Congrats to Amazon.
The chart shows a Mish calculation. I subtracted things one cannot or does not typically buy online from the total, then calculated the percent of nonstore sales of the remainder.
I exclude the obvious, like gasoline and eating out, but also motor vehicles and food store sales. One can buy groceries online but generally does not, especially fresh or frozen goods.
Nonstore sales spiked in Covid, and we have now finally taken out that pandemic high.
Nonstore Retail Sales as Percent of Advance Retail Sales Detail
Nominal Advance Retail Sales Motor Vehicles vs Nonstore
How long before nonstore is more important than motor vehicles?
Department Store Sales as Percent of Advance Retail Sales
The St. Louis Fed has not updated department store data for two months. I will contact them.
Do department stores serve any purpose now?
Related Posts
June 17, 2025: Retail Sales Down Much More than Expected, Drop 0.9 Percent
Retail sales declined 0.9 percent led by autos down 3.5 percent.
June 16, 2025: QCEW Report Shows Overstatement of Jobs by the BLS is Increasing
The discrepancy between QCEW and the BLS jobs report is rising.
June 15, 2025: Is the Fed Behind the Curve in Cutting Interest Rates?
That’s the claim in a Wall Street Journal op-ed.
Mish I have 2 questions about your assumptions.
1) If you order takeout (pizza delivery or Uber Eats) does that count as non-store sales since it’s ordered elsewhere? If it is it means restaurant ordering may count more than you think since an increasing number of people order food this way.
2) What’s considered groceries or a grocery order? I like to get my fresh veggies and meat in person just like you do but I am more than happy to do Walmart (or BJs Wholesale) orders for just about everything else food wise (canned goods, milk, bread, cheese, cereal, bottled water, liquor/beer etc) since there is nothing you need to be there in person to check. I’d say more than half my grocery order comes this way now and if this is considered non-store sales it also means your assumptions are far off in that industry.
I do not know how to answer your question
AI Overview
1. Identifying Nonstore Retailers: Nonstore retailers sell merchandise using methods other than traditional brick-and-mortar stores. This includes activities like broadcasting infomercials, direct-response advertising, electronic and paper catalogs, door-to-door sales, in-home demonstrations, selling from portable stalls (like street vendors), and using vending machines.
2. Data Collection: The Census Bureau collects data monthly from a sample of retail firms through surveys like the Monthly Retail Trade Survey (MRTS). These surveys include information about e-commerce sales, which are a major component of nonstore sales.
3. Weighting and Benchmarking: The collected data from the sampled firms are weighted and benchmarked to represent the entire universe of retail firms, including both store and nonstore retailers. This process ensures the estimates are representative of the overall retail landscape.
4. Imputation: For firms that don’t respond to the surveys, the Census Bureau uses imputation, a process of estimating missing data based on information from similar responding businesses within the same kind of business and sales size category.
5. Estimating E-commerce Sales: E-commerce sales are calculated separately from other nonstore sales. Horvitz-Thompson estimates are obtained by summing weighted e-commerce sales (including imputed data). These estimates are then benchmarked to previous estimates to ensure consistency.
6. Calculating Nonstore Sales Subtotals: For nonstore retailers, the Census Bureau calculates sales subtotals based on the primary business activity of the firm. These subtotals are then adjusted to align with the overall benchmarked monthly estimate for NAICS 4541 (Electronic Shopping and Mail-Order Houses), which is a key component of nonstore retailers.
7. Aggregating to Total Nonstore Sales: The various subtotals for nonstore retail activities are added together to arrive at the total estimate for nonstore sales.
8. Quarterly and Annual Estimates: Quarterly and annual estimates for nonstore sales are derived by summing the monthly estimates. The Census Bureau also adjusts these estimates for seasonal variation, holiday and trading-day differences.
Fantastic comment. The categories need to be revamped.
There’s a fair amount of auto parts slipping over into nonstore as well. Can order tires, wiper blades, fluids, various other parts online nowadays.
Gas won’t deliver or pump itself, and where I live there’s an incentive to return the spent batteries to a store for the refund, but that’s about it.
Google search: Average car price $46,700; Average years on road currently: 12.5
Average amount/yr/car: $3900
or $325/month.
So once the average spent in non-stores eclipses that, I’d say that’s when it becomes more important in dollar terms.
However the trend and nature of the spend in non-stores is growing, faster, and diverse in homeowners’ market basket. I’d say it’s already there in reality. Transactionally for sure. One purchase every 12.5 years vs many days per month in non-store.
The products I buy from non-stores are vastly more important, frequent, and growing larger. Because I’ve inherited 6 cats I’m already spending more in non-stores than on a vehicle. And as a consumer I can extend how long I keep a car by, ironically, purchasing parts from the non-stores.
In fact I just purchased $300 from Chewy, a lawn mower part, and a car part. One week. So for me and millions of others I’m sure non-stores have eclipsed autos.
This was a funny little thought experiment.