Amazon and Bricks: A Love-Hate Story

Last week, Amazon announced that it will be shutting down all its Amazon Books and Amazon 4-Star stores. The closures will affect a total of 68 brick-and-mortar stores across the U.S. and the U.K.

Over the years, Amazon has gradually launched several brick-and-mortar-type stores. From supermarkets to retail, some offer the top-selling products found online, while others offer Amazon’s private brand electronics such as Fire tablets and Echo speakers. 

Looking at Amazon’s Q4 2021, the firm’s online store generated $66.08 billion in revenue (with 18% YoY growth), while Third-party seller services made $30.32 billion (with 11% YoY growth). Physical stores brought in only $4.69 billion (less than half of the ad business, and a quarter of AWS), and this is across all 600 stores, including Whole Foods. Growth, however, was about 17% (in line with the growth of the Web store). 

However, the most interesting data is that “Physical stores, which includes Whole Foods and Fresh outlets, reported lower sales in 2021 than in 2018.”

Since these stores are only a small fraction of Amazon’s total revenues, the decision to close them is not surprising.

And it’s even less surprising if you have visited any of Amazon’s seven different brick-and-mortar store formats. They’re just not great, and this is not a case of 20/20 hindsight. 

In 2017, the New Republic published the following article: The Amazon Bookstore Isn’t Evil. It’s Just Dumb. The article emphasizes the fact that the stores had all books face-out, as seen in the photo below: 

The reason behind this format is that it “replicates” the main advantage of Amazon’s online store: it’s easy to discover new books.

“The most obvious way Amazon Books pushes discoverability is that every book is displayed face-out, as opposed to spine-out, as you would on your bookshelf at home. Every face-out has a small placard that features the book’s star review and a short customer review. These are both presented as innovations, but they’re really just very, very old bookstore conventions taken to an extreme. In my short career as a bookseller, I learned one thing: Face-outs and recommendations sell books.”

The trouble with this practice is that it forces the store to hold a significantly smaller number of books:

“The problem with a limited stock, however, is that it’s a limited stock. It’s not entirely clear how these books have been selected. The fiction section is a hodgepodge of mostly commercial fare, ranging from the midlist to the beach-y […]. The store’s history section runs from basic-dad book to basic-dad-who’s-interested-in-terrorism book, with, again, a few exceptions.”

This highlights one of Amazon’s main misconceptions: what works online will also work for brick-and-mortar. 

But this problem is not only present in the stores Amazon is planning to shut down. It’s clear that Amazon is not doing great with the stores it’s planning to keep either. 

And the pandemic only made these issues more glaring: 

“In fact, Whole Foods lost market share in 2020. In-store grocery sales at most other retailers — from specialists like Kroger and Sprouts to generalists like Walmart and Target — were up (a lot) in 2020. Overall, grocery spending has been up between 15% and 25% since the pandemic began in March. Whole Foods is an expectation and its stores have not performed well throughout the pandemic. Admittedly, some of this is a deliberate part of Amazon’s efforts to shift sales online, but there’s more to the story.”

If you are a Whole Foods customer, you will know that it’s been having its fair share of issues since it was acquired by Amazon. 

Comparing the data before and after the acquisition, we see that although existing customers increased their traffic by a bit (5%), the loss of new customers is at a rate similar to the one before the acquisition. 

All this without improving the basket size. 

But this is not only about traffic. The entire shopping experience has been hampered. Since the acquisition, Amazon has only made small modifications to Whole Foods (beyond the increase of stock-outs), with main changes including discounts and free home delivery for Prime members.

But In 2022, out-of-stock shelves are still a common sight. However, even in this dimension, we could say that Amazon remains a “pioneer”. Below is a photo of shelves at a Whole Foods location a year after the acquisition. 

Multiple interesting questions arise. 

First: Why does the firm that “invented” e-commerce even want to invest in brick-and-mortar stores?

Second: What can we learn from these store closures? 

Third: Is it possible that Amazon, the almighty retailer, doesn't know how to run the oldest form of retail… a.k.a. a store?

Let’s start with the last question.

Why did Amazon “Fail”?

The first reason is that running a store with a limited customer base (only those within the limited geography of the store) and with limited shelf space, requires vastly different capabilities than what’s needed to run a retail operation with an unlimited customer base and unlimited shelf space (albeit with more complex logistics). 

I would say that the two main issues are assortment management and inventory management (which are, of course, related). 

When it comes to assortment, there are very few firms that are as good as Amazon at continuously running experiments to elicit preferences from customers. When operating an online store, every interaction is one you can learn from. You offer two products, you see which one was clicked on and which one was purchased, and you can immediately improve the experience for the next customer. Amazon and Netflix are probably the best at this. However, this process is near impossible when operating a brick-and-mortar store. There is limited ability to experiment, and an even more limited ability to learn from every interaction. The list of books outlined in the New Republic article above, is a good example of this struggle. Amazon was always good at the “Long Tail”: niche products that people may never discover since regular stores don’t carry them. But by running a physical store, you cannot target the long tail. You will end up targeting the “average-dad”. 

A similar issue is exhibited when it comes to inventory management. Amazon’s main operational force comes from the notion of pooling and centralizing resources. By managing inventory centrally, it’s possible to maintain a high service level while carrying limited inventory, since you can ship to customers wherever they are. This is the same idea behind AWS; allocating centralized resources to people, based on need, without any precommitment. This is not the case when selling in physical stores, because you have to commit inventory. It’s not rocket science to run a store, and many retailers manage well, but Whole Foods (which should never stock out given its prices) frequently exhibits empty shelves. 

This brings me to the second reason for this “failure”: Amazon is really an online-first firm. You can see it with Whole Foods. Part of the decline of the brand (which has always been known for its high prices, yet high availability and good shopping experience) is the fact that in the store's aisles, you are competing with Amazon Fresh shoppers. These are not just stores anymore. They are fulfillment centers. And no one wants to shop at a fulfillment center for the prices Whole Foods charges. 

Is Amazon Really Failing?

But maybe the fact that these stores failed (and are probably still failing) is a good sign for Amazon's “Day 1” culture/strategy. 

Amazon is by definition, an experimentation firm, and in that sense, Amazon has demonstrated its low barrier for low-quality experimentation multiple times. A famous quote by Reid Hoffman, the founder of LinkedIn, says: “If you are not embarrassed by the first version of your product, you've launched too late.” Amazon may not be a startup, but it follows this idea for almost every new product or concept it launches. This is true for some of these store formats, which were clearly launched before being fully and well thought out, and it’s also true for many of the products Amazon has launched (such as the Amazon Fire). Most of these products wouldn’t even qualify as a prototype at Apple, but I don’t mean this in a negative way. It’s part of Amazon’s core competencies. 

Amazon also has a very very high willingness to admit when they have run in the wrong direction and need to course-correct. We saw it with grocery when Amazon decided to scrap years of work, acquire Whole Foods, and redesign its entire strategy. 

We are seeing it again now, with the closing of these stores. 

And indeed, an Amazon spokesperson said that the company will remain committed to building long-term physical retail concepts and technologies, citing the recently launched Style stores, the first foray into physical clothing stores. The company will also continue to focus on its Amazon Fresh and Whole Foods Market grocery chains, Amazon Go convenience stores, and embrace the Just Walk Out cashierless technology.

And this highlights an important aspect: Amazon is always on to “the next thing.” 

A few weeks ago, and while starting the process of shutting down its book and 4-Star stores, Amazon launched yet another store format, Amazon Style, and started making significant changes to Whole Foods. 

Last week, the New York Times published an article on a pilot Amazon is running:

“But this 21,000-square-foot Whole Foods just north of Georgetown has catapulted Amazon’s involvement forward. Along with another prototype Whole Foods store, which will open in Los Angeles this year, Amazon designed my local grocer to be almost completely run by tracking and robotic tools for the first time.

The technology is essentially what Amazon has in its Amazon Go: It identifies when we lift a product from a shelf, freezer, or produce bin; automatically itemizes the goods; charges us when we leave the store.”

The Brick-and-Mortar Store

This brings me back to the initial question: In a world where most of the growth is coming from e-commerce, why is Amazon even interested in brick-and-mortar? 

There are several reasons. 

First, there is the realization that some products need the brick-and-mortar format. 

Grocery: some aspects of grocery shopping will always remain in a brick-and-mortar format. Simply because, for certain items (e.g., fresh fruit or vegetables), people don’t trust anyone else to do the picking for them. 

The brick-and-mortar store also plays a role in stimulating demand: most food purchases are driven by how things look and smell. Sometimes what you decide to make for dinner depends on what the store has and how it looks (think fish, meat, etc). 

And finally, one of the unique features of grocery online shopping is the need for proximity to customers. Customers want cheap and quick delivery and (almost) the only way to do this is by being close. These stores can be both micro fulfillment centers, while also allowing the customer to stop by (further reducing the cost of shipping). 

Clothing: Buying clothes online is still far from perfect. As we mentioned earlier in some stores, 50% of apparel bought online is returned. By having a physical store you can reduce returns and allow people to measure before they buy. 

While Amazon is not known for style, I was surprised to find good statistics regarding its apparel and footwear business (which includes Zappos).

“Amazon’s U.S. apparel and footwear sales grew by about 15% in 2020 to more than $41 billion, Wells Fargo estimated, about 20% to 25% higher than giant rival Walmart. Amazon now has 11%-12% market share of all apparel sold in the U.S. and 34%-35% share of all apparel sold online, Wells Fargo analysts wrote in a research note.”

This brings one more potential role for any Amazon store. If you have had to return a product you recently bought online, you probably know that you can just drop it off at a nearby Whole Foods. 

If 25% of all products bought online are returned, the ability to cut shipping costs at the scale Amazon has, is extremely significant. In fact, I would argue that even if these stores just barely break even, their impact on cost reduction for Amazon’s reverse-logistics, is sufficient to justify their existence. 

However, and having said that, there are still some stores that never made sense. 

For books and electronics, which is what the stores in question specialize in, I don’t see a significant advantage of running a physical store. User-generated content is usually more important than what you get from a store associate, plus, you get it delivered to your door the next day. 

However, it’s not that there is no advantage at all. A few years ago, I needed a book light enough to carry on a multi-day hike (we did the Tour Du Mont Blanc), but big enough to last for 7 days of hiking. Unfortunately, finding it online was challenging, so I went to Barnes & Noble. This is the first book* I bought from a physical store in the last 20 years (and I do buy hardcover and softcover books much more than I buy online books). Assessing the size, weight, and “flexibility” of the book was important (on top of the content). But this is clearly a rare occurrence. 

*In case you are wondering which book I bought: Churchill and Orwell: The Fight for Freedom (the softcover, which is no longer available). 

So this brings me back to the fact that a brick-and-mortar store should be on the menu of almost every retail firm which is above a certain scale (given the significant fixed costs of running such a store).

They are clearly less scalable since they need to be local, but I don’t think we have seen the last of them. Firms are experimenting with more Omnichannel strategies such as “Buy Online, Pick-up in Store” (BOPIS).

As we see more cities banning “dark stores” (stores designed only for fulfillment), we will see firms trying to figure out how to offer good online and fast delivery, while rethinking the physical store.