algorithms, and community tools, including voting, rating, and reporting. YouTube votes, Facebook tags and Yelp ratings are excellent examples of governance tools.

Content governance mechanisms often fail as content increases. When the governance algorithm and audit mechanism can not adapt to the increase of content, the noise in the system will increase, triggering the reverse network effect, forcing the user to lose.

Quora has a strong content governance mechanism and benefits from a group of early adopters who are proficient in technology. As the size of the Quora expands, many people worry that the ability of less-able users to enter the system may increase noise, resulting in a rapid loss of value for existing users. It remains to be seen whether the audit mechanism will continue to function when the network is open to a wider user base.

Content Personalization

Content networks need to provide users with a personalized consumer experience and provide them with relevant content. The news feed stream of Facebook or Quora and the recommendation system of YouTube are good examples.

If the relevance of the content consumption experience cannot be maintained with the enlargement of scale, it is possible to have the reverse network effect.

Facebook's user experience revolves around news feeds. However, the content of the news feed is messy because Facebook has no threshold to share, which can lead to a reduction in the user experience as the scale of the network expands. This may be due to several factors:

1. When users increase their friends without any choice, the News feed flow will fill up irrelevant content.

2. When marketers and application developers can push information to the news feed, the noise increases.

3. When a network such as Facebook and Twitter deploys revenue-generating schemes such as promoted posts or promoted tweets, the noise level is further increased by the relevance of advertising content less than organic content.

With the enlargement of the scale, the content network is confronted with the threat of the anti-network effect. Facebook's biggest weakness in the expansion of the web is the poor signal-to-noise ratio in the news feed, not the overall size of the network.

Anti-network effect: influence

Influence networks have a system that distinguishes powerful users from ordinary users. Twitter, Quora and Quibb all have increased influence through one-way attention patterns. Active users are competing for a greater say in the use of such networks.

Influence networks tend to favour old users. Influence stems from what users create, and older users have more time to create content and attract fans than new users.

Ironically, this could worsen if you look at the high signal-to-noise ratio. Twitter always recommends superuser to potential fans, because they create content that may be better. As a result, the platform itself in the strong user and ordinary user differentiation process also played a role.

Late-joining users found it more difficult to attract fans, so it was possible to deactivate the network. These networks need a mechanism to ensure that new users have equal exposure to build their network influence. For example, 500px designed "best content" and "latest content" two lists to provide more opportunities for users with low exposure.

Anti-network effects often lead to the rapid decline of hit's large networks. With the expansion of the network scale, the higher signal-to-noise ratio becomes the important factor to maintain the network value. In fact, the network can both expand and retain users as long as they do the following:

1, the establishment of appropriate network use barriers to avoid abuse.

2, the establishment of a strong content management system, and the expansion of the scale of the network to adapt.

3, create a highly relevant and personalized user experience.

4, through the democratic model to allow users to gain a fair impact.

Networks with the above advantages have already achieved benign expansion. In a world of unprecedented network size, focusing on maintaining a high signal-to-noise ratio will help these networks continue to maintain their value and network effects in the growth process.