May 26, 2022 · 35m · another-podcast
Shein, TikTok and Netflix - thinking about limits
0:00 / 0:00
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Toni Cameron Brown and Benedict Evans examine the mechanics behind Shein's explosive growth, drawing cross-industry parallels with TikTok, Netflix, and Google to explore how algorithmic feedback loops and the elimination of physical retail constraints transform modern commerce and culture.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The hosts hold 78% of the talking time here. How this is scored →
The hosts as informed peer 7.2 Guest teaching 1.0 Guest disagreement 0.9 The hosts pushing back 1.4
speaking balance: gold is the hosts, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 21:05 Pushing on whether auto-generation is already live
Toni directly interrupts Benedict's hypothetical thought experiment on generative fashion to challenge whether algorithmic clothing design is already actively happening.
Hardest push from the hosts ▶ 29:45 Skeptical reframe of fashion echo chambersBenedict jokingly pushes back on social media panic assumptions and VC uniformity while reframing fashion trends as autonomous optimization feedback loops.
Biggest teaching moment ▶ 31:29 App SKU regional catalog comparisonToni checks live data from California showing 49,396 listed products, revealing distinct regional SKU differentials against Benedict's London app data.
The host holds their own ▶ 11:10 Historical analysis of WhatsApp and SMS inflection pointsBenedict demonstrates domain mastery by reciting exact historical message volumes and timeline milestones from WhatsApp's stealth rise alongside survey data.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The hosts as informed peer | Guest teaching | Guest disagreement | The hosts pushing back | Why |
|---|---|---|---|---|---|---|
| Fragmented Tech Landscape and Selecting a Meaningful Podcast Topic | 7 | 1 | 1 | 2 | Benedict synthesizes macro tech industry trends across Web3, AR/VR, gaming, and digital ad stacks to explain why finding fresh podcast topics is difficult. Toni collaborates closely, reflecting on their rare dilemma of needing multiple calls to pick a theme. | |
| Introducing Shein and the Evolution from ASOS Fast Fashion | 7 | 1 | 1 | 1 | Benedict introduces Shein's staggering catalog statistics compared to Zara and H&M, contextualizing its business model against historical online fast-fashion predecessors like ASOS. Toni adds relevant context regarding the transition from fast fashion to real-time fashion. | |
| Zero Retail Constraints and Predictive On-Demand Manufacturing | 7 | 1 | 0 | 1 | Benedict explains the mechanics of zero-inventory testing, predictive ordering, and real-time manufacturing batches. Toni enthusiastically builds on this concept to frame the overarching thesis about removing retail constraints. | |
| Shein's Explosive Ascent Compared to Skype and WhatsApp | 8 | 0 | 0 | 1 | Benedict demonstrates encyclopedic recall of tech history, comparing Shein's stealth growth curve to early metrics from Skype and WhatsApp before detailing credit card and teenager survey data. Toni validates the comparison with user retention observations. | |
| Infinite Inventory Parallels Between Shein, Netflix, and TikTok | 7 | 1 | 1 | 1 | Benedict articulates a structural parallel between infinite inventory dynamics across Shein, Netflix, and TikTok, noting how removing broadcast schedules and shelf limits forces rapid feedback cycles. Toni engages collaboratively with the premise. | |
| Users as Mechanical Turks and Algorithmic Fashion Generation | 8 | 2 | 2 | 2 | Benedict delivers an extended thesis framing consumer feedback on algorithmic platforms as a massive Mechanical Turk scoring system that enables generative design. Toni offers a brief interjection questioning whether auto-generation is already active in production. | |
| Points of Leverage from Google PageRank to Social Feeds | 8 | 1 | 0 | 1 | Benedict expands on historical points of technological leverage, detailing the operational differences between Yahoo directories, Google Street View, and PageRank's crowdsourced signals. Toni readily absorbs and reinforces the framing. | |
| Pop Culture Viral Hauls and Algorithmic Fashion Homogenization | 6 | 1 | 2 | 2 | Benedict and Toni engage in lighthearted banter about pop culture hauls, middle age, and VC dress codes before Benedict draws an analogy between social media radicalization loops and algorithmic fashion homogenization. | |
| Geopolitical Reactions and the Fashion Paperclip Maximizer | 7 | 1 | 1 | 2 | Benedict and Toni live-compare SKU numbers across regional versions of the Shein app, concluding with Benedict referencing Bostrom's paperclip maximizer and the Ratner jewelry scandal. Both summarize the environmental and supply chain implications. |
Statements from this episode (17)
Evans: Tech has fragmented into siloed conversations with no overlap
“It used to be that you basically, the thing was to understand what was happening in web two, and then the thing is to understand what was happening in smartphones. And now there's a fintech thing. So how do we think about the intersection of Square, which is n…”
Evans: Shein adds more SKUs every day than Zara carries in total
“Whereas Sheehan added something over 50,000 in the last seven days alone. And they add sort of seven, 8000 every single day. So they add more every day than Zara has in total.”
Evans: ASOS adds 5,000 new lines weekly and stocks 80,000-90,000 SKUs
“ASOS was this UK, is this UK online only fast fashion company, which does sort of 5000 new lines a week and stocks sort of a 80 to 80 to 90,000 SKUs at any given time.”
Evans: Shein uses browsing traffic to trigger 200-unit manufacturing batches
“You know, there's all sorts of different narratives about this, but you know, at a very, at a minimum, they're ordering in very short, small numbers. They may not even have actually ordered that product before they list it on the site. They're making 200 of th…”
Evans: Shein surpassed Zara and H&M combined in US sales
“No, Ernest, which is this company that does credit card, aggregates credit card bills, sort of said it's now the same size as Zara and H&M, and now, at the end of last year, it's bigger than Zara and H&M combined in US purchasing”
Evans: Shein ranks second only to Amazon in US teen brand survey
“Piper Sander does this study of US teenagers, and so what's the favorite brand, and for the last, like, four quarters, Shein, you could see it marching up every quarter, or every, every, every six months, it's a six-monthly survey, and the last two or three su…”
Evans: Online retailers redirect store rent savings into shipping and advertising
“Another big piece of it, of course, is that instead of spending the money on advertising, they spend the money, instead of spending the money on rent, they spend the money on advertising. So, you know, there's no, there's, you know, it's like a law of thermody…”
Evans: Shein and TikTok differ from Zara and Netflix via rapid feedback loops
“But it's also the feedback loop, and I think that's the change from ASOS, and certainly from Zara and H&M, and in some ways the change from Netflix, but it's what you have on TikTok, and what you have on on Shein is, is you have that very quick feedback cycle.”
Evans: TikTok, Shein, and Instagram users act as a giant Mechanical Turk
“And what you have in TikTok and what you have in Shein is you've got enough people looking at it that you've actually got that feedback loop. So basically the Shein user base or the TikTok user base is almost like a giant mechanical turk. Telling the system wh…”
Evans: E-commerce product imagery will almost certainly become entirely artificial
“I mean, a subsidiary point, of course, is how are you doing the product shops, and the answer almost certainly is that the product shops will all be artificial.”
Evans: IKEA has used artificial product imagery for a decade
“I mean, Ikea has been doing that for a decade and, you know, there's an awful lot of startups that now make it very easy to do garment, garment product shops artificially, and you can't tell the difference.”
Evans: Yahoo abandoned its directory around 3.5 million websites
“And they basically gave up in the early 2000 because they were like, I think they gave up at three and a half million sites and the whole thing just didn't work anymore.”
Evans: YouTube historically relied on watch data rather than video understanding
“Until very recently, or, you know, maybe now, YouTube doesn't know what's in the videos. They have no idea what's in the videos. They know who watched them.”
Evans: 'Shein Haul' hashtag has 5.3 billion views on TikTok
“So if it goes into TikTok hashtags, Sheehan Hall has 5.3 billion views, which it sounds like a lot to me.”
Evans: There will probably be a US political panic over Shein
“There hasn't been a Xi in panic yet, actually, even though it's another Chinese thing that every American teenager is using. But there probably will be. I don't know what form that will take.”
Brown: Shein excels at localizing regional fashion catalog styles
“Shein has been very good at localizing its content and giving people exactly the type of styles, whether you're based in America, whether you're in Europe, etc. So if it doesn't, I think they've done very well in not making people feel like they're buying some…”