Microsoft has six weeks to complete one of the strangest deals in the history of the technology industry. On Sunday, the company announced publicly that it was in talks with President Trump to buy parts of TikTok from its Chinese parent company ByteDance. For months, Trump has been voicing national security concerns about TikTok and even threatening to ban the app. Now he presents Microsoft with a sale as a last chance to save her. If the deal comes, it would give Microsoft a new perspective on social networking and solve a variety of US national security concerns about the newly discovered popularity of TikTok.
But there is a problem at the core of the business that no one has addressed on either side ̵
In practice, Microsoft is likely to limit the deal to four countries because of the need to do so. The United States Foreign Investment Committee (CFIUS) empowers Trump to force ByteDance to sell its U.S. investments. Canada, Australia and New Zealand are the countries most closely linked to the United States on national security issues – best known as part of the Five Eyes Intelligence Sharing Network (excluding the United Kingdom) – and it’s likely that they’re on this list because of you are ready to take a similar step. When the United States expressed similar doubts about Huawei, Britain followed America’s leadership, but only under duress, and Europe still hasn’t made the leap. Given Trump’s confused justification and pariah status among other major democracies, this is likely as far as the US can expand its influence in this particular struggle.
Although the four-country approach makes sense for Trump, it is not clear that it makes sense for Microsoft. No one has yet acquired a regional part of a social network, and it will be more difficult to pull off Microsoft’s own part of TikTok than it looks. As TikTok employees keep reminding us, the app is in California, but the majority of users are still in Asia or Europe. The Indian TikTok ban slowed these numbers down, but Microsoft would still buy less than a third of the entire platform. We don’t know exactly which parts of TikTok’s activities Microsoft would include in the deal, but when it comes to separating TikTok from China, the company needs to rebuild any teams or infrastructure that are currently operated by ByteDance.
Essentially, TikTok is split into two apps (let’s call them MS-TikTok and BD-TikTok), with separate servers, a separate code base, and separate users. This would affect almost everyone who works with the app: advertisers would reach fewer users with a single ad purchase, influencers would have a smaller pool where they could go viral, and users would have less content to choose from. TikTok is successful when it can show users interesting posts, and without K-pop stans or African dance memes, it will be more difficult to accomplish. Even if users could share content from one network to another – for example, sharing a Turkish TikTok with a US account – it would require significant technical work that would only be more difficult as the two apps evolved on different tracks. Given the concerns about algorithmic propaganda, it is not clear whether such disclosure would be allowed at all.
Splitting the network would also be expensive. Even after the initial shock of rebuilding the ByteDance infrastructure, Microsoft will face a smaller audience and a smaller revenue pool. All investments in the platform – such as important work to develop new features or a new ad format – are spread across a much smaller pool of users, which means less money, less investment and less growth. This is the iron scaling law, the power that made Facebook as profitable as quickly as its user base grew. MS-TikTok would go the other way and take similar costs with far fewer users and far less revenue. This new version of TikTok would be far less profitable, and since BD-TikTok already has the market in the rest of the world, it is not clear how much scope there would be for scaling.
Of course, that’s not the only factor in business. As Tom Warren pointed out yesterday, Microsoft would have many advantages if it had a social network, even a small, regional one. The danger of CFIUS measures is really an exceptional circumstance, and it’s possible that ByteDance may be sold at fire retail prices, which makes it all worthwhile. There is so much we don’t know about the business and so much that remains to be determined. So it would be stupid to devalue it completely.
However, the proposed deal is based on an unprecedented type of business and technology divide. The Internet is inherently limitless, and most social networks use it to operate global empires from relatively regional headquarters. It is a monumental task to reduce the network back to national terms, with users seeing a different app in Korea than in Hawaii. In the best of circumstances, it would be difficult. But it seems absolutely impossible to do so on a hectic six-week schedule, fueled by a belligerent president and an ever-changing set of national security restrictions.