Data story
Payday
Do Americans buy sneakers when the paycheck lands? We lined up 514 days of resale sales against the pay calendar. The days after the 15th and the end of the month run about 6% busier. The 1st on its own, benefit days and prices: not much at all.
YZY YS-01 Black (Under $100); ASICS Gel-1130 Black Pure Silver ($100–$200); Jordan 4 Retro Black Cat (2025) ($200+). Photos: StockX
Source: SneakerPulse records of completed GOAT and StockX footwear sales, January 2025 to September 2026. Photos: StockX.
Twenty months of sneaker sales, one dot a day
Each dot is one day of completed GOAT and StockX footwear sales, from January 3, 2025 to September 16, 2026, by US Eastern date: 514 days, 13,047,854 pairs. Orange dots are the 1st and the 15th.
Big drops make their own spikes, so we took out every shoe's launch window, from 30 days before release to four weeks after (2,504,987 sales). What is left is the everyday market.
The line is a normal day for that time of year: it rises and falls with the seasons and with the set of shoes we track. Left out: StockX's collection gap (May 9, 2026 to July 3, 2026), 7 other days when a feed was not fully loaded, and holidays (Thanksgiving to Cyber Monday, Christmas to New Year, long weekends). Hover or tap a dot.
Divide by a normal day
To compare a Tuesday in March 2025 with one in August 2026, divide each day by the median of the 28 days around it. A day at +10% sold 10% more pairs than a normal day of that month.
That removes changes in our coverage and the seasons (tax-refund season, back to school, the holidays) and leaves the rhythm of the month. 72% of days land within 10% of normal.
A calendar of busy and quiet days
Now fold those days into a calendar: one row per weekday, one column per date. Teal is busier than normal, brown quieter.
The rows tell the first story: Thursdays and Fridays run teal, Sundays and Mondays brown. The columns tell the second: look at the 1st, and at the 15th to 17th.
Use the menus above the chart to pick a platform, a brand or a price band. A shoe's band is its typical resale price, for example:
YZY YS-01 Black
typical $51 · Photo: StockX
ASICS Gel-1130 Black Pure Silver
typical $122 · Photo: StockX
Jordan 4 Retro Black Cat (2025)
typical $309 · Photo: StockX
Each cell averages two to four days, so single cells are noisy. The next steps add intervals.
The 1st and the 15th stand out
Take the weekday out (a regression with a weekday effect and a date effect) and the dates line up. The 1st sold +5.1% more pairs than the average day; the 15th, 16th and 17th +4.0%, +6.1% and +7.3%.
Between paydays, sales sag: the 9th to 12th ran −2.8% to −4.0%, the 19th −4.1%.
Read single dates with care. With 31 dates tested, only the dips on the 11th, 12th, 19th pass a false-discovery correction (q < 0.05); the 1st (q = 0.085) and the 17th (q = 0.085) come close. The sharper test uses the days people are actually paid: next.
Bars: each date vs the average day of the same weekday; line: 95% interval from a bootstrap that resamples whole weeks. Faded bars are not significant.
The three days after payday run about 6% busier
We tested common US pay schedules directly: payday (moved to the business day before when it falls on a weekend or bank holiday) plus the two days after, against every other day, weekday taken out.
Paid on the 15th and last day: +5.6% (95% interval +2.9% to +8.4%). 1st and 15th: +5.1%. Last business day: +4.2%. The 15th alone: +5.4%. The 1st alone: +3.2% (95% interval −0.5% to +6.9%), not clearly different from zero.
Social Security's Wednesdays: −1.5%; the 3rd: +1.8%. Nothing.
Fridays sold +5.5% more than Monday to Thursday, but Thursday was the busiest day of all (+7.6% vs the average day; Friday +5.9%, Sunday −4.5%, Monday −5.5%). Workers paid every other Friday are paid on different Fridays, so in a national total that pay is spread over every Friday and cannot be told apart from the weekly rhythm of shopping.
All eight tests share one Benjamini-Hochberg correction. GOAT alone: 15th and last day +5.9%, the 1st +3.6% (q = 0.031). StockX alone: 15th and last day +5.5%, Fridays +1.3% (q = 0.285).
No payday premium
If payday brings eager buyers, do they pay more? For 40,355 listings we compared each day's sale prices with the same shoe's price over the 28 days around it.
Barely. In the three days from the 1st, the same shoes sold for +0.39% (95% interval −0.04% to +0.83%). Every date sits within about 2% of normal.
The one clear move goes the other way. On StockX the 16th and 17th sold 1.51% and 2.03% below normal (q = 0.003 for the 17th), in the busiest stretch of the month; GOAT showed no dip (−0.25%, −0.25%). Busier days at lower prices look more like extra sellers, or a StockX promotion cycle, than buyers flush with cash. We cannot tell which.
Same listing, so the mix of shoes cannot drive this; sizes can (GOAT also sells used pairs). Weekday prices moved even less: Monday +0.53%, Tuesday to Thursday −0.23%.
SNAP states show no first-of-month bump
Many government payments land early in the month: SSI on the 1st, older Social Security on the 3rd, and SNAP food benefits, which most states stagger over the first one to three weeks. SNAP buys only food, but it frees up cash.
If that cash went to sneakers, states with more people on SNAP should do more of their buying on the 1st to 3rd. Using GOAT sales with a buyer state (460 days, 4,637,803 sales), they do not. The fitted line is flat: +0.10 points per point of SNAP participation (95% interval −0.44 to +0.63).
Top fifth of SNAP states vs bottom fifth: +1.1% (interval −2.9% to +5.4%). Over the 1st to 10th: −0.9%. A well-measured nothing.
SNAP participation: USDA, fiscal 2023 (the rates on our SNAP dashboard). Buyer locations cover 18 months from January 2025 to September 2026 (none from March to June 2026). The year's biggest payout, tax refunds, did not show up either: our refund-season analysis.
Pick your pay schedule
Choose how you are paid. The calendar marks the paydays and the two days after in September 2026, and the result is the pattern for those days over all 514 days, for the platform, brand and price band in the menus.
Pairs sold in the three days from this payday: +5.6% vs other days (95% interval +2.9% to +8.4%; q < 0.001; 99 payday-window days). That is unlikely to be chance.
Colours show each date's past pattern, not a forecast. Paydays move to the business day before a weekend or bank holiday (Labor Day, September 7). For every-other-Friday pay, every Friday is marked: yours is half of them.
In three numbers
What this shows, and what it does not
Payday shows up in sneaker resale, modestly: about 6% more pairs in the three days after the paydays of twice-monthly and month-end schedules, on top of a much bigger weekly rhythm (Thursday to Monday swings by about 13 points). The 1st on its own is weaker than its reputation, benefit days show nothing, and buyers do not pay more per pair.
These are national daily totals. We cannot see who is paid when, so a payday window is a calendar test, not a payroll record, and every-other-Friday pay cannot be separated from the weekly rhythm. The payday lift pointed the same way in both years (15th and last day: 2025 +6.2% (+3.1% to +9.4%), 2026 +4.2% (−0.8% to +10.1%)), but 2026 on its own, with 184 usable days after StockX's gap, is too short to settle it. None of this is a forecast or advice.
Questions
When do people buy the most sneakers?
On resale, Thursdays and Fridays, and the days right after a payday. From January 2025 to September 2026, Thursdays sold +7.6% more pairs than the average day and Mondays −5.5%; the three days from a payday on the 15th or the last day of the month sold +5.6% more than other days (GOAT and StockX, launch sales excluded).
What is the best day to sell sneakers?
The data shows when buyers are busiest, not when prices are highest: Thursday and Friday had the most sales, and the days after the 15th and month-end about 6% more, but the same shoes did not sell for more on those days (within about 1%). Monday had the fewest sales and slightly higher same-shoe prices (+0.53%). This is past data, not selling advice.
Do people spend more on sneakers on payday?
They buy more pairs, not pricier ones. Twice-monthly and month-end payday windows were +5.1% to +5.6% busier than other days; the 1st alone was +3.2%, not significant. Same-shoe prices in the days after the 1st were +0.39%.
Do SNAP or Social Security payments affect sneaker sales?
We found no sign of it. States with more SNAP recipients did not buy a bigger share of their sneakers on the 1st to 3rd (slope +0.10 points per point of participation, 95% interval −0.44 to +0.63), and the days after Social Security's Wednesday payments were −1.5% vs other days.
Do tax refunds boost sneaker sales?
Not in 2025 or 2026: our analysis of weekly IRS refund data found no link between refund dollars and resale volume. This story's day-by-day method removes seasonal waves like refund season on purpose.
Method. Completed GOAT and StockX sales recorded by SneakerPulse, sales between $0 and $30,000, footwear only (cleats excluded), grouped by US Eastern calendar day.
Analysis days January 3, 2025 to September 16, 2026; the baselines use days through September 30, 2026. Feed health: a platform-day counts only if it holds at least 70% of the median
of the 28 days around it (our trending page's rule, centred) and lies outside StockX's collection gap (May 9, 2026 to July 3, 2026, when StockX ran at under half its usual count); dropped:
goat 2025-02-03, 2026-04-14, 2026-04-15, 2026-04-16, 2026-04-17, 2026-04-18; stockx 2025-10-20. The combined series needs both platforms. Holidays left out: Easter Sunday, Memorial Day and Labor Day weekends, July 3 to 5,
Thanksgiving to Cyber Monday, December 20 to January 2. Launches: every listing's sales from 30 days before to 27 days after its StockX release date (GOAT listings matched by SKU), and
the first 28 days of any listing's history in our data, are removed (2,504,987 of 18,304,736 sales). Ratio to expected: each day's count over the median of the
valid days within 14 days either side (16+ needed), on a log scale. Model: ordinary least squares with a weekday and a day-of-month effect; date effects are relative to the average day.
Intervals and p-values: 2,000-draw bootstrap resampling whole calendar weeks (80 weeks), which keeps the dependence between neighbouring days; Benjamini-Hochberg q-values across the 31 dates,
and separately across the eight payday tests. Payday windows: payday (moved to the previous business day for weekends and bank holidays) and the next two days, as one indicator added to a weekday-only model;
Fridays: the Friday effect minus the Monday-to-Thursday average. Robustness (15th and last day / 1st and 15th / the 1st): with launch sales kept +6.7% / +6.2% / +4.5%;
with a 56-day baseline +5.5% / +6.2% / +4.5% (wider intervals, q = 0.071 for the first).
Prices: for each listing with 30+ sales, each day's mean log price minus the mean log price of the same listing's other sales within 14 days (3+ needed), weighted by sales,
same model (40,355 listings). Price bands use each listing's typical (geometric mean) price over the window, not the sale price, so a band never selects on the outcome.
Benefits: GOAT sales with a US buyer state; per state, the share of located sales on the 1st to 3rd (and 1st to 10th) over the share on other days; weighted regression on USDA FY2023
SNAP participation, week-bootstrap interval; BH across the four benefit tests. Counts include international buyers (StockX records no location); GOAT also sells used pairs. The menus' segments are
exploratory: with 72 combinations some will look striking by chance, so lean on the intervals and q-values.
Query: tools/stories/payday_query.py; data: JSON (free to reuse with credit). Product photos: StockX.