5/20/22: How Shein Works - An Update
View this email in your browser
Insider Digest 5/20/22: How Shein Works - The Latest
Housekeeping / Announcements / Fun:
Splitting today’s newsletter into two parts, as there were two other posts that have been published on Circle about Chinese VC/PE and overseas investment. I’ll send that over the weekend with some collected thoughts and links on the situation surrounding both of those (decline in VC and proposed capital controls). Neither good news I’m afraid!
Although Shein continues to remain under the radar for many in the US (famed venture capitalists Chamath and crew apparently had never heard of the company as recently as this week’s All-In tech conference in Miami), in other circles, especially amongst fashion consumers and ecommerce investors, it has become the next breakout Chinese company to watch after ByteDance. Remarkably, it is listed as the fourth most valuable privately held tech startup by Crunchbase currently, (edged out by SpaceX), thanks to its recent $100Bn valuation. For reference, its 2021 revenue is rumored to be $15.7Bn, a bit over half of Zara’s (Inditex) at $29Bn (but market cap: $68Bn), and still just two-thirds of H&M, which was $23Bn (market cap: $19Bn). Too expensive? Well, definitely a valuation it will have to do some (a lot of??) work to justify, and IMO only works if it becomes a true platform versus just a brand. But I’ve already talked about that before, and so won’t belabor the point here.
I receive lots of questions on its tactics and as someone interested in cross-border ecommerce specifically (with quite a few deals in the space coming up for the TBC Syndicate) and even stating my thesis as hoping to find “the next Shein,” I thought I’d compile all of the latest and greatest Chinese research on this company. Especially since its “flexible supply chain” in Guangzhou province is really the heart and soul of Shein, I find that only Chinese reporters and analysts on the ground have been able to do any real quality due diligence. So here we go:
Shein’s Flexible Manufacturing Supply Chain
Shein has had a few years to work on its “flexible manufacturing supply chain,” which we describe below. Per a major supplier, this system was largely in place as of 2017. The years after that were mostly about scaling.
In fact, the Chinese supply chain began upgrading back in 2008, when Metersbonnwe started trying to copy Zara. The second wave was in 2013 when Taobao brands such as Handu came online and also wanted to lower inventory risk. 2015 brought with it a bunch more “wanghong” AKA influencer brands like Ruhnn. All of these helped, but none were as successful in truly remaking the supply chain as Shein, who’s turned this into a true infrastructure capability.
For fast fashion, the triumvirate — or before Zara & Shein, the “Impossible Triangle,” as it’s been called — is composed of 1) quickly onboard lots of new styles at 2) low prices while 3) being extremely efficient in inventory management. Handu, Ruhnn et al. managed the first two, but could never figure out the last.
Shein did. Its 2019 inventory turnover was 4.6x, higher than Zara’s 4.2x, Uniqlo’s 2.7x, and far better than Chinese fashion average of 1.85x
From a different perspective, we can see that VIPShop’s continued growth (86% in 2019, 50% in 2020) shows that many fashion companies are very bad at getting rid of their inventory.
Why did Shein succeed? It is hard to say. According to one of its long-time suppliers, it is mostly because they persevered and focused on the “dirty work” of improving the supply chain when pretty much everyone else was focused on marketing and sales. From 2014 to 2017 and even onwards, Shein invested the bulk of its funds into the supply chain, versus marketing and sales. Its flexible supply chain, its ability to deliver so many SKUs so quickly, is its sales strategy.
It must also be mentioned that once Shein has built up this capability, other brands will no doubt be able to leverage it as well. That’s why some people think Shein’s greatest contribution isn’t at the company level, but at the industry level.
Shein’s Suppliers:
As of Aug. 2021, Shein has 3-400 core suppliers, the bulk of which are located in Nancun village, Panyu district of Guangzhou city. Baidu Baike this village has only ~100K residents, but I’m guessing that doesn’t include the many migrant workers who work in its factories. Besides these 3-400 core suppliers, Shein works with another 1,000 or so smaller-scale suppliers, who are often just receiving outsourced work from the core suppliers. Nancun is also where Shein’s Guangzhou headquarters is located, and it’s estimated that half of the manufacturing capacity in Panyu is taken up by Shein.
Shein does also source from The Thirteen Factories, which is a historical site but is now also a wholesale bazaar for clothing. (Video of what it looks like now here.)
Many people thought Zara’s Chinese suppliers were also part of its flexible supply chain system, but this is not true, as Zara primarily manufactures standardized items in China, and its real flexible supply chain is in Spain.
But is Shein’s manufacturing “automated”? No. While its management software (which we discuss below) is, the actual method by which the clothes are produced is not. This kind of makes sense — profits are thin and order volumes start small, the bigger, higher-tech factories that are able to work with name brands aren’t fighting over Shein orders. Work mostly goes to the smaller factories. Per garment workers interviewed, they can sew up to 1,000 pieces a day if it’s a simple design, or 3-400 if it’s more complicated. Other workers who sew buttons can also do up to 300 pieces a day.
Workers are paid per completed piece, and typically work 10-12 hours a day, with two breaks in the middle, ending usually at 1030PM. Sundays are the only days they don’t work overtime. They do make a decent amount of money, about 8-10K RMB on average, about $1200-1500 per month, but it’s obviously extremely grueling.
Shein’s supply chain center of Panyu district is actually largely due to the fact that this is where the factories never really scaled up too much and most shops remain small.
Per Shein’s factory recruitment materials, FOB (free on board) suppliers, ie the ones who don’t come up with their own designs and are generally making more basic items, need only be at least 2,000 sqm in size. ODMs who do come up with their own designs need be just 1,000 sqm. This is the same requirement for MOTF, which makes higher priced / quality products.
But don’t forget, even all these small shops, at least when they work with Shein, must use the same centralized MES (manufacturing execution system, see below).
Because the shops are small though, this means they don’t have the capital or appetite to expand overseas as labor costs rise in Guangzhou. Thus, they are primarily looking at moving operations to inland China — Jiangxi, Hunan, Hubei are popular destinations. As an example, Jiangxi costs of living can be half as much or even lower of Guangzhou, and it’s just a 5 hour car ride away.
In fact, the latest reporting claims that the number of total suppliers for Shein are supposedly a lot less than in 2020. And 50% of the volume is now made in those cheaper inland provinces.
Shein Speed:
Samples for the initial order needs to be mailed to the buyer and updated within the MES (IT system explained below) within 3 days. Initial orders are typically 3-4 days (more than 5 days and suppliers face being eliminated, unless the item is something the buyer really really wants), while re-orders are 9 days or less. This is versus Zara’s well-known 14-day turn around.
Of course, we do have to add another 7 days for shipping. Conservatively, the entire supply chain from starting a design to when it reaches the mailbox of the final individual US buyer is 28 days (up to 14 days for design / sample, 7 days for initial order, 7 days to ship to customer).
The workflow is pretty intuitive. The ODM designer finds pictures on the internet and sends a selection to Shein’s internal buyer for consideration. The buyer picks some, and her manager does another round of selection. After that, the ODM starts working on creating samples, mails them to the buyer, and another employee who reviews the patterns (not quite sure how to translate, but the Chinese is 审版师) will provide feedback. Depending on feedback, there may be 2 or even 3 rounds of changes, before manufacturing can commence. This entire time, everything needs to be recorded in the MES, even chat logs.
Beginning this year, suppliers need to even include information on their materials and pricing into the MES. They’re not happy about this because to provide this level of transparency is not in their interest — suppose the deal doesn’t go through, there’s nothing stopping Shein from going to another supplier and negotiating for a better price.
Another point on the materials suppliers. Shein will actually call and check pricing on materials to make sure they’re getting the absolute best deal.
Shein is ruthlessly efficient when it comes to evaluating its suppliers. Timeliness of procurement and delivery, stocking and delivery, rate of defects, and success rate of new products make up 40% of how Shein evaluates its suppliers. The remaining 60% is based on order volume. They are then tiered into 5 levels, and bottom 30% of the lowest tier suppliers are culled.
Although Shein’s pricing is significantly lower than Zara’s (its most popular items are $5-25 whereas Zara’s are $35+), it is estimated that it can still make about a 5-10% profit margin.
Up to 1/2 of Guangzhou’s clothing suppliers may be working for Shein, but attrition is high. The reason is not because Shein is not doing well, but because QA (quality assurance) has become more and more stringent in recent years and many suppliers can’t keep up.
Suppliers are Struggling to Make Money Though:
Profit margins are very low. Basically, suppliers are either making no money, or often times, losing money on the initial order. They’re mostly hoping to make a viral item that can result in a large volume order for that item. (Recall that Shein’s m.o. is to make 100 pieces of each item, and only re-order at larger volumes when actual sales prove that they should be ramping up, so as to reduce inventory and working capital to a minimum. In startup terms, this is a sort of “minimum viable product” testing.)
But why do suppliers stay with Shein? Doesn’t this sound painful? Because, as people have mentioned before, they pay on time. Unlike many other clients, they pay on a monthly basis, which is the fastest in the industry.
Also because the fashion industry has been hit hard by the poor economy, the factories will still work with Shein even if they are making very little money because there’s no other choice.
As expected, OEMs (who do not design, and just make whatever design Shein asks of them) make even less than ODMs. Profit per piece ranges from 1-5RMB (or 15-75 cents, but typically 30-44 cents) and average costs are typically 10 RMB per piece or $1.48.
Initial orders used to be more than 100 pieces, but are now pretty much only 100 pieces, even for items with multiple colors, which used to start at 200 pieces. At these volumes, it’s a given that you are making very little money or losing money. (Quick math, if each item’s profit is 1-5 RMB, multiply that by 100, and you may not even be able to pay for the 500 RMB it typically costs to set up a sample.)
Shein’s Massive SKU Business Model
40-50K new SKUs per week, 660K SKUs on hand (as of October 2021)
This speed to market creates a virtuous cycle whereby new data points give Shein even more market intelligence about what they should make next. (However, I think there is a ceiling to this as fashion is one of those things where data goes quickly out of date.)
It’s hard to know if this speed and constant A/B testing (Shein tests 30 styles at a time versus 1-6 styles for Zara) is what accounts for the difference in “best sellers,” Zhongtai Securities estimates that only 20% of Zara’s items become hot sellers, vs 50% of Shein’s.
To give you a comparison of how this compares to other brands, H&M updates seasonally, and had 1687 new SKUs in Fall 2021 for the US market. Zara updates its inventory every 2 weeks, but does about 75K new SKUs per year (12000 for US market).
This is all enabled by … The “Technology”:
Shein requires all its suppliers to be plugged into its MES (manufacturing execution system). It pretty much does what it sounds like it does. “MES works as real time monitoring system to enable the control of multiple elements of the production process (e.g. inputs, personnel, machines and support services).”
Everything must go into this MES and so everyone is trained in it, including all of Shein’s ODM (original design manufacturing) partners. Per suppliers, it’s well designed and very intuitive to use, and everything is intelligent and automated. They just follow what the system spits out in terms of how many pieces to make in which sizes.
While both Zara and Shein don’t come up with original designs as much as they copy brands, Zara does try to hire more experienced designers and send them to global fashion events. Shein, on the other hand, just hires local Chinese vocational school grads who complete their research online.
Shein’s Warehouses:
Shein has 3 different types of warehouses:
Main Foshan warehouse in Guangdong province sends out 95% of the global orders. 30mm pieces in inventory, and 400-500K SKUs
Overseas transit warehouses: In Saudi Arabia, Dubai, Italy etc. 6 regions. Only receives returns. 10mm pieces in inventory, 150K SKUs
Overseas operational warehouses: 270 locations, but all small in size, as only accounts for 5% of global orders.
Marketing: Nothing Too Special, But Efficient:
It’s really about accommodating Gen Z, who number 2.4Bn as of 2019 and represent 32% of the global population. They love fast fashion. Supposedly 75% of Shein users are between the ages of 18-27.
Because of founder and CEO Yangtian XU’s SEO roots, Shein has been extremely successful at figuring out where to market, and has been able to take advantage of Facebook, Instagram, and now TikTok as acquisition channels. The company is at $100K of marketing spend a day on TikTok (unsure where this number comes from, but feels low?), and has double Zara’s views on TikTok (4.8B vs 2.4B).
Shein pays 10-20% affiliate marketing fee to influencers and doesn’t interfere with content creation. They also don’t really care about how many fans you have as it is % based.
Shein’s DAU was 22mm in April 2021, up 175% YoY
Traffic by geography (top 10): US 38.5%, Italy 9.8%, France 8.7%, Brazil 6.9%, Spain 3.7%, Canada 3.7%, Australia 3.5%, Germany 2.5%, Israel 2.3%, Russia 2.2%
Traffic by origin: search 42%, direct link 39%, social 8%, display ads 5%, referrals 3%, email 2%. (Thus, organic traffic is ~60%)
Do note these are all from various third party platforms as collected and summarized by Zhongtai Securities, and may differ from actual data
Sales Breakdown by Geography:
North America 35%, South America 10%, Europe 40%, Middle East 7% and 8% for Russia + Japan + Southeast Asia
Shein Company Structure
What’s Next?
Shein is putting in noticeable efforts to upgrade its brand from super-cheap to … well, less cheap and better quality. It’s taking sustainability much more seriously (although I think that’s due to media pressure), and collaborating with designers for more higher priced items. Although exactly how much higher priced? It looks like going from $10 to $20 … so maybe just trying to overlap more with Zara territory, but not quite hitting the same range.
As explained above, Shein has re-made much of the fashion supply chain in China into a more flexible one. Going forward, it is certain that Shein will not be the only beneficiary. Expect many other disruptive brands just like Shein to be birthed.
Shein has also started to apply its same flexible manufacturing supply chain strategy to other categories, most notably accessories, home and pets. It seems that they are integrating these suppliers into the same MES, although I imagine there are big differences in the workflow and inventory management. As I like to say, it is becoming much more of a platform than a brand, and it’s not inconceivable that it becomes more like an Amazon, with some more 3P elements.
What do you think, is it worth $100Bn? At what valuation would you consider an investment reasonable? $50Bn?
Reply here on the Circle forum
Have any comments or questions? See you on the Discord server!
Copyright (C) *|CURRENT_YEAR|* *|LIST:COMPANY|*. All rights reserved.
*|IFNOT:ARCHIVE_PAGE|**|LIST:DESCRIPTION|**|END:IF|*
*|IFNOT:ARCHIVE_PAGE|**|HTML:LIST_ADDRESS_HTML|**|END:IF|*
Update Preferences | Unsubscribe
*|IF:REWARDS|* *|HTML:REWARDS|* *|END:IF|*