TikTok and the Sorting Hat

Why TikTok's algorithm is more disruptive than you think

As a reminder, I’m Eugene Wei, a former product guy at companies like Amazon, Hulu, Flipboard, and Oculus, and I’ve kept a personal blog Remains of the Day since 2001, offering what I describe as “Bespoke observations, 80% fat-free”, though the perceived fat ratio will vary depending on your tastes in technology, media, and all the other random topics I cover.

---

Apologies for the long hiatus from posting here. If you’re like me, you’re probably buried in unread newsletters anyhow. I’ve been in a prolonged state of mourning for many things, not the least the state of the world, and I’ve been booked on back to back to back consulting engagements. I haven’t stopped writing, but a lot of it has been private client work.

This one is a short update, simply to point to my new blog post TikTok and the Sorting Hat. I’d post it in its entirety here but it includes copious sidenotes that can only be displayed on my blog.

In the post, which I think will be the first and longest of roughly three pieces on TikTok, I don’t address the proposed divestment or ban of TikTok, nor do I cover the geopolitical or security issues involved. There’s plenty of writing on that already, and I don’t have anything new to add.

Instead, I dig into why I find TikTok so fascinating and disruptive a company. Namely, it has an algorithm so perceptive and efficient that it allowed an app built by two Chinese guys in Shanghai (Musical.ly) to finally level up to TikTok and pierce the cultural veil in markets as diverse as the U.S., India, and the Middle East.

I have friends who work at Bytedance, and they’d be the first to admit, they don’t understand most of the video memes on TikTok in the U.S. Hell, I don’t understand some of them myself, and I was born and raised here. In the past, Chinese apps have really struggled to make inroads in the West. I can’t think of a single one that has made a real dent. The cultural differences, especially the language barriers, and the varying design aesthetics (often driven by the differences in the rendering Chinese versus English), have made international market expansion a real challenge for China.

But that’s the magic of TikTok’s algorithm: it abstracts away culture! Machine learning figures out who will like what videos, you don’t have to even understand what the videos are about. In doing so, TikTok bypasses all the hard work other apps have done to try to derive the interest graph.

Using a social graph as a way to reverse engineer the interest graph has always been a roundabout way of building it. But the interest graph is where almost all the money is: for targeted advertising, for e-commerce, for SVOD subscriptions, and so on. Knowing what people like is one of the fundamental ways to achieve product-market-fit. That algorithms can now do much of that work for you may not be new, but the way that TikTok has employed that, in conjunction with short video, is new and disruptive.

In TikTok, I see an alternate path to competing with tech giants like Facebook/Instagram, Twitter, Amazon, YouTube, et al. An algorithmically driven video-based attack that doesn’t try to beat them by being them but by attacking from an oblique angle.

There is some irony that at the very moment Chinese tech finally figured out a successful approach into the U.S. market, the Trump-led administration might lock them out. However, for potential purchasers of TikTok in the U.S., this is a once-in-a-generation opportunity to pick up an invaluable asset. The only surprise, to me, is that more companies aren’t rumored to be aggressively bidding. Microsoft is hardly the most logical destination, yet even they can recognize what a gem of a firesale this is.

If you’re interested, check out the piece and let me know what you think. I’ll ping this list again when my next entry comes out. I promise that the follow-up won’t take months.