• December 1, 2022

Disappointing Nowcasts For Upcoming Inflation Highlight Fed’s Concerns

U.S. Federal Reserve Chair Jerome Powell indicated that rate hikes may be ending in early 2023, but he still worries about inflation. Though we’ve seen encouraging inflation data for the month …

Student Loan Forgiveness To Remain Blocked, Says Supreme Court, As Justices Prepare To Consider Arguments

The Supreme Court on Thursday allowed a nationwide injunction blocking President Biden’s signature student loan forgiveness program to remain in place, while signalling that it will consider legal arguments regarding the …

CFDA Turns 60 With Sandbox Metaverse Exhibit, Bored Ape NFT Dress & More

To celebrate its sixtieth anniversary, The Council of Fashion Designers of America is making its metaverse debut with an exhibition in The Sandbox alongside a collection of commemorative NFTs. The seven …

Multi-sided platforms like Twitter create value for each user by adding more users. However, some users are more impactful than others. While former President Trump would probably increase Twitter’s activity, audience, and market value, he could also hold Twitter hostage.

[Author’s note: this article is not about politics. I am an economics professor, focusing on multi-sided platform marketplaces. This article is about economic theory.]

Let’s start with a comparison between YouTube and Twitter. Beyonce’s music video, Single Ladies, has been viewed 8,880,000 times on YouTube. Clearly, many people have derived great pleasure and value from watching it. And that’s the way that platforms work. More suppliers (for YouTube, these are content producers) provide a wider variety of goods and services which, in turn, attract more buyers (for YouTube, these are viewers). The aggregation of buyers then attracts more supplies. This is labeled the network effect, where the value of the platform for each person is directly related to the number of other people on the platform. The network effect is responsible for the accelerated, exponential growth that many platforms have achieved in the last 10 years.

In its simplest form, the network effect treats all users the same. But clearly they are not. Some add more value – as suppliers or buyers – than others. In an article in the Harvard Business Review in 2019 entitled “Why Some Platforms Thrive and Others Don’t”, Feng Zhu and Marco Iansiti distinguished between a weak network effect and a strong one, and their conclusions are surprising.


A mature platform company with a strong network effect has many users who make roughly equal contributions. If any single buyer or supplier depart, the value for all of the other users is not dramatically reduced. Despite Beyonce’s contributions, YouTube has thousands of contributors with accumulated views in the millions.

A weak network effect, on the hand, has a few super-stars that contribute more than others. If one of them leaves, the decline in value for each user could be significant. Just as the network effect can fuel rapid growth, it can also herald rapid decline. So the departure of a few super-stars might cause the marketplace to collapse as the herd departs.

Twitter is not yet mature. It does not have stable revenues or costs, and has not yet made a profit. This puts Elon Musk, Twitter’s new CEO, in a bind. If he allows a popular figure onto the platform who would attract lots of attention (positive or negative) like former President Trump, he will boost activity and audience growth. But he will also move from a strong network effect to a weak network effect, beholden to these few major attractions. If they depart, the audience – and platform revenues – will shrink. This reality might also complicate decisions to eject a major contributor (again) from the platform for fear of market collapse.

As of yesterday, Mr. Musk elected to allow Mr. Trump back onto Twitter. And as of today, Mr. Trump has yet to participate with a tweet. The next chapter will be instructive for platform entrepreneurs and for platform theory.


Leave a Reply

Your email address will not be published.