Sixty days before peak, a Chinese seller I know started doing the math. His California warehouse normally runs with 12 people. Last Black Friday week, order volume tripled, shipments slipped by two days, and the marketplace fined him. This year he wised up and started the labor plan in early September. The lesson is simple: with peak-season staffing, starting late is unfixable.
Look at the big picture first. Amazon announced it would hire 250,000 seasonal workers across the US for the 2025 holidays, at average pay above $19 an hour. Meanwhile the Challenger, Gray & Christmas outlook projected fewer than 500,000 new seasonal retail positions in Q4 2025 — the smallest holiday hiring wave since the 2009 recession. Giants sweeping up labor on one side, a shrinking total pool on the other: for a mid-size seller's overseas warehouse, hiring will only get harder.
Do the Headcount Math Before Anything Else
Do not guess. Take last peak season's average daily orders, divide by one experienced packer's daily output, and that is your baseline gap. Then multiply by 1.2 — temps in their first two weeks run at roughly seventy to eighty percent of veteran output, and you must bake that discount in.
An example: last Black Friday week averaged 3,000 orders a day. One veteran packer ships 250 a day, so in theory you need 12 people. At eighty percent temp productivity, that becomes 15. Add two floaters and 16 is your starting line. Once that number is on paper, every conversation about channels and timelines has an anchor.
The first US-China labor trap sits right here: the habit of "panic-hiring one week before peak" does not work in America. Yes, employment is at-will, but the chain of recruiting, background checks, and onboarding is no faster than back home. In logistics hubs, staffing agencies have their rosters spoken for by the big warehouses a month before peak. Arrive late and you get whoever is left.
Two Channels: Agencies and Direct Hiring
My recommendation is a 60-40 split: sixty percent through local staffing agencies, forty percent hired directly.
Agencies are fast. Sign the agreement and a batch of workers shows up within a week, and underperformers can be swapped. The downside is cost — agency markups typically run thirty to fifty percent on top of the hourly wage, and the closer to peak, the higher the markup. So lock the agency channel early: sign 45 days before peak, with headcount and start dates written into an addendum. Negotiating in November means negotiating with no leverage.
The forty percent you hire directly goes through local job boards and social media. Blue-collar recruiting in the US works better through these channels than most Chinese sellers expect — post the hourly rate clearly, spell out the night-shift premium, and applications come in fast. Direct hires also tend to be more loyal, making them your best candidates for post-peak retention.
Second trap: do not compare hourly rates, compare total cost. US employer burden is real — workers' comp, unemployment insurance tax, payroll tax share, and in California, overtime at 1.5x after 8 hours in a day. Agency quotes and direct-hire wages are quoted on different bases. Normalize them before comparing, or you are lying to yourself.
The 14-Day Ramp: From Job Post to Solo Shift
This is the schedule I have used with several sellers. It assumes the headcount math is done and agencies are locked 45 days out.
| Days | Action |
|---|---|
| Day 1-3 | Job posts go live, agencies get their requisitions; internally, finalize training materials and the one-page SOP |
| Day 4-7 | Batch interviews. Screen for three things only: can they lift, will they work nights, any warehouse experience |
| Day 8-10 | Onboarding in waves: half a day on safety, half a day on process, the rest shadowing |
| Day 11-12 | Buddy shifts — temps work independently with a veteran watching |
| Day 13-14 | Solo shifts with ramping output; 15-minute end-of-day review of what went wrong |
Three things matter most. First, keep interviews simple — you are hiring hands and feet, not resumes; three rounds of interviews waste everyone's time. Second, training materials must be finalized before recruiting starts, not improvised after people arrive. Third, never skip the day 13-14 reviews. Eighty percent of a temp's first-two-week problems cluster in the first three days, and one review session clears them — saving two weeks of rework later.
The third trap is safety training. The US has hard requirements for temp safety training, and California's warehouse quota transparency law, effective 2022, requires employers to disclose production quotas in writing. "They are only here for two months" is not an excuse to skip it — an injured temp costs you exactly the same as an injured full-timer.
The One-Page SOP: Compress the Job Until a Newcomer Gets It
Temps turn over fast. You cannot expect them to memorize a 20-page manual. My iron rule: each role's SOP fits on one sheet of paper — flowchart on the front, FAQ on the back.
Take the pack station. The front lists five steps: scan the order, verify the pick, box it, add dunnage, label and weigh. Each step gets a real photo with the critical spot circled. The back answers three things: what to do when a label is wrong, who to call when a box is damaged, who to call when the system freezes. That is it. One extra word is dead weight.
The sheet is not done when it is written. During day 8-10 training, have the first wave of temps work from it and mark every spot where they get stuck. By day 13 it should be on its second or third revision. A revised sheet is a usable sheet; a first draft is usually scrap paper.
After Peak: Who Stays, Who Goes, How to Close Out
Mid-January, volume falls, and the hardest step arrives. My advice: say "seasonal role, ending mid-January" on day one of onboarding. No false hope. This is the easiest US-China workplace culture trap to fall into — American workers have near-zero tolerance for being strung along, and mishandling it earns you bad reviews and complaints.
Retention runs on three metrics: attendance, error rate, learning speed. Rank on each, take the intersection, and talk conversion to the top 20%. Send those offers before the end of December — wait until January and someone else has already hired them.
The offboarding process must be clean: final paycheck on the last day (California requires same-day final pay, no exceptions), collect badges and uniforms, and run a 10-minute exit interview. The interview is not theater. Two questions suffice: what part of the process was the most awkward, and would you come back next peak season. Everyone who says yes goes on a list you call next September — your cheapest recruiting channel.
Three things you can do tomorrow: first, pull last peak season's order data and run the headcount formula above; second, call two local staffing agencies and ask about peak availability and markups; third, take your pack-station SOP and see if it fits on one page. Do those three, and your peak-season labor plan has a skeleton.
