The 2026 US holiday season is likely to set a record for how much retail sales each retail worker carries. The specific call, and the one this piece should be judged on: nominal holiday-season retail sales per retail-trade employee likely rise roughly 4% or more year over year in the November 2026 through January 2027 window, setting a series high. The check arrives by mid-February 2027, when the Census Bureau publishes its final December retail sales report and the Bureau of Labor Statistics completes its fourth-quarter retail employment picture. Three signals observed between August 26 and September 24, 2026 point the same way, and the tension between them is the whole story.
That tension is simple to state. The demand forecast for this holiday season accelerated in early September, the seasonal hiring forecast fell in late September, and the parcel network spent late August pricing for a volume surge it has said little about staffing. When the numerator rises and the denominator does not, the ratio between them is where the signal lands.
In short
- The prediction: US holiday retail sales per retail-trade employee likely set a series record this season, rising roughly 4% or more year over year in nominal terms.
- The timeframe: verifiable by mid-February 2027, using Census holiday retail sales against BLS fourth-quarter retail-trade employment.
- Signal 1: Deloitte forecast holiday retail sales of $1.70 trillion to $1.71 trillion, up 4.0% to 4.8%, with e-commerce up 7.5% to 8.4%.
- Signal 2: Challenger, Gray & Christmas forecast 450,000 Q4 retail hires on September 23, down from an already depressed 461,500 in 2025 and 543,100 in 2024.
- Signal 3: UPS published a demand surcharge schedule on August 26 with flat service-level charges up roughly 22% to 25%, while its chief executive has pointed to US volume rising about 24% from Q3 to Q4.
- The main risk: this is a nominal ratio, and tariff-driven price inflation plus a shrinking statistical denominator could manufacture much of the record without any real productivity gain.
Why this matters now
Most holiday previews ask whether retailers will hire enough people. That question has been asked and largely answered, and the answer has been the same for three years running. The more useful question for the next six months is what happens to output per worker when demand keeps growing and headcount does not.
The answer matters because it is the mechanism behind several things retailers will report in February and March 2027. Gross margin resilience, store labor commentary, and the recurring claim that automation investment is paying for itself all run through this ratio. If sales per worker sets a record, executives are likely to describe it as productivity, and that framing will shape capital allocation into 2027.
It also matters because the ratio is measurable, which most holiday narratives are not. Two federal series, both published on a known schedule, settle it without requiring anyone to trust a company’s characterization of its own labor model. That makes it a rare holiday prediction that a reader can check rather than relitigate.
Signal 1: the demand forecast accelerated while the labor forecast fell
Deloitte released its holiday retail forecast in early September 2026, projecting sales of $1.70 trillion to $1.71 trillion for the November 2026 through January 2027 period. That represents growth of 4.0% to 4.8% over the same period a year earlier, according to the firm’s press materials. The prior season grew 4.1%, so the midpoint of the new range points to acceleration rather than a slowdown.
The e-commerce component is the sharper part of the number. Deloitte put holiday e-commerce at $316.1 billion to $318.9 billion, growth of 7.5% to 8.4% year over year. That is roughly double the pace of the total, which means the channel mix is still shifting toward fulfillment-intensive orders. Per the firm’s stated methodology, disposable personal income growth of 4.5% to 5.2% is the primary input behind the forecast.
What makes this a signal rather than background is its direction relative to the labor forecast that followed it two weeks later. A demand forecast that accelerates is unremarkable on its own. A demand forecast that accelerates while the corresponding hiring forecast declines is the setup for a productivity print, because the sales have to be moved by someone or something.
The caveat worth stating early: Deloitte’s figure is nominal, and so is the prediction built on it. Some meaningful share of the 4.0% to 4.8% is price rather than units, particularly in tariff-exposed categories. That does not break the ratio, but it does change what the ratio means, a point developed in the caveats below.
There is also a measurement subtlety worth flagging for anyone planning to check this later. Deloitte defines its holiday window as November through January, which captures gift card redemption and post-Christmas clearance. The Census Bureau convention most commonly cited for holiday sales is November through December. A reader running the verification should hold the window constant across years rather than mixing conventions, because the January tail has been growing.
The forecast’s dependence on disposable personal income is the part most likely to break. Deloitte attributes its range primarily to projected income growth of 4.5% to 5.2%, which is a reasonable predictor in ordinary years. It is a weaker predictor when the savings rate is moving or when confidence turns late in the quarter. If income growth lands at the bottom of that range, the sales forecast likely drifts toward 4.0% rather than 4.8%.
Signal 2: the seasonal hiring series is heading for a third straight decline
Challenger, Gray & Christmas published its 2026 holiday hiring outlook on September 23, 2026, forecasting 450,000 retail jobs added in the fourth quarter. The framing in the title of the firm’s own outlook is instructive: the 2025 season was already the lowest since 2008. The new forecast points to a third consecutive annual decline rather than a rebound from a trough.
The multi-year shape is what carries the argument. The series moved from 543,100 in 2024 to 461,500 in 2025, a decline of roughly 15%, and the 2026 forecast of 450,000 would extend the slide. Transportation and warehousing told a similar story, adding 266,500 positions in Q4 2025, down about 12% from 2024 and the lowest reading since 2018’s 259,500.
| Measure | 2024 | 2025 | 2026 |
|---|---|---|---|
| Q4 retail jobs added | 543,100 | 461,500 | 450,000 (forecast) |
| Year-over-year change | n/a | about -15% | about -2.5% (implied) |
| Q4 transportation and warehousing | about 302,000 (implied) | 266,500 | not separately forecast |
| Series context | above trend | lowest since 2008 | third straight decline |
The outlook also notes something that complicates any hiring tally: several of the largest employers have said nothing. Amazon, Target, Bath & Body Works and Kohl’s had not published 2026 commitments as of the September 23 outlook, which the firm reads as continued reliance on flexible staffing rather than large seasonal waves. The announcements that did land were mid-sized, including Spirit Halloween at 52,000 associates and Michaels at more than 10,000 seasonal team members.
Macy’s, Inc. reinforced the pattern the following day. Its September 24, 2026 press release announced seasonal hiring across Macy’s, Bloomingdale’s and Bluemercury stores plus distribution and fulfillment centers, describing part-time store roles and full- and part-time supply chain positions. The release contains no headcount number at all, and no comparison to prior years. This desk has argued before that the quiet announcement channel means the headline figure likely overstates the true decline in Q4 retail labor, which is precisely why the productivity ratio is the cleaner thing to measure.
The identity of the employers still publishing numbers is itself a pattern. Spirit Halloween operates a business that exists only during a season, so it has no permanent workforce to flex and no option but to hire and announce. Michaels sits closer to the same position, running a craft calendar that peaks sharply and staffs against it. Businesses with no base workforce to lean on are structurally obliged to disclose a seasonal intake.
The employers going quiet are the ones with the largest permanent payrolls and the most developed scheduling technology. That asymmetry means the announcement-based series is not simply getting smaller, it is getting biased toward employers who cannot substitute hours for heads. A tally that increasingly samples only the businesses that must hire will understate total peak labor, while a productivity ratio built on federal payroll data sidesteps the selection problem entirely.
Signal 3: the parcel network is pricing for a volume surge it has not staffed for
UPS published its updated demand surcharge schedule on August 26, 2026. Flat service-level charges rose approximately 22% to 25% versus the prior season, while handling and size-based demand charges rose roughly 6% to 10%. The demand surcharge on Ground Residential, Air and Ground Saver runs $0.50 to $2.50 per package, with high-volume shippers exposed to a band reaching $9.35.
The calendar of those fees is itself informative. Additional handling, large package and over-maximum-limits charges run September 27, 2026 through January 16, 2027, while the broader residential demand surcharge starts October 25 and runs to the same January end date. Peak pricing concentrates between November 22 and December 26, 2026. FedEx announced its domestic holiday increases in July on a closely parallel schedule, with nonstandard package fees from September 28 and most residential surcharges from October 26.
The volume expectation sitting behind those fees is the part that matters here. UPS chief executive Carol Tomé has pointed to US volume rising about 24% from the third quarter to the fourth, broadly mirroring the prior year’s step-up. A network expecting a roughly quarter-sized volume increase, pricing it through per-parcel fees rather than announcing a matching labor intake, is describing a throughput plan rather than a hiring plan.
That repricing is a story this desk has tracked separately, because the all-in delivered cost per parcel is likely to outrun the published general rate increase by a wide margin this peak. The labor read is the mirror image of the cost read. The same automated network that lets carriers charge more per parcel is what lets them move more parcels per employee.
The early start date deserves more attention than it usually gets. Oversize and additional-handling fees beginning September 27 and 28, nearly two months before the true volume peak, function as a demand management tool rather than a revenue grab. Charging for awkward freight early pushes shippers to reconfigure packaging before the network is stressed. That is capacity creation through pricing, and it substitutes directly for capacity creation through hiring.
The surcharge bands point the same way. A spread running from $0.50 to $9.35 per package for high-volume shippers is steep enough to change behavior rather than merely tax it. Shippers who flatten their volume curve and standardize their cartons pay the bottom of the band, and in doing so they hand the carrier a more automatable parcel mix. The pricing and the automation reinforce each other.
| Signal | Date | Source type | What it measures | Direction |
|---|---|---|---|---|
| Holiday sales forecast | Early Sept 2026 | Consultancy forecast | Numerator: demand | Up 4.0% to 4.8% |
| Seasonal hiring outlook | Sept 23, 2026 | Outplacement research | Denominator: seasonal heads | Down to 450,000 |
| Headcount-free hiring release | Sept 24, 2026 | Company press release | Disclosure behavior | Number withdrawn |
| Carrier demand surcharges | Aug 26, 2026 | Carrier rate schedule | Peak volume expectation | Fees up 22% to 25% |
What the pattern suggests
Put the three together and the arithmetic does most of the work. Demand is forecast up 4.0% to 4.8% in nominal terms, with the fulfillment-heavy e-commerce slice up 7.5% to 8.4%. Seasonal headcount is forecast down for a third year, and the largest employers are not committing to numbers. The volume still has to be received, picked, packed, shipped, sold and returned.
The mechanism is substitution, and it appears to have three parts. Flexible hour pools let existing employees absorb peak demand without appearing in any seasonal hiring tally, which is why Target’s on-demand pool of roughly 45,000 workers is prioritized ahead of new hires. Automation absorbs a second slice, particularly in warehousing. Price absorbs a third, by rationing demand toward the periods and package profiles a network can handle.
The calendar mildly reinforces this. Thanksgiving falls on November 26, 2026, leaving 28 shopping days through December 24, against 27 in 2025 and 26 in 2024. A longer runway spreads the same volume across more operating days, which reduces the single-day peak that seasonal headcount exists to cover. Longer seasons favor hours-based flexing over headcount-based staffing.
None of this requires anyone to be doing anything unusual. It requires only that the substitution trend of the past two seasons continues at roughly its current rate while demand grows at the forecast pace. That is why the prediction is framed as a record rather than a discontinuity: the ratio has likely been climbing for two years, and 2026 looks like the year it clears the prior high with room to spare.
It is worth making the arithmetic explicit, using index values rather than absolute levels to keep the logic visible. Set the 2025 season at 100 for both sales and retail employment. Apply the forecast sales growth and a flat to modestly negative employment path, which is what the Challenger series and the August payroll reading jointly imply.
| Index (2025 = 100) | Low case | Mid case | High case |
|---|---|---|---|
| Holiday sales | 104.0 | 104.4 | 104.8 |
| Q4 retail employment | 100.0 | 99.5 | 99.0 |
| Implied sales per worker | 104.0 | 104.9 | 105.9 |
| Year-over-year change | 4.0% | 4.9% | 5.9% |
The table shows why the call is framed at roughly 4% or more rather than at a precise figure. Even the least favorable combination, sales at the bottom of the forecast range with employment perfectly flat, still clears 4%. The prediction fails only if sales undershoot the published forecast range or employment rises, which is exactly what the falsifier section specifies.
This is also why the call is more robust than it might first appear. It does not depend on automation working, on any company’s guidance, or on a particular view of the consumer. It depends on the forecast range being roughly right and on retail employment not growing, which would be a reversal of a three-year trend.
Wider context: the denominator is moving too
The employment side of this ratio is unusually unstable right now, which cuts both ways for the prediction. The August 2026 Employment Situation described retail trade employment as little changed over the month, with overall payroll gains concentrated in food services and health care. More consequentially, the annual benchmark revision cut retail trade by 154,600 jobs, the largest downward reduction of any sector.
A denominator that gets revised down by more than 150,000 jobs mechanically raises sales per worker without a single additional item being sold. An honest version of this prediction has to name that, and it is why the caveats section treats the revision as a live threat to the interpretation rather than a supporting detail. The September Employment Situation, scheduled for October 2, 2026, will be the next read on whether retail employment is flat or genuinely eroding.
The automation backdrop is older than the 14 to 30 day signal window but it frames the trend. Amazon crossed one million deployed robots in mid-2025, against a workforce of roughly 1.5 million. Reporting on leaked internal planning documents has described an ambition to automate up to 75% of operations by 2033, avoiding roughly 600,000 additional US hires, with savings framed around 30 cents per shipped package.
Wages are the other moving part, and they push against the cost-saving reading. Amazon’s decision to raise minimum pay to $20 an hour in a $1.5 billion commitment means fewer workers does not automatically mean cheaper labor. Sales per worker can set a record while labor cost per hour also rises, and both can appear in the same quarterly filing.
The store and warehouse halves of the ratio are also likely to diverge, which matters for interpretation. Warehousing has the clearer automation path and the sharper measured decline, with Q4 2025 transportation and warehousing additions at their lowest since 2018. Store labor is harder to automate away, because shelves, fitting rooms and checkout still absorb hours that no robot currently takes. A blended record could therefore conceal a warehouse productivity surge sitting alongside flat store productivity.
Implications for retailers, operators and investors
For retailers, the practical consequence is that peak performance in 2026 likely depends more on scheduling systems than on recruiting funnels. If the flex pool is the primary peak instrument, its accuracy becomes the operational risk. Under-flexing on a single high-volume weekend is harder to correct than under-hiring in October, because there is no bench to call.
For logistics operators, the surcharge schedule sets the shape of the quarter. Fees that begin September 27 and peak from November 22 reward shippers who can flatten their own volume curve. Operators that have already converted variable labor into fixed automation capacity capture the spread between per-parcel pricing and per-parcel cost, which is the same dynamic that made warehouse hiring hold up better than store-based seasonal hiring through recent seasons.
For investors, the February and March 2027 reporting round is where this becomes tradeable rather than interesting. The specific thing to listen for is whether management attributes fourth-quarter gross margin to pricing, to mix, or to labor productivity. A productivity attribution, backed by a record sales-per-worker ratio, likely supports continued automation capex into 2027.
For brands and marketplace sellers, the read-through is about service levels rather than headlines. A retail partner running peak on a thin flex pool has less slack to absorb a late purchase order, a re-ticketing request or an off-profile carton. Sellers who lock in carton standards and delivery appointments early are likely to see materially better treatment than those negotiating in the last week of November.
For anyone benchmarking internationally, the pattern is not uniquely American. UK retailers have been cutting stated Christmas intakes as well, with John Lewis trimming its Christmas hiring to 10,400 roles at stated pay of £13 to £14.80. The direction travels, though the labor market institutions differ enough that the ratios are not directly comparable.
Scenarios and what would falsify this
A prediction that cannot fail is not worth publishing, so here is the scenario grid with explicit falsifiers. The base case assumes both forecasts land near their midpoints and the substitution trend holds.
| Scenario | Rough likelihood | What we would observe by Feb 2027 | Falsifier for this call |
|---|---|---|---|
| Base: record sales per worker | Most likely | Holiday sales up 4% or more, Q4 retail employment flat to down | None; this is the call |
| Demand disappoints | Plausible | Holiday sales up under 3%, hiring still weak | Ratio rises under 4%, call partially wrong |
| Late hiring surge | Less likely | November and December retail payrolls beat, Challenger revised up | Ratio flat or down, call wrong |
| Classification shift dominates | Plausible | Retail trade shrinks as roles move to warehousing NAICS | Record is real but meaningless, call hollow |
The cleanest single falsifier is straightforward. If Q4 2026 retail-trade employment rises year over year while holiday sales growth comes in below 4%, the ratio likely stalls and this call fails on its own terms. A reader can settle that with two public releases and a calculator.
Caveats: what could go wrong
The strongest objection is that this is a nominal ratio dressed up as a productivity claim. If a meaningful share of the 4.0% to 4.8% sales growth is tariff-driven price increases rather than unit growth, then sales per worker rises without anyone moving more goods. Deflating by a goods price index could erase most of the record, and the prediction as stated would be technically right and analytically empty.
The second objection is the denominator problem described above. A benchmark revision that removed 154,600 retail jobs raises the ratio by construction, and further revisions could move it again in either direction. Statistical artifacts and genuine productivity are difficult to separate in real time, and readers should treat any single-year record with that in mind.
The third is classification drift. As e-commerce grows faster than total retail, work migrates from retail trade into transportation and warehousing in the statistical categories, shrinking the denominator for reasons that have nothing to do with efficiency. The e-commerce forecast of 7.5% to 8.4% growth makes this a live concern rather than a theoretical one this season.
The fourth is that the substitution may be reaching its limit. Flex pools and automation have absorbed two consecutive seasons of growth, and there is no guarantee they absorb a third at the same rate. Service degradation is the tell: if conversion rates fall or fulfillment times slip, retailers could add headcount late in the quarter, which would show up in December payrolls and undercut the call.
The fifth is simply that Challenger’s forecast is a forecast. The firm publishes an outlook in September and a tally in January, and the gap between them has been wide in some years. Announcement-based tallies are also increasingly incomplete, as the Macy’s release demonstrates, so the hiring side of this argument rests partly on a measure that is itself degrading.
Frequently asked questions
What exactly is being predicted here?
That nominal US holiday retail sales per retail-trade employee likely set a series record for the November 2026 through January 2027 season, rising roughly 4% or more year over year. It is a ratio prediction, not a sales prediction or a hiring prediction. Both underlying series are public.
How would a reader verify it?
Take Census Bureau holiday retail sales for the season, published in final form with the December report in mid-February 2027, and divide by BLS fourth-quarter retail-trade employment. Compare the result to the same calculation for prior years. The employment series is published through the BLS Current Employment Statistics program.
Is this just a restatement of the weak hiring story?
No, and the distinction matters. The weak hiring story is about the denominator alone and has been widely covered for three seasons. This call is about what happens when an accelerating numerator meets that falling denominator, which is a different and more checkable claim.
Could the record be an accounting illusion rather than real productivity?
Yes, and that is the most serious counter-argument. Price inflation, benchmark revisions and NAICS classification drift could each produce a record ratio without additional goods moving per hour worked. A real-terms check, deflating sales by a goods price index, is the appropriate follow-up test.
Why does a longer shopping calendar support the prediction?
Because seasonal headcount exists largely to cover peak-day capacity rather than total volume. The 28 shopping days between Black Friday and December 24 in 2026, against 26 in 2024, spread the same demand across more operating days. Flatter peaks favor flexing existing employees over hiring temporary ones.
What would make this call clearly wrong?
A late hiring surge showing up in November and December payrolls, combined with holiday sales growth below 4%, would likely leave the ratio flat or lower. Equally, a demand miss driven by consumer resistance to tariff-affected prices would undercut the numerator. Either outcome is visible in the same February 2027 data.
Does fewer workers mean lower labor costs for retailers?
Not necessarily, and assuming so is a common error. Hourly wages have been rising even as headcount flattens, so total labor cost can grow while sales per worker sets a record. The two measures answer different questions and can move in the same direction at once.
Why treat a press release with no number as evidence?
Because the absence is the data point. A company that published headcount targets in prior seasons and announces hiring without one this season has changed what it is willing to commit to publicly. One release proves little, but it corroborates the pattern of large employers staying silent that the September 23 outlook independently described.
How much of this is specific to Amazon?
Less than the automation coverage implies. Amazon is the clearest case because its robot deployment is disclosed, but the mechanism runs through department stores, specialty chains and carriers alike. The Macy’s release and the UPS surcharge schedule are both non-Amazon evidence of the same substitution.