Corporate TransportationSmart Mobility

Seasonal workforce transportation is a Q2 decision, not a Q4 one

· 13 min read
Empty marked shuttle bay at a distribution center in summer, a planner with a tablet beside it, loaded trailers behind

You contract trailer capacity in June for November. Parcel injection slots, temporary sort space, peak surcharges, additional air lift: all of it committed and priced months out, because you already know that every competitor is bidding for the same slots in the same ten weeks. Then you get to the people, and the discipline stops. Headcount turns into a requisition line, and transportation turns into an assumption that the bus will run, the lot will hold, and somebody will find a charter operator in October.

That assumption fails for exactly the reason you book freight early. Employment in couriers and messengers climbed by 275,200 between September and December 2025, which means 21% of that industry's December workforce had not existed three months earlier (BLS, Current Employment Statistics, not seasonally adjusted). Peak-season commute capacity behaves like freight capacity, not like a cost line that scales with headcount. Your surge arrives at the same hour as everybody else's, onto shared capacity already sitting at its own annual ceiling.

Your peak-season workforce surge moved from the store to the sort center

Retail trade is the shallow end. Between September and December 2025, retail trade employment rose 3.0%, from 15,345,600 to 15,807,100 (BLS CES, not seasonally adjusted). Couriers and messengers rose 26.6% across the same four months. Warehousing and storage, the industry most people picture when they hear "peak-season hiring," rose 0.8%: 14,100 people on a base of 1.86 million.

So the mental model most operators carry is off by roughly a factor of ten, and it points at the wrong buildings. The deep surge lives in the parcel and last-mile tier: sort centers, delivery stations, driver pools. Indeed Hiring Lab caught the same migration on its postings index. As of 14 November 2025, seasonal driving postings ran 153% above the prior year and seasonal loading and stocking postings 49% above, against +1% for seasonal retail postings and −10% for seasonal sales. Postings, not hires. The direction is still unambiguous.

Volume explains the split. E-commerce was forecast to grow 7–9% across the 2025–26 holiday season against total holiday retail growth of 2.9–3.4% (Deloitte, September 2025). Each point of that gap is a package somebody sorts and drives instead of a shopper carrying it out of a store.

Geography is what turns the split into a transportation problem. A mall-based chain hires 60 seasonal associates onto a site a bus already serves, in a place people already drive to for other reasons. A delivery station hires 400 onto a parcel of land chosen for interchange access and price per acre. The store peak lands on infrastructure that already exists. The sort-center peak lands on infrastructure nobody built.

The bus route your site sits on peaks when you do

Correlated capacity is the part no peak budget models. On Tuesday 28 November 2017, five days after Thanksgiving and squarely inside peak, the Mid-Ohio Regional Planning Commission ran an on-board survey of riders on the GREAT shuttle network serving the Rickenbacker logistics cluster south of Columbus. The five employers those riders worked for most: Eddie Bauer, CTDI, FedEx, Gap and Toys "R" Us. Retail and distribution, every one, all peaking in the same window, all drawing from the same corridor.

The survey's finding about the connecting public route is the one to sit with. COTA Route 22 buses were "often filled to capacity, forcing riders to wait for the next one," and the report names peak season specifically as when that happens. COTA's mitigation was tripper buses, an extra vehicle running five minutes behind the scheduled one, which MORPC's own write-up calls "a temporary fix intended for only peak-ridership periods."

Read that as an operator rather than as a planner. The agency's surge capacity is itself a surge product, rationed across every employer on the line, and it shows at the same moment your competitor's 300 new hires show at the same stop.

Two other findings from that survey carry further than the ridership count. 56% of riders had zero working vehicles in the household and another 25% had one, so the commute had no private fallback. And the study recorded that "workers are given no exceptions on lateness, so COTA timeliness directly affects job retention in the area" — on the survey day, Lines 22 and 24 ran more than five minutes late on about 15% of trips.

Caveat the sample honestly: 77 usable responses out of 134 riders, one corridor, one day, nearly nine years ago. Percentages from a survey that size do not generalize to a national rate. The structure generalizes fine. A seasonal workforce with no car, an employer with a zero-tolerance lateness policy, and a shared bus line already at capacity is a combination that produces no-shows in week one.

3,573 jobs in 30 minutes, measured at 8 a.m.

The Inland Empire is one of the country's largest warehouse clusters and among its worst-connected labor markets. Riverside–San Bernardino–Ontario holds 1,640,973 jobs at an average density of 60 per square mile, and the average resident reaches 3,573 of them inside a 30-minute trip by transit and walking, per Access Across America: Transit 2024 (Owen, Liu and Lind, University of Minnesota Accessibility Observatory, December 2025). That is 0.22% of the metro's jobs. Stretch the trip to a full hour and the reach grows to 35,781, or 2.2%. The metro ranks 13th of 50 by total employment and 46th of 50 by weighted transit access.

New York, for contrast, puts 182,046 jobs inside 30 minutes at a job density of 1,512 per square mile. Columbus manages 9,841. Memphis, a parcel hub by definition, manages 3,874.

Now the detail that should change how you read every number above. The report computes travel times "using detailed pedestrian networks and full transit schedules for the 7:00 – 9:00 AM period," and reports what a rider reaches "starting at 8 AM on a typical weekday." That is the best hour of the best day. Your night sort is not in the dataset. Neither is the 4:30 a.m. inbound, the 11 p.m. release, or the Sunday shift you only run in December.

TARC's own Route 28 timetable in Louisville puts a size on that gap: 37 scheduled stops at UPS Worldport on a weekday, 11 on a Sunday. Peak season is when you start adding Sundays.

One correction worth making. A September 2025 release from the same university, headlined "For drivers and transit users, access to jobs in nearly all major US cities is on the decline," travelled much further than the report itself, and it described the 2023 data year. Transit job access did not fall in the most recent year measured: Riverside gained 5.26% year over year, Los Angeles 16.74%. The authors' own framing is the honest one, and the useful one for planning — access "remains well below pre-COVID levels."

The peak is shrinking, which argues for a contracted route rather than against one

Grant the objection in full, because the evidence behind it is real. Challenger, Gray & Christmas projected the smallest retail seasonal gain in 16 years for 2025, and Andy Challenger said on the record in September 2025 that "the traditional seasonal spike in warehousing and shipping jobs is no longer as pronounced, and that shift could be permanent." Q4 job adds in transportation and warehousing fell from 552,300 in 2021 to 303,700 in 2024, on Challenger's compilation of non-seasonally-adjusted BLS data. Forecast seasonal retail hiring for 2025 came in at 265,000 to 365,000 heads against 442,000 the year before (National Retail Federation, 6 November 2025). If the spike is flattening, why build for it?

Because a smaller peak is a cheaper peak to serve and a more concentrated one is easier to route. Six coaches on two fixed loops covering 300 seats twice a day is a purchasable, priceable, contractable thing; a diffuse national surge is not. A thinner roster also makes every absence heavier. Cut the seasonal headcount and raise the volume per head, and one no-show on a sort shift backs up a belt.

Rickenbacker is the empirical version of that argument, and it runs the direction skeptics do not expect. A fare-free workforce shuttle network runs south of Columbus today, and its origin is a single peak season. Transportation planner Daniel Haake, writing in Governing in 2017, traced it back to 2014, when warehouses around Groveport could not staff the holiday push and one of them paid for a bus link to workers back in Columbus proper. The arrangement was supposed to expire with the season. Instead the wider logistics industry asked for a version that would not. The permanent version is Groveport Rickenbacker Employee Access Transit — GREAT. Its 2016 ridership came to 24,274 on Haake's figures, and the inland port it serves employed nearly 20,000 people at the time he wrote. COTA still lists the service in 2026. Ten weeks was long enough to prove the demand that justified permanent infrastructure.

There is also a wage arithmetic that does not move with the size of the peak. Amazon put average pay for its 2025 seasonal roles at over $19 an hour; 40 hours a week for ten weeks grosses about $7,600 before tax. Owning and running a new vehicle driven 15,000 miles a year cost $11,577, or $964.78 a month (AAA, Your Driving Costs 2025). A ten-week job cannot finance a car, and a seasonal hire who arrives without a reliable one is not going to acquire one for the season.

Adam Porter, business co-chair of the Harrisburg community group Friends of Midtown, put the same arithmetic more bluntly to WHYY in 2016: "A lot of people want to apply to a $14-hour dollar job, but you still need to be earning $17 to $19 to actually truly afford a car." He was talking about the commute to a local Amazon distribution center. Capital Area Transit redrew four of its bus lines around where Amazon's workers actually lived, in Harrisburg and Enola, and reset the timetables to the company's 12-hour shifts. Wages have moved a long way in the decade since. The ratio between the entry wage and the cost of the car that gets you to the entry job has not moved with them.

Book the coaches on the freight calendar, not the hiring calendar

Motorcoach supply is smaller and far more fragmented than the word "charter" suggests. The American Bus Association Foundation's 2025 Motorcoach Census counted 1,891 companies running 50,968 coaches across the US and Canada, with 87.3% of those companies operating fewer than 25 coaches and average employment of 41 people per company.

Do the arithmetic from the operator's side of the table. A six-coach, ten-week Q4 block is a quarter of a typical operator's entire fleet, committed straight through the weeks that also carry school contracts, sports charters and holiday tour work. That block is not sitting unsold in October waiting for your call. The constraint is fragmentation rather than a driver shortage. Coaches exist in quantity; the census counted 50,968 of them. Uncommitted Q4 coaches, in your county, on your shift clock, are the scarce item.

The window is also creeping earlier. On NRF's winter-holiday FAQ numbers, roughly two shoppers in five are buying before November opens, and those two months carry about 19% of the year's retail sales. Demand pulled forward; the seasonal transport plan has to pull forward with it.

Three specifications separate a seasonal staff shuttle that holds from one that unravels in its second week. First departure and last return set against the actual shift-change clock rather than the day-shift default. Weekend coverage on the days your peak adds volume and the transit timetable subtracts trips. And a named on-time percentage with a remedy attached to it, which is the part most shuttle service level agreements leave vague — a workforce operating under a no-exceptions lateness policy has no tolerance for a late bus.

A ten-week program also has no recovery window. A year-round pilot that sags in week six still has five months to correct itself. A peak route that misboards in week two has burned a fifth of its life.

Pull two things before your next planning cycle, and neither of them needs a vendor in the room. Geocode last year's seasonal roster to home ZIP and lay it over the routes that already exist; MORPC's survey found rider home addresses clustering in a handful of ZIPs around Rickenbacker, and yours will cluster too. Then put the first-departure and last-return times of every route serving your site next to your real shift-change clock, Sundays included. The distance between those two columns is your no-show forecast, and it is knowable in April.

Then move the money. Most peak plans already carry an over-hire buffer sized to absorb attrition that transportation caused in the first place. Reprice part of that buffer as contracted seats, and buy them on the calendar you already use for trailers. If the seasonal spike keeps shrinking, that block gets cheaper every year — which is an argument for locking it now. Ryde builds contracted route capacity for warehousing and distribution sites.

Sources