A Shopping Center Has to Sell You Something Amazon Can't

The story I had in my head about retail real estate before working through this material was simple: Amazon is winning, malls are dying, end of story.

That's partly true. The e-commerce share of total retail sales has been climbing steadily for two decades, and the pandemic accelerated the trend in a way nobody fully reversed afterward. There are entire categories — books, electronics, certain kinds of apparel — that have largely migrated online and aren't coming back.

But the more I looked at how successful retail centers are operating in 2026, the more I realized the story is more interesting than "Amazon is winning." The shopping centers that are working today are ones that figured out how to sell something Amazon can't.

The future of retail real estate isn't about competing with online retail on price or convenience — Amazon will win that fight. It's about offering categories, experiences, and services that don't translate to a screen. The shopping centers that survive and thrive are the ones that lean into food, entertainment, fitness, services, and the small set of physical-goods categories where touching the product still matters. The ones that don't lean in become the ones being torn down or repurposed.

That reframing is what I want to walk through here.

The e-commerce share that retail can't ignore

The single number that defines this entire conversation is the e-commerce share of total US retail sales. It's been climbing steadily for two decades, with one major break around 2020.

A few things stand out from that trajectory.

First, the growth was already steady before 2020. From 2000 to 2019, e-commerce roughly doubled every five to six years — a slow, compounding migration that retail developers were already adjusting to long before COVID forced everyone to.

Second, the 2020 jump was real but partly a step-function. When stores closed during lockdowns, customers who had never bought groceries or clothes online learned how to do it. Some of that behavior reverted afterward. A meaningful chunk of it didn't.

Third — and this is the part that gets lost in the headlines — even at 16%, e-commerce is still a minority of total retail sales. The other 84% happens in physical stores. That's a smaller number than it used to be, but it's still a very big number.

What that math means for retail developers is twofold. The total retail pie that physical stores share is shrinking relative to e-commerce. But the absolute size of in-store retail is still enormous, and the categories where physical retail keeps winning are different from the categories where it's losing.

The categories that resist Amazon

If you walk through a successful shopping center today and look at what's still thriving, a pattern emerges. Almost all of it falls into a few categories that resist e-commerce for structural reasons.

The common thread is that each of these categories offers something a screen can't.

Restaurants and coffee shops sell prepared food on-site, which can't be shipped at the same quality and which is partly an experiential purchase — you're paying for the room as much as the food. Food and beverage has been one of the strongest growth categories in retail real estate for the last decade.

Entertainment — movie theaters, climbing gyms, bowling alleys, escape rooms, live music venues — sells experiences that only happen in person. The pandemic was hard on this category, but the survivors came back strong because there's no e-commerce substitute for actually being somewhere with other people.

Personal services — haircuts, manicures, massages, dry cleaning, tax prep — require physical presence and don't ship. This category has quietly grown into a larger and larger share of shopping center tenant mix as the goods-only retailers have left.

Fit-sensitive physical goods — clothing where fit matters a lot, mattresses, furniture, large appliances — still benefit from in-person evaluation. Online has captured a meaningful share even here, but a substantial chunk of customers want to see and try the product before buying.

Specialty and local retail — boutiques with curated selection, hobby stores with knowledgeable staff, pet stores with on-site grooming — compete by offering something Amazon's algorithm can't replicate, which is human curation and relationship.

The retail centers that have leaned into these categories are doing fine. The ones still clinging to a 1995 tenant mix — anchored by department stores, surrounded by clothing chains, with a small food court bolted on as an afterthought — are mostly struggling.

Omni-channel: when stores become endpoints

The other adaptation reshaping retail is omni-channel — the idea that the physical store and the website aren't separate channels but two parts of the same shopping experience.

The big retailers that survived the e-commerce wave — Target, Walmart, Best Buy, Home Depot — all built omni-channel infrastructure that uses the physical store as more than a place to walk in and buy something. The store became a pickup location, a return point, a try-before-you-ship room, a fulfillment hub for last-mile delivery.

That changes what a store needs to be. The classic 1990s retail box was designed around customers walking in, browsing, paying at a register, and walking out. The 2026 retail box has to handle that flow plus an order pickup desk, a curbside pickup operation, a returns counter that's actually staffed, and often a small back-of-house fulfillment area for shipping out web orders from store inventory.

For shopping center developers, omni-channel changed the spec. A new community center built today is more likely to have dedicated curbside pickup spaces, drive-aisle reconfigurations to support it, and back-loading access for tenants doing fulfillment. Those features don't show up on a marketing brochure, but they make the difference between a center where modern retailers want to lease and one where they don't.

The reverse migration nobody expected

One of the more interesting trends I learned about is the reverse migration — online-only retailers opening physical stores.

Tesla opening showrooms in shopping malls was probably the first highly visible example. Apple expanded enormously in physical retail throughout the 2010s. Then Amazon — the company that supposedly killed retail — opened actual physical bookstores. Glossier, Warby Parker, Casper, Allbirds, and dozens of other digitally native brands followed.

The pattern is the same in each case. The online-only retailer hit a ceiling on how cheaply it could acquire new customers through digital advertising. Opening a physical store turned out to be a relatively efficient way to acquire those marginal customers — the store generates awareness, builds brand, lets people touch the product, and (this is the part that surprised me) measurably increases online sales in the ZIP codes around the store.

For a shopping center developer, that's a meaningful trend. It means the universe of potential tenants has expanded — not just traditional brick-and-mortar brands, but digitally native ones that have decided physical presence is part of their growth strategy.

The "consumer services" reinvention

The most interesting tenant mix shift is what happens when shopping centers move beyond retail entirely.

The classic shopping center was 90%+ retail tenants. The successful 2026 shopping center is much more diverse.

The reasoning is simple. People go to shopping centers anyway, for groceries or coffee or a movie. If you can also get your taxes done, see a dentist, take a fitness class, and pick up dry cleaning while you're already there, the trip becomes more efficient — and you visit the center more often.

The tax prep example is the one that really clicked for me. Why would H&R Block be in a shopping center instead of a freestanding office? Because most people procrastinate on taxes, then realize they need to get them done. If H&R Block is in the center next to the grocery store you already visit weekly, you can drop off your documents, do your shopping, and pick everything up an hour later. That's an efficiency gain the freestanding tax office can't offer.

Healthcare followed a similar logic. Urgent care clinics have been moving aggressively into shopping centers because the same convenience math works — patients want care when they want it, and being near where they already go reduces friction. Dental practices, optometrists, physical therapy, and concierge medical practices have all followed.

The result is that a successful 2026 shopping center looks less like a retail destination and more like a daily-life infrastructure hub. Which, when you think about it, is what neighborhood centers always were — just with a broader definition of what counts as "neighborhood."

What this means for shopping center developers

If you're trying to develop a new retail center today, the analysis isn't "can we build a great-looking center on this site?" It's "can we assemble a tenant mix that genuinely benefits from being together, in this location, with these surrounding rooftops, given where consumer behavior is heading?"

The tenant mix question has gotten harder. The tenants that historically anchored centers — department stores, big-box electronics, bookstores — are mostly not options anymore. The tenants that work today require more curation and more market understanding than the old "lease the obvious anchor and fill the rest" approach.

That's also why the shopping centers being built today are smaller and more specialized than the regional malls of the 1990s. A grocery-anchored neighborhood center with a strong food and services tenant mix can still pencil. A new 1.2 million square foot regional mall with three department store anchors probably can't — and the data has been telling developers that for at least a decade.

The redevelopment side of this is also interesting. The vacant Sears box at a struggling regional mall isn't being re-leased to another department store. It's being converted into a Costco, an Amazon distribution facility, a self-storage operator, an entertainment complex, or in some cases torn down entirely and replaced with apartments or mixed-use. The building was built for a kind of retail that doesn't exist at scale anymore, and the highest-and-best-use calculation now points elsewhere.

What changed how I look at retail

What this material did to my mental model of retail is to reframe it: retail real estate isn't dying — it's reorganizing.

The categories that thrived in the 1990s mall model are mostly gone or going. The categories that benefit from physical presence — food, services, experiences, fit-sensitive goods — are growing into the space. The shopping centers that recognize that and curate around it are building healthy tenant mixes. The ones that don't are becoming candidates for redevelopment.

It's also more honest about what retail real estate actually offers. The old story was that the shopping center was where you went to buy things. The new story is that the shopping center is where you go to do things — eat, be entertained, get services done, see a doctor, take a class, meet people — and where you also sometimes buy things while you're there.

That's a more durable value proposition. Amazon can ship you almost any product faster and cheaper than the store down the street. What it can't ship you is dinner with your kids, a haircut, a yoga class, a movie with friends, or a doctor's appointment. The retail centers that lean into that gap are the ones that will still be standing in 2040.

A few things I'm taking away

• E-commerce is a real and steady share shift in retail — roughly doubling every five to six years before 2020 and stabilizing higher since — but physical retail is still the larger share of total spending and likely will be for a long time
• The categories that resist e-commerce for structural reasons — food and beverage, entertainment, personal services, fit-sensitive goods, specialty retail — are where shopping center growth still happens
• Omni-channel changed what a retail box needs to look like — pickup spaces, returns counters, in-store fulfillment, and curbside operations are now table stakes for major chains
• The reverse migration of online-only brands into physical retail expanded the pool of potential tenants and showed that physical stores still pay back as a customer acquisition channel
• The most interesting tenant mix shift is the move away from pure retail toward consumer services — taxes, healthcare, fitness, personal care — because they leverage the foot traffic the center is already generating
• The successful 2026 shopping center looks less like a retail destination and more like a daily-life infrastructure hub serving its surrounding neighborhoods
• Big regional malls anchored by department stores are mostly past tense — the new builds are smaller, more food-and-service-heavy, and tightly curated to what their specific trade area is missing
• Redevelopment of struggling retail real estate is a real opportunity — vacant department store boxes are being converted to fulfillment, fitness, entertainment, healthcare, and mixed-use rather than being re-leased to another department store

The thing that stayed with me most is how different the modern retail story is from the doom-and-gloom version I had in my head. Retail real estate isn't being killed by Amazon. It's being filtered. The categories that needed to be in stores are still in stores. The categories that didn't have moved online. And the centers that figured out which categories belong in which bucket are doing something Amazon will never quite do — they're giving people a reason to leave the house, get together, and spend an afternoon somewhere that isn't their living room. As long as that's worth something, retail real estate has a future.

This post is the third in a sub-series on retail within my broader Real Estate Development series. Earlier retail posts cover the percentage rent structure that defines retail leases and trade area underwriting.