Plot twist: Amazon’s future may soon be tied to physical stores

It’s hard to underestimate the success and increasing power of Amazon. Their market cap hovers just under $1 trillion. Their growth rates have been astounding. By most estimates Amazon now accounts for nearly 50% of all US e-commerce revenues, roughly 5% of all retail and is significantly bigger than their next 10 competitors combined. One study has some 55% of all online product searches starting at Amazon.

Last week a report that Amazon is considering opening up to 3,000 of their Amazon Go cashierless convenience stores by 2021 grabbed a lot of attention, despite their only having opened up a fourth location a few days ago. Advocates enthusiastically tout the concept’s potential ability to revolutionize shopping. Skeptics challenge the high capital cost, the reliability of the underlying technology and whether the stores really offer enough added value to take on well established players like 7-Eleven. I think both miss the larger point.

From a strictly pragmatic view, Amazon is not bound by the limitations of most retailers. They have patient investors who are much more focused on growth than short-term profits. Amazon has a strong commitment to innovation and has enormous capacity to invest for the long-term. While the economics of these stores do look rather challenging, the costs are certain to come down. And besides, at least for now, Amazon is not held to the conventional ROI hurdles that their traditional competitors face.

Whether or not the world sees 10 or 10,000 Amazon Go stores 5 years from now, what’s important to understand is that for Amazon to sustain anything remotely close to current growth rates over the long-term–much less defend against Walmart, Alibaba and others–they MUST significantly expand their physical store presence. You don’t have to possess a highly functioning crystal ball to see that one key to unlocking major growth in certain large product categories will require a substantial brick & mortar footprint. There are a few reasons for this.

The physical limitations of direct-to-consumer. Until someone invents a teleportation device (Elon, you on it?), considerable retail volume is impulsive driven, demands immediate gratification, is dependent on proximity to point of sale or is just stupid expensive to absorb the “last mile” delivery cost. Maybe Amazon is willing to have a robot or drone deliver a Slurpee to you, but that doesn’t make it a scalable business model.

The difference between buying and shopping. Amazon is really good at the “buying” process, i.e., those occasions where the customer values a highly efficient transaction, great pricing, vast (or very specifically curated) assortment and the particular convenience of direct-to-consumer delivery. “Shopping” on the other hand is less search and more discovery. It leans heavily on experience, be that the ability to interact in-person with a sales associate, see first hand the quality and/or fit of the product, figure out a broader, more complicated solution (like assembling an outfit or visualizing a re-decorating project) or simply to enjoy the social or entertainment dimensions that a brick & mortar location uniquely provides. While a customer “shopping” journey may be digitally informed, a physical dimension is often essential to conversion and customer delight.

When we understand this, it’s no surprise that most “shopping” dominant segments not only have much lower e-commerce share (groceries, prepared foods, furniture, home improvement, luxury fashion, etc.) but many digitally-native vertical brands (Warby Parker, Bonobos, Indochino, Casper) are investing in physical locations to reach consumers for whom pure online shopping is an obstacle to becoming frequent and profitable customers. Given the barriers to meaningful growth without a physical presence, Amazon will either have to place big brick & mortar bets (through their own formats and/or through acquisitions like Whole Foods) or accept a material deceleration of their growth rates over time.

Brick & mortar can be more profitable. Online shopping has two big profit drivers: the cost of acquiring (and retaining) customers with solid lifetime value and the per order dynamics of fulfilling orders. If the marginal cost of acquiring customers is greater than the marginal value of the lifetime value of those newly acquired customers the business model is unsustainable. This may well be the achilles heel of brands like Blue Apron and Wayfair. As many once online only brands are learning, it’s often cheaper to acquire a customer in a physical location than to pay the marketing tollbooth operators (Google, Facebook and, increasingly, Amazon) to target and convert the best prospects.

High fulfillment costs can make many e-commerce orders profit proof. There often is not enough gross profit per order for lower-priced items to offset the cost of picking, packing and shipping. This only gets worse when items are prone to high rates of returns or exchanges. This also helps explain why many online only brands are now opening stores and seeing their marginal fulfillment costs as a percentage of sales drop markedly. Amazon, on the other hand, is continuing to see fulfillment costs go in the wrong direction, thereby setting up a major headwind to improving lackluster margins. To reach more customers, improve marginal profitability and offset certain advantages of current (Walmart, Best Buy) and important future competitors (Nordstrom, Home Depot, Walgreens, Nebraska Furniture Mart) a significantly expanded brick & mortar presence is not nice to have, but essential.

While important, it is by no means urgent for Amazon to make an immediate big move. There is still plenty of solid growth within their core business model, including tapping into international markets. They have their hands full figuring out Amazon Go and Whole Foods. But in my mind, the long-term math leads to one inevitable conclusion. If Amazon wants to be the world’s largest retailer and significantly improve their margins a lot more physical locations are virtually certain to be a big part of that future.

A version of this story appeared at Forbes, where I am a retail contributor. You can check out more of my posts and follow me here.  

Over the next few weeks I’ll be in Chicago, Dallas, Toronto and San Antonio delivering an updated version of my keynote “A Really Bad Time To Be Boring.” For more info on my speaking and workshops go here.