The TikTokization of Everything
This is a weekly newsletter about how tech and culture intersect. To receive Digital Native in your inbox each week, subscribe here:
The TikTokization of Everything
Over the past decade, two primary forces have powered technology: mobile and cloud.
Mobile facilitated the rise of massive consumer internet companies: Uber and Lyft, Instagram and Snap, Robinhood and Coinbase. Each was founded between 2009 and 2013. Digital advertising rapidly shifted to mobile in the 2010s, and desktop-era companies like Facebook had to scramble to reinvent their businesses.
Cloud, for its part, underpinned an explosion in software-as-a-service (SaaS) and enabled data to become the most prized resource in a business (“Data is the new oil” and all that). Emergent companies—again, each founded between 2009 and 2013—included Slack and Airtable, Stripe and Plaid, Snowflake and Databricks.
The percentage of corporate data stored in the cloud doubled from 2015 to 2022:
Few charts are more impressive than that of Amazon Web Services revenue over the past decade (with 35% profit margins to boot!):
Mobile and cloud made the 2010s a very, very good decade in technology. But over the past few years, we’ve seen a lot of clamoring for what comes next. Virtual reality? Augmented reality? Autonomous vehicles? Crypto? Web3?
Each is interesting for distinct reasons and within distinct use cases—but each is also very, very early. The entire VR industry is equivalent to just 2% of Apple’s 2021 revenue. Much of the hype surrounding new technologies and vaunted “platform shifts” derives from anxiety around mobile and cloud being…old. AWS launched in 2006; the iPhone came out in 2007. Neither mobile nor cloud are saturated, but they aren’t as ripe for greenfield opportunity as they once were. At the same time, we’ve seen an unprecedented influx of private capital chasing startups:
Perhaps the most compelling—and the most likely—force to power tech in the 2020s is artificial intelligence. AI has improved dramatically within the past few years. Until recently, Gmail’s auto-complete sentence feature was almost unusable; now it’s scarily good. Facebook users will recognize how good AI has become at identifying friends in your photos; Facebook’s DeepFace engine is now actually better at facial recognition than humans.
What we’re seeing is AI emerge from the infrastructure layer to the application layer—the parts of technology that everyday people interact with. The breakthrough use case in recent months has been text-to-image generative art, which I wrote about earlier this month in When Art and Technology Collide. Here’s the Stable Diffusion output for—
A distant futuristic city full of tall buildings inside a huge transparent glass dome, In the middle of a barren desert full of large dunes, Sun rays, Artstation, Dark sky full of stars with a shiny sun, Massive scale, Fog, Highly detailed, Cinematic, Colorful
Incredible.
But perhaps no use of AI is more visible to more people than TikTok’s For You Page. TikTok pioneered the immersive, algorithmic For You Page to curate content…well, for you. Each posted video is pushed to an initial set of viewers, then evaluated based on how those viewers respond to it—how long they watch, if they like the video, if they comment on it, and so on. If viewers respond well, the video is pushed to even more viewers, and the cycle continues.
Last week, I was talking with my colleague Martin about what’s coming next in technology after mobile and cloud. We were chatting about AI, and thought back to the famous “Unbundling of Craigslist” graphics from a decade ago.
Here’s a detailed view of Craigslist unbundling:
Here’s a cleaner view of the same concept:
The basic premise of these graphics was that major categories were being reinvented by more focused, better products. Often, the disruptor leveraged a new technology. Tinder, for instance, was one of the first mobile-only dating apps.
A similar reckoning may come from AI applications. Major categories—dating, real estate, job searches—may find themselves completely upended by better use of artificial intelligence. Why swipe endlessly on Tinder when AI can surface your perfect match? A decade from now, we could be looking at a version of the graphic above with a completely different set of logos—AI-first companies that reimagined each category.
Let’s look at an example—commerce.
SHEIN, in many ways, is TikTok’s sister company. SHEIN and Bytedance (TikTok’s parent company) are both Chinese companies and are two of the world’s three most-valuable startups (America’s SpaceX slots in at #2 to break up #1 Bytedance and #3 SHEIN).
Just as TikTok has infiltrated U.S. media, SHEIN has infiltrated U.S. fast fashion—
Here’s a different view, comparing SHEIN to H&M and Zara sales:
SHEIN’s explosion is nothing short of remarkable: SHEIN has grown over 100% every year for eight straight years (!), and its latest private market valuation makes it worth more than Zara and H&M combined. In June, SHEIN dethroned Amazon as the No. 1 shopping app in the iOS and Android app stores.
SHEIN’s velocity is something to behold: 8,000 new items are added to SHEIN every day, while Zara adds 500 every week. SHEIN is basically an internet-native reincarnation of Zara and H&M, leveraging better technology to squeeze three week design-to-production timelines into three days. SHEIN combs competitor’s websites and Google Trends to figure out what’s in style, then creates designs quickly, forecasts demand, and adjusts inventory in real-time.
To bring us back to AI, one aspect of SHEIN that’s impressed me is its recommendations. Just as Bytedance anticipates the content you’ll want to watch, SHEIN anticipates the clothes you’ll want to buy. SHEIN is to commerce what Bytedance is to content.
Over the weekend, I went online shopping for a friend’s upcoming 30th birthday party. The party is Euphoria-themed, meaning you basically come dressed like Maddie or Cassie or Nate Jacobs from the HBO show. I’ve never shopped on SHEIN before, but I typed in “men’s mesh black top” to look for a shirt. Then I clicked on the “Pants” category and was met with this screen:
Based only on a single search for that mesh top, SHEIN was able to anticipate pants that were very much the same style and theme. Impressive. (Also, please don’t assume these are the clothes I normally wear.)
In some ways, this is a more sophisticated version of the concept Stitch Fix pioneered with its personal styling subscription boxes. Stitch Fix had humans in the loop, but also leveraged data science based on a lengthy onboarding questionnaire. SHEIN, meanwhile, made spot-on recommendations based on only four words I typed (and likely a lot of data around what I clicked on, where my mouse hovered, and so on).
Stitch Fix’s personal styling market has proven relatively niche, and the stock has gotten clobbered. Active clients are down to 3.9 million, a 200,000 year-over-year drop (down 5%). The company is pivoting hard to its Freestyle product—a more normal shopping experience—but that segment is still a small portion of the business.
Though Stitch Fix is struggling, it was on to something groundbreaking—personalized commerce. The company just arrived at the concept a few years too early, when AI wasn’t yet sophisticated enough to take the place of a lengthy questionnaire and small army of data scientists. SHEIN is a step in the right direction, but we’re still only at the cusp of AI-driven recommendations.
Imagine a company that combs your camera roll and—with stunning accuracy—recommends a new wardrobe for you. Or the company simply asks you to link your Instagram account, and then it digests every like and follow you’ve ever made to deliver incredibly accurate, incredibly personalized fashion recommendations.
The major consumer applications of AI will lean heavily into sophisticated recommendations that anticipate your wants and desires before you even know them—in the same way that the TikTok For You Page has shown people that they’re gay before they themselves have even arrived at that realization. Perhaps the example company above reinvents commerce in an FYP feed, allowing you to only browse one highly-curated item at a time—double-tap to buy, swipe up to see the next item.
Today’s AI applications are still rudimentary to those we’ll see in the coming years. Think of every Craigslist category above—education, books, home decor. Each one is ripe for reinvention.
Sources & Additional Reading
How TikTok Holds Our Attention | Jia Tolentino, The New Yorker
A deep-dive into SHEIN | Green Is the New Black
Thank you to my partner Martin Mignot for being a thought partner on this
Thanks for reading! Subscribe here to receive Digital Native in your inbox each week:
