DEEPER DIVES: We're Seeing The Start Of The A Supply Chain Strategy

Good Afternoon,

First, as always, thank you for joining.

Costs are on the top of everyone’s mind. Just when you think you get a breathe from the global economic impacts, the next challenge is in front of you.

While most organizations are stuck reacting to the news, diversification is being taking to a whole new level by others.

If there’s one thing that’s clear, operational excellence, innovation and cost control are the foundation to your future success.

Here’s what this issue brings:

• Shein’s building a new mega warehouse complex in Vietnam. I’m sharing my thoughts on what I see as their strategy and why it’s going to be a disaster no one sees coming
  

• Veho x Rivr announced a robot-assisted last mile partnership. I’m sharing why I think these projects are a waste of time (right now)
  

• You can now run your own AI model … on your phone.

Have US Tariffs Created An Even Bigger (And Uncontrollable) Monster?

Global trade winds are shifting.

Shein seems to be taking on a strategy that will completely change how major retailers access the market.

Confirmed reports point to a huge supply chain move, a mega-warehouse in Vietnam. One so big that it would be the equivalent of 26 standard football fields.

But this really isn’t about some state-of-the-art warehouse. They already have a massive manufacturing and logistics footprint in China.

This is Shein laying the groundwork for a multi-country manufacturing ecosystem to dominate mass-market demand, particularly in tariff-sensitive regions like the US (or any other country that’s looking to slow down “low cost goods from China”.

I’m going to unpack what this means, why it's happening, and who stands to win and lose.

Why Vietnam is Suddenly Center Stage

It’s no secret that the US-China trade relationship has been… complicated.

For Shein, a company that built its empire on an army of Chinese suppliers and the (now changed) "de minimis" rule, the game has fundamentally changed.

Consider this:

• The Old Way
  Packages from China worth $800 or less entered into the US without tariffs
  

• The New Reality
  Shein (and everyone else) now faces tariffs of around 54% on those same packages. For a company with annual sales north of $30 billion, there’s no way you accept the loss in profitability

The timing of Shein’s Vietnam warehouse decision near Ho Chi Minh City, isn't coincidental.

It’s a direct response to the new tariff reality.

The truth is it would be dangerous for Shein not to diversify.

The Vietnam Hub - A Nexus for a Diversified Empire

Here’s what I’m seeing.

Shein is building this Vietnam facility to be the backbone of a multi-country Southeast Asian (SEA) manufacturing operation.

The strategy appears to be:

• Diversify Manufacturing
  Encourage and incentivize suppliers (reportedly with 15-30% higher procurement prices and larger order guarantees) to set up or expand operations across various SEA countries (not just Vietnam).
  
  This creates a distributed manufacturing base
  

• Consolidate in Vietnam
  Funnel finished goods from these diversified SEA manufacturing points into the central Vietnam mega-hub
  

• Streamline and Ship
  From this D2C (Direct-to-Consumer) warehouse, manage inventory, process orders, and ship globally (leveraging Vietnam's favorable trade status with key markets like the US)

Ho Chi Minh City is a strategically brilliant choice.

Why?

• Vietnam’s largest port for imports (potentially from Chinese component suppliers or other SEA nations)
  

• Another major port handling exports to the US
  

• An international airport
  

• A government ambition to become a regional and global logistics hub

This ends up creating a resilient, agile, and cost-effective supply chain that can adapt to the ever-changing policies and regulations tied to international trade.

If one manufacturing country faces new restrictions, Shein can dynamically shift production to another within its SEA network, all while maintaining the central consolidation point in Vietnam for efficient export.

The Ripple Effect

What’s even more impactful is the how it will affect the mass retail market in general.

Temu's troubles highlight Shein's strength.

As a marketplace, Temu doesn't have the same direct control over its manufacturing sources.

This makes it extremely vulnerable to supplier disruptions and the direct impact of tariffs on its many foreign (often Chinese-based) sellers.

Reports from Business Insider show TikTok Shop (which shares many similarities with Temu's seller model) saw US sales from foreign sellers slump by 20-25% month-over-month in early May 2025, partly due to tariff headaches.

Sellers become hesitant, pricing becomes unpredictable, and the entire model starts to crawl under the strain.

By orchestrating its own manufacturing network, Shein gains a significant advantage in stability and cost control.

And if you don’t think the market responds immediately to changes in the tariffs, think again.

The second the US and China announced a temporary 90-day pause on most tariffs, Alibaba.com saw an instant reaction.

Orders from US businesses surged by 40% year-over-year.

Everyone rushed to stock up.

This shows just how sensitive eCommerce volumes are to tariff policies (since so much volume is based directly on the price the consumer will pay) and validates the huge financial incentive behind Shein's diversification and “tariff-arbitrage” strategy.

While it might all feel like high level strategy for billion dollar brands, the squeeze is really going to come to smaller businesses.

D2C brands (and less mature retailers) simply don't have the capital, scale, or leverage to support a multi-country manufacturing strategy (let alone build any dedicated overseas consolidation hubs).

As Shein optimizes its costs and delivery times through this new model, the competitive pressure on smaller players, already having a hard time dealing with rising costs and (much more) complicated logistics, will only get worse.

They are caught between a rock and a hard place, unable to match these types of strategic moves.

Big retailers Can (and will) follow suit.

Amazon and Walmart are watching closely.

They also have the resources and global reach to implement similar strategic manufacturing diversification.

The "China plus one" strategy isn’t new and has been a talking point for years already.

Shein is just executing it on a scale and with an integrated D2C focus that sets a new benchmark.

Necessity is the mother of invention

The New Playbook

So, is Shein's new model unbeatable?

Not necessarily, but it's definitely changing the game, and Shein still has their own challenges.

US authorities are aware of goods being illegally rerouted through countries like Vietnam to avoid tariffs.

Shein will face increased scrutiny to ensure genuine value-added manufacturing is occurring in its declared countries of origin, not simply a network of transshipments.

(Their investment in actual manufacturing incentives and a massive physical hub in Vietnam suggests they are preparing for this)

Managing a multi-country supplier network feeding into a central hub isn’t as easy as a China-centric model.

It requires sophisticated supply chain management, quality control, and logistics coordination.

But, Shein has shown that they are able to make what others thought was impossible, possible.

The Uncomfortable Truth for Policymakers

One of the biggest realities of Shein's strategy is how difficult it becomes to target "low-cost goods from China" without inadvertently penalizing other countries or even domestic retailers.

If Shein is legitimately manufacturing across multiple SEA nations and consolidating in Vietnam:

• Tariffs aimed solely at China become less effective against Shein (or anyone else copying this playbook)
  

• Broader tariffs or restrictions on goods from the entire SEA region could harm many other businesses, including US companies that have also diversified their manufacturing there
  

• It highlights that no single country of origin will be a silver bullet for brands seeking to de-risk or for policymakers trying to manage trade flows

This new model, if successful, positions Shein to maintain its cost leadership and rapid delivery promises.

Even if all SEA countries were eventually impacted by similar US trade policies, Shein would likely still be in a more favourable position than competitors due to the sheer scale, efficiency, and resilience built into this diversified, centrally consolidated system.

The Game Has Changed

Don’t get stuck thinking that what Shein is doing won’t impact you. They are making a proactive move that will reshape global operations.

Investing billions towards a future where supply chain agility, diversification, and cost management are paramount.

Here are the key takeaways you need to be thinking about:

• Diversification is no longer optional, it's essential.
  But real diversification is complex and resource-intensive
  

• Country of origin will become a fluid concept.
  The focus will shift to the resilience and adaptability of the entire supply network
  

• Smaller brands face an ELE if they cannot find flexible, collaborative solutions to compete with the scale of businesses like Shein
  

• Large retailers have a new strategy to study.
  The pressure to optimize global manufacturing and logistics will only increase (for everyone)
  

• Controlling this new model is a wicked problem.
  There are no easy answers for policymakers that don't also risk broader economic impacts

Shein is building a a fortress designed to withstand anything that comes their way. They are making sure that they stay a dominant force in fast fashion.

The rest of the industry needs to pay very close attention.

Why Last Mile Delivery Robots Are A Waste Of Time

There was a lot of hype across the logistics networks after the announcement of Rivr x Veho made regarding a new pilot project for robot-assisted last mile delivery.

You can check out a promo video here.

I got a ton of questions because I said that “while cool, I think this is a waste of time”.

Too many people thought I was tearing down innovation.

Or don’t see the future potential.

Not true.

But anyone that has actually designed and run delivery networks knows that there are two things that drive profitability.

Density and speed.

(And by speed, I don’t mean ultra-fast delivery).

Making money in last mile is all about transactions per hour.

The more you do, the better of you’ll be.

The fact is delivery robots at this stage are simply too slow.

Too slow to maneuver.

Too slow to load.

Too slow to unload.

Too slow to charge.

Too slow to get in and out of vehicles.

Even under the assisted model Veho is thinking (driver delivers one package and the robot another), you still end up with challenges one you hit critical density thresholds.

A driver that has to take time to load a second package into a robot multiple times a day is going to be slower than simply doing the delivery themselves.

Amazon has become notorious for their “multi-stop” package consolidation. They go as far as grouping multiple side-by-side houses as one stop, because they feel a delivery person should only leave and return the vehicle once when addresses have a certain proximity.

Where a robot can help is when you don’t have enough density on your run and you want to have the robot cover more ground while the driver is doing something else.

In this specific case, you could see a benefit.

But to get that benefit, you need to have the robot moving at a rate of speed that’s at least double that of the driver (RIVR states their robot can get up t o 27 km/h).

While this speed is a massive shift, you still have to think of the realistic applications and the health and safety concerns that come with it.

These days, no one is expecting a small robot to be moving as fast as someone peddling on a bike.

Especially a robot that’s a lot lower to the ground and harder to see.

Kids in neighbourhoods. Drivers moving around the block. It’s an accident waiting to happen.

To prevent that, everyone slows their machines down.

Another recent test by a robotics company here in Canada that has partnered with SKIP to do meal deliveries has the robot speed limited to just under 5 km/h. This is a touch less than most adult’s regular walking speed.

So you get little advantage at these levels other than some marginal parallel activity (that completely vanishes with improved density).

The other big thing about these technologies is that companies are not realizing the negative consumer sentiment that will come with them once they go mainstream.

No one cares about seeing a novel delivery robot today.

There’s barely any of them out there.

But what happens when every street in every neighboourhood in every city has dozens, or hundreds of these things moving around each day.

Below is my favourite example to explain this situation.

With a lower capacity per robot, it means that you will have way more “vehicles” on the roads and on sidewalks delivering packages.

Like it or not, city infrastructure was not made to support it.

That means that as you invest more into these ‘delivery options’, you’ll be creating more and more frustration for the customers you are looking to serve.

Now, do I think robots are a complete lost cause?

Absolutely not.

But if they are going to be used, they need to open up capacity and options that don’t really exist today.

There’s no value in using them the same way you would a human labourer while they are smaller and less efficient.

Keep that in mind as your organization is looking at new innovation.

Are you solving a problem or do you have a solution looking for something to solve.

Big difference.

Don’t Get Sucked Into ‘Easy’. Not If You Want To Get The Most Value From AI Anyways

Whether you are working in your own business or within a team at a large organization, there’s little to no chance you haven’t seen a huge upswing in AI adoption.

Tool after tool is trying to sell you on its solution.

How it’s the only one you need.

Full end-to-end visibility and control.

The truth is this.

AI is moving too fast and changing too quickly to be locked into a platform that doesn’t offer enough control.

I shared similar thoughts a number of weeks ago (check out this Youtube video), and my position is only getting stronger with what’s going on in the market.

Google released an app last week that allows users run their own AI model.

Called Google AI Edge Gallery, the app is currently available for Android and lets users download and run compatible models (from AI dev platform Hugging Face) that generate images, answer questions, write and edit code, and more.

Here’s the wild part….

You run the app and run the bots ON YOUR PHONE.

Now, I’ve played with some self-hosted AI solutions, and right now I am comfortable saying that the performance is nothing like what you have gotten used to by using ChatGPT or Gemini or any other major online model.

(Unless of course you run absolutely killer hardware - then the sky’s the limit).

They key here though isn’t the performance.

It’s that model selection and features are going to continue to be the differentiators and what will really help your business standout.

You can’t realistically think that you can use the same platforms, the same models and execute them in the same ways but somehow get dramatically different results than your competition.

Generic AI is going to be the lazy button that keeps you and your team locked into mediocrity.

What you want to be looking for are platforms or tools that allow you to:

• Choose your own models

• Select your own additional tools

• Load in your own data (to create amazing RAG systems)

This is the only way that you will truly get AI that’s dialed into your business and will be capable of being an extension of your brand promise.

Simple chatbots and walls of weakly generated text will become as annoying as commercials are (for most people anyways) today.

Build your own future.

Don’t buy into what someone else thinks what you need.

That’s it for this week. Thanks for being here.