Analyzing Retail Industry Disruptions (Part 2 - Digital)

The Disappointment of Digital Disruption

This continues a multi-part series that seeks to better understand potential A.I. disruption to the Retail industry.  Part One is available here: Intro/Favorite Retail Industry Chart

The evolution of E-commerce/Digital certainly disrupted the Retail industry, but not in the way almost all companies originally expected.  Ten years ago, it was consensus that E-commerce would drive higher margins yet fast forward to today and the exact opposite is what has happened.  To his credit, Gary Friedman at RH was really the only person I remember to openly challenge this narrative, way back in 2017:

“the web is not the most profitable channel. And I still don't know why anybody believes that…I spent 14 years at Williams-Sonoma playing the channel game that, quite frankly, was just moving cost or assets from one channel to another...most retailers who have increased online sales have decreased operating margins…So why are people trying to shift sales online? Why would a retailer take business from a highly profitable store, okay, and try to spend capital, spend expense to try to move sales to another channel and create a whole new cost structure, not really understand what the true metrics and how to make money there are. And I get Amazon has become $140 billion, but it's not a lot of operating margin.”

MARGIN

Gary was prescient with those comments as we are still waiting for the magical margin expansion associated with digital growth at most retailers.  For this discussion, I picked several leaders in their respective Retail industry sub-sectors to see what has happened to margin since 2015.  While we have seen some expansion at roughly half of the names, it is nowhere near the overly bullish levels most investors expected ten years ago.  Furthermore, we have seen meaningful margin compression at several names.  Mind you, these are leaders with scale so likely represent the best positive examples of Digital disruption for those respective sub-industries.  Lastly, COVID by itself should have been a major competitive accelerator towards margin expansion for names such as these.   

As you dig into these individual stories, it’s hard to find one that can clearly claim the Digital evolution as the primary reason for margin expansion.  Logistics is likely why Williams-Sonoma is one of the few retailers that has enjoyed meaningful margin expansion since 2015, as they already had a 50%+ E-commerce sales mix back then (happy to talk more about that in detail but RH stands as a testament).  Hard to tell if it’s Lululemon’s Product improvement or E-commerce growth that delivered meaningful margin expansion since 2015, but probably more the former.  Autozone has certainly not enjoyed enough E-commerce growth to enter into this conversation – let’s consider this the placebo group for margin expansion.  Home Depot likely has received some benefit here but continues to enjoy an oligopolistic market share position that really was never challenged much online.  Dicks Sporting Goods might be the single biggest success story here, but also had a major push into vertical integration during this time period through somewhat successful private label brand introduction (Calia, etc.).  The main point being that nobody outside of LULU and WSM has reached anywhere near what consensus used to think about the margin opportunity of E-commerce. 

EMPLOYEE INTENSITY

To better understand what happened, let’s first start with my favorite chart as discussed in part one.  You might be surprised to learn that most retailers today have even more employees per square foot than they did back in 2016.  That factor was originally the more alluring thing about the Digital channel as previous margin expansion expectations were largely based on reducing the employee content per sales dollar.  And they ended up being right!  Across my sample set, Gap, Nordstrom, and Williams-Sonoma have had outright declines in their ESF (employees per 1,000 square feet).  And every single retailer has had a bigger increase in their SSF (sales per square foot) than their ESF. 

So, everyone largely nailed the main factor they were focused on (employee intensity to sales), yet still ended up massively wrong on forward margin.  Why?  There are numerous potential reasons for why margin upside from the Digital channel failed to materialize.  The retail industry’s original approach to shipping fees materially changed, likely as a direct result of Amazon’s decision to acclimate the consumer to free shipping.  Return policies ultimately became a major and costly logistical challenge that was largely overlooked in early retailer margin optimism.  Google, Amazon, Meta, and other Digital advertising platforms have largely used their ever-increasing scale to maintain, if not increase, pricing power.  There has also been a perpetual investment to “Keeping up with the Joneses” as newer, yet unproven technology is constantly introduced and never really seems to create anything other than a short-term benefit at most before either failing or commoditization.  To that last point, an individual physical retail store typically only requires a meaningful investment every 7-10 years, which leads to long periods of leverage in-between. 

MONOPOLY

Alas, I firmly believe the true culprit to retailer margin pressure for the past decade+ is simply the transformation of a historically monopolistic industry into one much closer to perfect competition.  Not just nationally, but globally.  While the Digital evolution provided opportunities for many retailers to expand their reach to consumers, at the same time, it eliminated the monopolistic like features of the physical store world that retailers used to enjoy.  As an example, for the first time ever, the individual Abercrombie & Fitch store in Tysons Corner, Virginia, directly competed with ASOS, a British E-commerce site.    

The Retail industry prior to E-commerce required a significant upfront cost to open physical stores that only competed for the TAM (total addressable market) of a relatively small area.  The limited TAM acted as a natural barrier to new competition in those areas.  

Nowhere is this transition more evident than the death of several “Category Killer” business models.  Toys R Us, Bed Bath & Beyond, Linen’s N Things, etc.  These retailers competed primarily on Product: Wal-Mart might have 500 SKU’s of toys while Toys R Us had 50,000.  Effectively Toys R Us only had to be Price competitive on those 500 overlap SKUs and could charge what it wanted on the remainder that faced no real competition (this is an oversimplification for illustrative purposes).  People forget that Bed Bath & Beyond used to be the envy of the entire Retail industry with its mid-teens operating margin selling commoditized Product, but then nobody else in the physical world had the space in their stores to compete (funny enough, the Bed Bath CEO for that amazing run was Steven Tamares, whose background was Real Estate law).  E-commerce changed all of this and allowed consumers to buy anything they wanted from anyone outside of their physical location.  Best Buy survived but went through an exceptionally challenging business model transition period where Price and vendor relationships (akin to a light form of vertical integration) became important competitive factors.  Best Buy is an interesting case study itself as they were one of the first to flip conventional wisdom at the time and prove that physical stores can still matter in the Retail industry.  And now we have newer Category Killers based on a similar premise, such as At Home and Floor & Décor. 

In summary, investors somewhat got what they expected in terms of employee intensity with Digital disruption but not with margin expansion.  Nobody ever really addressed the competition subject back when they originally talked about the opportunity of Digital growth.  Just as all the current discussion towards A.I. seems to only focus on potential positives for the industry.  Now that we have looked at recent Retail history, let's attempt to see how A.I. disruption could “rhyme” in the future for our next piece. 

PS: There is so much I wanted to write on this subject but am keeping it tight because thats not the main focus of this series. One day...