TJ Maxx's Moat Isn't the Treasure Hunt. It's That You Can't Ship One.

Walk into a TJ Maxx looking for a specific thing and you will fail. There is no aisle for it, because there are no permanent aisles — the selling floor is, in the company's own words, 'flexible, without walls between departments and largely free of permanent fixtures.'⁶ The cashmere sweater you found last week is gone and will never come back. What's there instead is a wall of mismatched lots: a designer handbag beside an off-brand one, twelve of this and three of that, all of it 20% to 60% below what full-price retailers — including the major online ones — charge for the same goods.² Shoppers call this the treasure hunt. It is the most quoted retail moat of the last decade. And almost everyone describes it as a feeling. It is not. It is a logistics fact that happens to feel like fun.

The official story is that TJX thrives because the treasure hunt is irresistible and addictive — a dopamine machine that Amazon can't copy. That's half true and entirely beside the point. The deeper truth is that the same randomness that makes the store a hunt is exactly what makes a website impossible to run. The chaos isn't a marketing flourish on top of the inventory. The chaos is the inventory, and you cannot ship chaos.

The same buy that fills the store breaks the website

Start with where the goods come from. TJX doesn't order a season of merchandise from a catalog months out. Its buying organization — over 1,300 people in the FY2025 filing, sourcing from more than 21,000 vendors across 100-plus countries — chases closeouts, order cancellations, and manufacturer overruns wherever they surface.²⁶ A brand over-produced; a retailer canceled; a factory ran long. TJX takes the lot, small and specific, and moves on. Tomorrow's lot will be different goods, different sizes, different stores. This is what 'opportunistic buying' actually means: the assortment is assembled from accidents.

Now try to put that on a website. E-commerce demands the opposite of an accident. It needs a SKU that can be photographed once, described once, searched, restocked, and reordered when it sells. A national web catalog wants depth — hundreds of the same item, available everywhere, replenishable. TJX's buy is the inverse: shallow, random, one-time, and varying — as one analysis put it — more on a store-by-store basis than any other class of retailer.⁸ The cost of cataloguing a forty-unit closeout that will never be reordered is brutal against the thin margin a bargain carries. The store can absorb that randomness for free, because a shopper walks the floor and does the searching with their own feet. A website has to do the searching for the customer — and the inventory refuses to hold still long enough to be indexed.

This is the asymmetry that gets missed. Department stores went online because their inventory was already online-shaped — branded, deep, replenishable. They could ship what they sold. TJX cannot, not without changing what it sells. And the moment it changes what it sells to suit a warehouse, it stops being able to offer the vendor what the vendor wants: a fast, quiet, no-questions buyer for unwanted lots. Break the inventory model to fit the channel, and the discount that draws both shopper and supplier evaporates. The moat is self-reinforcing in the cruelest way for a competitor: you can't copy the front end without copying the back end, and the back end won't survive the copy.

The scoreboard while everyone else went digital

The years that were supposed to kill physical retail were the years TJX grew fastest. It surpassed $50 billion in annual sales in fiscal 2024, then reported $56.4 billion for fiscal 2025 — up 4%, with $4.9 billion in net income and comparable store sales up 4%.¹³ Tellingly, that comp figure explicitly excludes e-commerce.⁷ The growth was the floor, not the screen. And the company kept building floor: 4,954 stores at the end of FY2024, 5,085 a year later, over 5,200 after that — against a stated long-term target of 7,000 locations.²³⁴⁵ TJX runs six e-commerce sites, but it labels them with a single word that tells you everything about their role: 'complementary.'⁵

There's a second mechanism hiding in the physical bet. Because the floor has no permanent walls or fixtures, TJX can swing whole categories overnight — pour more apparel in when handbags are scarce, expand home goods when a furniture closeout lands.⁶ The store is a buffer that absorbs whatever the buyers happen to catch. A website has no such elasticity; it is a fixed grid of slots that wants to be filled the same way every week. The flexible floor is the physical expression of opportunistic buying. One could not exist without the other, and neither has a digital equivalent.

But couldn't a smarter operator just out-logistics them?

The fair objection is that 'you can't ship random inventory' sounds like a failure of imagination, not a law of physics. Plenty of marketplaces sell one-of-a-kind goods — eBay, RealReal, Poshmark all move singular items at scale. So why couldn't a well-capitalized player build an off-price site that catalogs the chaos faster and cheaper than TJX assumes?

The honest answer is that some could, at the margin — but they'd be solving a different problem and winning a smaller prize. Resale marketplaces work because the seller does the cataloguing and the buyer pays a price that funds the friction. TJX's model funds nothing of the kind: its whole promise to the vendor is speed and discretion — take the lot, no fuss, no listing it next to the brand's full-price line online for the world to see. Force that lot into a public catalog and you've handed the vendor a reason to say no, and undercut the very channel separation that makes off-price tolerable to brands. The benchmark hides in the 10-K's own pricing language — '20% to 60% below full-price retailers, including major online retailers.'²⁴ TJX isn't dodging Amazon. It is positioned against the online full-price world by being the place that price can't follow. The moat isn't that nobody can build an off-price website. It's that nobody can build one that keeps the vendors, the discounts, and the margins all at once.

TJX spent the digital decade doing the unfashionable thing — opening rooms with no walls and filling them with whatever the world over-made that week. The competitors who fled to the screen were optimizing for a customer who wanted to know exactly what they'd get. TJX bet on the opposite customer, and on a supply chain that could only ever feed a floor. The treasure hunt was never the moat. The moat is that you have to be standing in the store for the treasure to exist at all — and there is no shopping cart in the world that can hold an accident.