Where they spend has changed dramatically, with only two winners that gained share among the retailer categories. The rest lost share, some are getting wiped out.
By Wolf Richter for WOLF STREET.
Consumers spent $8.5 trillion over the past 12 months at brick-and-mortar and online retailers to purchase goods, from shoes and groceries to RVs. In June alone, they spent $720 billion. But how are consumers divvying up these trillions among the retailers? And how has it changed over the past 10 years?
The big winner is ecommerce, whose sales soared by 212% over the 10-year period. As a result, its share nearly doubled to 17.2% of total retail sales over the past 12 months through June, up from 8.9% in 2015, gaining 8.2 percentage points in share in 10 years; and up from a share of 5.4% in 2003, when retailers were still brushing off ecommerce as a threat to their brick-and-mortar business. Over the past 10 years, ecommerce became the #2 retailer category behind motor-vehicle and parts dealers, up from #5 in 2015 (dotted red line in the chart).
The only other winner: restaurants and bars (“food services and drinking places”). Sales soared by 88% over the 10-year period, causing the share to increase by 1.8 percentage points, to 13.5% over the past 12 months through June, up from a share of 11.6% in 2015, bypassing food and beverage stores and general merchandise stores to become the #3 retailer category. All other major categories lost share to ecommerce.
Over this 10-year period, total retail sales have increased by 63%. For retailers to maintain their share, their sales would have had to also increase by 63%. If sales increased by less than 63%, they lost share. If sales increased by more than 63%, they gained share.
Only two major categories increased their sales by more than 63% and gained share: ecommerce and restaurants and bars. They are eating everyone’s lunch, so to speak. Their gains are the result of big structural changes in how Americans spend their money and what they spend it on.
The #1 category is still motor vehicles and parts dealers (black line in the chart above). They sell new and used vehicles, recreational vehicles, motorcycles, ATVs, other motor vehicles, boats, tires, and parts. Over those 10 years, their sales rose by 51%, but that was less than the 63% increase of total retail sales, and so their share dropped to 19.2%, from 20.7% in 2025 (top black line in the chart above).
Those top three – motor vehicle and parts dealers, ecommerce, and restaurants and bars – accounted for half (50%) of total retail sales over the past 12 months through June, up from a share of 41% in 2015.
Sales at restaurants & bars, the #3 category (purple in the chart above), increased by 88% since 2015, while overall retail sales increased by only 63%. They blew by food and beverage stores in 2019.
During the pandemic, many restaurants were shut down, and sales collapsed, while grocery store sales spiked as people couldn’t eat out anymore. But then the restaurant industry re-opened more vibrant than ever and left food and beverage stores in the dust.
Restaurants benefited from the post-pandemic “revenge spending,” when people wanted to get out of the house and have “experiences,” and this trend, despite predictions that it would fizzle in 2023, has continued full force, as consumers are in the mood to splurge, and have the money to do so, with record high earnings and historically low unemployment.
Food and beverage stores dropped to #4 (double green line in the chart above), as their sales increased by 48% since 2015, while overall retail sales increased by 63%.
In 2015, they were #2. They were bypassed by ecommerce during the lockdown and by restaurants and bars in 2022.
General merchandise stores and ecommerce ate their homework. Food and beverage stores lost sales to general merchandise stores that also sell food, such as Walmart (now the largest grocery seller). They also lost sales to ecommerce, including to online-only food retailers. And they lost sales to restaurants.
General merchandise stores dropped to #5 now, from #3 in 2015, and their share declined to 10.7% now, from 13.0% in 2015. Over those 10 years, their sales increased by only 35%, while total retail sales increased by 63%.
These brick-and-mortar stores lost a large part of their nonfood sales to ecommerce retailers (such as Amazon), or shifted sales to their own ecommerce channels (such as Walmart, now one of the largest ecommerce retailers in the US). Part of the brick-and-mortar stores’ loss of sales to ecommerce of their nonfood items was covered up by their increased food sales, at the expense of grocery stores.
Department stores were once the iconic way in which Americans shopped. Their slow demise started quietly in 2001 when their sales began their relentless decline.
Ecommerce brought about a structural change in how Americans shopped, and it pulled the rug out from under department stores (brown line at the very bottom of the first chart).
Here is the long-term sales performance of brick-and-mortar department stores. This does not include the ecommerce sales of the few remaining department store chains, only their brick-and-mortar sales:
In 1992, department stores had a share of over 9% of total retail sales. By 2004, their share was down to 5.4%. By 2015, their share was down to 2.0%. Now their share is down to 0.46%. They really don’t matter anymore.
As this shift occurred, countless independent department stores, regional chains, and national chains were liquidated in bankruptcy court, while the few surviving chains closed thousands of stores. Since 2016, WOLF STREET’s Brick-and-Mortar Meltdown has documented the fate of many of these department stores along with other brick-and-mortar stores.
The share of gas station sales, the #6 category, dropped to 7.3% over the past 12 months, from 8.6% in 2015, largely driven by the price of gasoline (blue line in the middle, first chart).
But there is also the long-term issue that gasoline consumption per capita has been declining for over 20 years, as a result of fewer miles driven per capita, and ever more fuel-efficient vehicles (including now increasingly EVs).
In 2024, gasoline consumption in the US was 3.9% below the 2018 peak and 3.4% below the prior peak in 2007, and about where it had been in 2003 (details and charts are here):
In case you missed it: My Thoughts about those June Retail Sales
Enjoy reading WOLF STREET and want to support it? You can donate. I appreciate it immensely. Click on the mug to find out how:
Just curious – if someone orders bags of garden soil (or any other inexpensive bulky item that is impractical to ship) online at Home Depot or Lowe’s website with a same day store pickup, is this attributed to e-commerce or brick&mortar in the category?
Most people would still buy those at Home Depot or Lowe’s, they are just using a website as a convenience tool to make sure they have their stuff reserved for them for when they come to the store, or because they want to use an online coupon or to save some legwork on searching small items in a store.
I guess, my question here is if this is truly a shift from b&m to e-commerce, or if it is just an added convenience, by had it not been offered by retailers, customers would’ve still shopped in store?
This also somewhat “explains” why car dealers are going strong – because they are not too much affected by e-commerce (except for Tesla or whatever new models like Carvana/Vroom exist)
When you buy online and pay online, it’s ecommerce. It doesn’t matter how the product gets inside your house. Whether someone delivers it, or whether you pick it up at a locker or at a store doesn’t matter.
The delivery business is a service and belongs to the transportation industry, not retail. By picking it up yourself, all you’re doing is taking share away from a transportation service provider.
“This also somewhat “explains” why car dealers are going strong”
No, they’re not going that strong. They lost share in 10 years, their share dropped to 19.2% now from 20.7% 10 years ago, as the article points out. Some EV makers and used vehicles dealers are doing brisk business online. But the state franchise laws still protect franchised dealers from their own automakers selling online directly to consumers.
So when people order groceries to pick up, that gets put into the e-commerce category? That’s fine, but obviously a huge chunk of that green line has now been shifted to the red line, so that kind of alters the original perception of grocery buying being lower than restaurant buying and e-commerce, right?
Under the second chart, there is this paragraph:
“General merchandise stores and ecommerce ate their homework. Food and beverage stores lost sales to general merchandise stores that also sell food, such as Walmart (now the largest grocery seller). They also lost sales to ecommerce, including to online-only food retailers. And they lost sales to restaurants.”
I worked commission sales in a major luxury department store from 1983 to 2015, and made a good if not great living at it. My experience mirrors the Department Stores Graph above perfectly. Sales just kept steadily rising up into the early 2000s; we then inexorably went from flying to driving to running to walking to crawling in the next two decades, and, no matter what alternate, rationalizing explanation someone might have for you, it was 90% ecommerce driven.
Disagree somewhat. I assert that a sizable drop in luxury department store sales was also due to the younger generations who have little interest or ability to partake of that lifestyle. Yes, ecommerce has played a role, but there is also a demographic piece that was involved.
My college roomate made great money as an undergrad working at Nordstrom part time in the early 80’s (so much that they flew him to talk to other part time salespeople out of state to talk about about how he created a strong customer base by writing personal letters to his good customers). E Commerce is a big reason for the drop but as as @Mike R. says most kids today are not interested in what Nordstrom, Nieman Marcus and Macys are selling. My college age son was just commenting that he thinks even less guys under 30 today can tie a tie than drive a manual transmission and that there are probably more certified SCUBA divers under 30 than guys that can tie a bow tie (I taught my son “and” daughter to tie a bow tie in High School and I still wear the Robert Talbott self tie bow tie my roommate got me with his employee discount 40+ years ago to black tie events that seem to come up every year ow two)
Wolf – great article on the continuing evolution of ‘Main Street’ ‘Murica! Thanks again for your creation and bartending of this most-excellent establishment!
may we all find a better day.
Topmost, 10 yr chart is very interesting and useful. Long term data tends to reveal important trends that frequently get lost in the weeds of short term data.
One unfortunate factor, ….
I deleted your post because it was just too much BS piled into one heap.
But I will shoot down ALL of it by telling you that ecommerce = retail sales of GOODS online, and does NOT include the sale of services, such as lodging, rental cars, plane tickets, reservations, subscriptions, insurance, etc., and does NOT include online B2B sales, etc.
Ecommerce = retail sales of goods only. Nothing else.
Re: “ The rest lost share, some are getting wiped out.”
I recently stumbled on the topic of elevator strikes, and have been pondering AI implementation, as a metaphor.
“The Elevator Strikes were a series of labor strikes that took place from the 1920s to the 1960s across the United States”.
As a first step, it’s easy to understand the impact the internet had on department stores, and commerce, in terms of efficiency. Consumers were easily influenced by lower online prices, and cared less and less about retail sales spaces and people.
After twenty years of infrastructure buildout and conditioning — the pandemic forced dominate global adoption of online interactions — accelerating the cord cutting — not too different from smartphone adoption and the extinction of phones with cords.
Back to elevators, even though elevator operators became extinct — elevators essentially serve the exact purpose one hundred years later, moving up and down, ferrying people and stuff in between floors of buildings.
Now, with AI, and post pandemic dynamics, including fewer people in buildings — I think there’s a curious concern unfolding about an epic societal transformation related to technological transformation.
In a weird perspective, the extinction of elevator operators is evolving towards the extinction (decline) of office space and hence people in tall buildings — and without expanding this too far, the function of buildings and cities are undergoing a transformation that will be supercharged by AI.
I think that process will eventually mimic the elevator strikes, as people are forced to adopt new ways of living, that seem alien (as many people lose jobs to automation).
As with the elevator timeline, that process of change took multi-decades, but more than likely, we’re dealing with multi-years.
The evolution of commerce and consumer spending will undoubtedly be connected to efficient supply chains that have fewer people in the loop — which brings me the primary question:
How do the monkeys, squirrels and drunken sailors make money and drive spending, in a continuous parabolic upward spike — especially if a tsunami of jobs are headed to extinction?
I’ll find a better day all right…
Nothing better than to wait around for a better day. But with e- commerce, instead of meeting my new bride at the bar, now I can go online and order one. Not sure she will provide a ” better day”.
But looking for a date/bride online is NOT part of ecommerce. Ecommerce = retail sales of goods online.
Department stores data makes no sense.
Only Kohl’s has like $3B revenue per quarter with 30% being online, so the rest $2.1B is brick and mortar.
$8B per year, only Kohl’s!
It does make sense. All sales charts are monthly sales. So multiply by 12 to get close to annual sales. Total department store sales over the past 12 months = $40 billion.
Department stores are narrowly defined by store type, not by brand.
There are only a handful of department store chains left, all of them have shifted a big portion of their business to ecommerce as a survival strategy because that’s what is doing well. Macy’s and Nordstrom have been over 1/3 ecommerce for many years.
Also outlet stores and discount stores from department store brands are NOT department stores.
“Small format” stores from department store brands are NOT department stores.
Other businesses that department store brands have, such as restaurants, are not included in department store sales. For example, 13% of Macy’s total revenues are from “other” which is primarily restaurants.
Fees that department stores earn, such as on branded credit cards, are also not part of department store sales. For example, 4% of Macy’s total revenues are from fees, including credit card revenues and its media channel.
So you take ecommerce (33%), fees (4%), and restaurants and other (13%) out of Macy’s revenues, and what’s left is 50% of Macy’s total revenues that may be department store sales.
(I’m using Macy’s because I’m a lot more familiar with it than Kohl’s and have followed it for many years).
In 2006,the CEO of BP predicted the gasoline volumes have peaked and would gradually decline in the coming years…
Americans are addicted to cars. Literally. Nothing else can explain how an entire country can continue to spend what we do on private transportations.
And of course, this addiction is well understood by the automakers and all downstream players.
If Americans stopped buying these overpriced, pimp mobiles enmass, automakers would start providing cheaper, more economical (but still comfortable by any reasonable standard). And many would go out of business as the entire industry is bloated.
This will happen one day when things implode, but until then, happy motoring.
Americans are addicted to buying new cars. Not many of us out there that are passionate about keeping 10-20 year old cars on the road.
That’s why cars aren’t as reliable as they used to be – the manufacturers know they don’t need to be. Why build an engine to last 200-300k miles when you know the original buyer will probably trade it in before 100k? Computer-aided part design and materials engineering is also a factor since you can more precisely design parts to last 100k miles without over-engineering them.
The vast majority of today’s auto engines will easily go 200,000+ miles with normal maintenance.
@Mike R. Americans are not “addicted” to cars and I would say that only a small percentage of America is like me and gives a crap about cars (my son estimates that less than half the young men his age even know what kind of engine is in their car). Public Transit is not time efficient for most people. Google says it will take me 1/2 hour to drive to downtown SF but “just over” a full hour if I take public transit. If I drive to Bolinas in Marin it will take about an hour and a half, If I take public transit Google says it will take almost 4 hours.
All of the components in the first graph can be and are purchased under the category “ecommerce” except gas stations, and maybe restaurants and bars, although many get restaurant food delivered by ordering it on-line. Even some cars can be bought on-line, e.g., Tesla, Carvana. Maybe a better delineation would be “stuff you can buy at home without moving your lazy a&% to a store” and “stuff where you have to go to a store to buy it”.
It’s interesting that department stores such as Kohls, Macy’s, Dilliards, etc., which mainly sell clothing, are losing market share, but discounters like TJMax and Marshalls and Ross, etc., which also primarily sell clothing, are doing great.