The short answer: A surge shows up on the floor as four signals before it reaches a headcount plan. Watch for overtime asked for two weeks running, the same people borrowed across lines, shifts covered only after a scramble, and changeovers that slip. Overtime is easiest to count: manufacturing production workers averaged 4.0 overtime hours a week in September 2026 (preliminary), per BLS.
Written for shift leads, line supervisors and the operations managers they report to at light industrial employers in manufacturing, warehousing and food processing; how workers handle peak-season schedules is a separate subject.
Production and nonsupervisory workers in manufacturing averaged 4.0 hours of overtime a week in September 2026, a preliminary figure, up from 3.8 a year earlier, according to the Bureau of Labor Statistics. Their workweek reached 42.0 hours, while manufacturers still held 522,000 open jobs in August, a 4.0% openings rate. That’s work being carried by the people already on the floor. It also means the extra work lands in the overtime log weeks before it lands in a request for more people. Your shift leads read that log first.
At Horizon America, our discovery conversation with a site covers four things before we recommend a placement strategy: wage signals, scheduling structure, onboarding depth and local market conditions. Scheduling structure is the one your shift leads live inside every shift. And the employers we work with buy staffing through operations, HR and plant managers, with line supervisors shaping the decision without signing it. That suggests the earliest read often sits with people who don’t make the call.
This guide covers what a surge looks like before it becomes a number, the four floor signals and which combinations should move headcount. It then shows how to get the floor’s read into the plan, and three numbers that show whether it arrived in time.
What Does a Surge Look Like Before It Becomes a Number?
Before it’s a number, a surge looks like extra hours. In the Federal Reserve’s October 2026 Beige Book, Kansas City District manufacturers reported adding hours, overtime options and longer shifts “because of the limited labor availability within the market.” When firms can’t find people, the existing crew works longer. That’s why a headcount plan sees the surge late.
Across sectors, the same report found that most Districts “described little or no net change in overall employment levels.” One Kansas City manufacturer put the strain plainly: “It is hard to expand the workforce when they leave faster than you can replace them.” When people leave as fast as you hire, overtime fills the hole.
Economists already treat hours as an early signal. Average weekly hours in manufacturing is one of the ten components of The Conference Board’s Leading Economic Index for the US. Its place on that list is the point. Hours tend to move early nationally, and the same logic applies on your floor.
That’s the gap between the floor and the plan, and it’s a timing gap, not a bad tool. A headcount plan runs on data that arrives on a schedule. Federal job openings figures for October 2026 won’t publish until December 1, after most peak ramps are under way. The plan runs on a slower clock. Your shift leads run on tonight’s.
The question for your own site is one the floor answers every day: how much of this week’s output depends on hours nobody planned for? A model can’t see that until the hours become a trend.
How Do You Read the Four Floor Signals?
Read the four floor signals by counting, not by feel. Each week, log overtime hours by line, people borrowed across lines, shifts covered only after calls, and changeovers that ran late. Six weeks of logs give you a baseline. Without one, a normal busy week and a real surge look the same.
1. Overtime that stacks up for two weeks
The detail most sites miss is that overtime hides the surge from the plan. Every hour a crew absorbs is an hour that never becomes a headcount request. One week above normal is noise. Two running weeks on the same line is a forecast.
Compare each line with its own history, not a national figure. In 2025, BLS put weekly manufacturing overtime at 3.8 hours, against 3.6 in 2023 and 2024, but the spread underneath is wide. BLS data show food manufacturing averaged 4.0 overtime hours a week in 2025, while transportation equipment manufacturing ran 5.3. A line that’s normal at five hours tells you nothing at five; a line that’s normal at two tells you a lot. How the hours are shared matters too, and Horizon America has written about rotating overtime fairly through a peak.
2. Borrowed people and scrambled coverage
Scheduling structure is one of the four inputs in our discovery conversation. A schedule that needs a scramble to fill has already run out of slack. Watch for a line that keeps borrowing two people from receiving, or a Saturday shift that only fills after phone calls the night before.
That scramble costs more than a supervisor’s evening. In a 2026 Harris Poll survey for DailyPay of 2,208 hourly US workers aged 18 to 45, 45% ranked schedule among their top three reasons for taking their job. Seventy-two percent called schedule control important, and only 51% were satisfied with the control they get. Last-minute coverage works against the thing they value. That’s a retention risk.
3. Changeovers that start to slip
A changeover that runs late is a capacity signal before it’s a performance problem. Ask each shift lead to log every late changeover with one word for the cause: people, parts or equipment. When “people” keeps coming up, the surge is already on the line, and the log shows when it started.
Manufacturing overtime ran 4.0 hours a week in September. Pull six weeks of overtime by line. A line above its normal two weeks running carries a surge your plan can’t see. Seeing two? Let’s compare notes.
Which Combination of Signals Should Move Headcount?
Escalate combinations, not single readings. One signal is a bad week; two signals on the same line for two weeks is a surge. Waiting is expensive: transportation, warehousing and utilities had 338,000 job openings in August 2026, a 4.5% rate, per BLS. A request that starts late competes for a thin pool.
The table below shows which combinations of floor signals should move a staffing request, and what to send up with each.
| What the floor shows | What it usually means | What to send up |
| Overtime above baseline, one week | Normal variation | Note it; no request yet |
| Overtime above baseline, two weeks, one line | Demand outrunning that line’s crew | Flag the line, with dates |
| Overtime plus borrowed people | The shortage is moving, not shrinking | Request added coverage for that line |
| Scrambled coverage plus slipping changeovers | The schedule no longer covers the work | Request headcount now, with the log |
Two reasons the table reads this way. First, overtime on its own often recovers when volume dips, but overtime paired with borrowing means the shortage has a home on that shift. Second, absences don’t pause for peak. In a given week of 2025, BLS found 2.9% of full-time manufacturing workers missed time for illness, injury, family or other personal reasons, and 3.1% in transportation and warehousing. A crew with no slack has no answer to that.
The move most sites miss is timing the handoff. Send the signal log to your staffing partner on the same day it goes to the plant manager, so your partner can start planning while the request is still being approved. Our peak season staffing guide walks through the planning side of that same window.
Getting the Floor’s Read Into the Plan Before Peak
Get the floor’s read into the plan by giving it a fixed time, a fixed format and a named owner. The American Staffing Association reported temporary and contract staffing employment 5.8% above a year earlier in the four weeks ending September 13. More employers are leaning on temporary and contract help, so a request that waits competes with them.
- Hold a weekly floor read. Each shift lead reports the four signals by line, with dates, in fifteen minutes.
- Agree the trigger in advance. Two signals on one line for two weeks sends a request to the plant manager that day.
- Name who answers. One person owns the reply within 48 hours: approve, adjust or explain.
Automate the count, never the conversation. A weekly overtime-by-line report can build itself from your timekeeping system. The call on what it means should stay with the person who walked the line, because only they know whether Tuesday’s overtime was a late truck or a missing crew.
Overtime isn’t a free bridge while that call waits. A 2005 study of US workers across industries, published in Occupational and Environmental Medicine, found jobs with overtime schedules carried a 61% higher injury hazard rate than jobs without overtime. It’s older research, and it still describes the trade a late request makes.
Building a flexible workforce before peak, and hiring fast without shortcuts once it arrives, both start with a request that goes in on time.
Three Numbers That Show the Floor Was Heard in Time
Three numbers show whether the floor was heard in time. Track days from the first logged signal to an approved request, weeks of overtime above baseline after approval, and how many added workers finished training before the peak. In Travelers’ indemnity claims across industries, first-year employees made up about 36% of injuries over five years.
That’s a share of claims, not an injury rate, and it’s still a reason to count training as one of the three. Set the baseline before peak, this week, while the floor is still calm enough to count. Write down each line’s normal overtime, its normal borrowing and its normal changeover time. Without a baseline, every number after it is a guess.
Then read the three together. If lead time is short but overtime stays high after approval, the request moved fast and added people arrived slowly, so the bottleneck is screening and onboarding, and your partner needs the log sooner. If lead time is long but overtime drops soon after approval, staffing caught up quickly and the delay sits in the approval path, so the trigger needs fixing. If both run high, the floor’s read isn’t reaching the plan, and the weekly read is the fix.
This works alongside the weekly coverage reviews distribution sites already run, and with forecasting labor alongside production. After peak, compare the three numbers with last year’s. That comparison is the case for the next plan.
If your shift leads flagged this peak more than two weeks before the request, overtime paid for that gap. Rushed onboarding often does too. Check both dates. Longer than two weeks? Ask where it stalls.
Frequently Asked Questions About Surge Staffing Signals
How much notice does a staffing agency need to add workers?
It depends on the role, the shift and the local market. Ask your staffing partner for its usual lead time on your most common roles, then plan backward from your busiest weeks. Roles that need certifications or longer training take more notice.
Who is responsible for headcount planning in a plant?
It varies, but the plan usually sits with operations or plant management, with HR and finance shaping the budget. Supervisors and shift leads feed it with what they see on the floor. The clearest plants name one owner who signs off on the number.
What is a leading indicator?
A leading indicator is a measure that tends to turn before the wider economy does, so it’s watched for early warning. New orders and building permits are common examples. They’re read alongside slower data, such as payroll counts, that confirm a change after it happens.
What is the difference between temporary and temp-to-hire staffing?
Temporary staffing covers a surge, a season or a project, and the assignment ends with the need. Temp-to-hire starts as a temporary assignment with the option of a direct hire later. It lets both sides try the fit first.
Is there a limit on how many hours an employee can work?
Federal law sets no cap on hours for adults. The US Department of Labor says the Fair Labor Standards Act “does not limit the number of hours per day or per week” that employees 16 and older can be required to work. Overtime pay rules and state laws still apply.
The Floor Sees It First, So Let the Floor Start the Clock
Run the two-week test on your own last peak. If the request went in more than two weeks after the overtime started, the plan wasn’t wrong; it was late. With manufacturing overtime at 4.0 hours a week and 522,000 manufacturing jobs open in August, late is the expensive option.
The floor’s read is your shift leads’, and the call on headcount belongs to the managers they report to. Screening, placing and supporting the added workers, 24/7 and 365 days a year, is what Horizon America does, starting with an onsite look at how your site runs. See our employer services, then tell us what’s happening.
References
- “Average weekly hours and overtime of production and nonsupervisory employees on private nonfarm payrolls by industry sector, seasonally adjusted (Table B-7).” US Bureau of Labor Statistics, Employment Situation. Published October 2, 2026. Verified live 2026-10-25.
- “Overtime hours in manufacturing industries.” US Bureau of Labor Statistics, The Economics Daily. Published March 31, 2026. Verified live 2026-10-25.
- “Beige Book, October 2026.” Federal Reserve. Published October 14, 2026. Verified live 2026-10-25.
- “The Conference Board Leading Economic Index for the US, technical notes.” The Conference Board. Published October 22, 2026. Verified live 2026-10-25.
- “JOLTS release schedule.” US Bureau of Labor Statistics. Published undated (schedule page). Verified live 2026-10-25.
- “Job Openings and Labor Turnover, August 2026.” US Bureau of Labor Statistics. Published September 29, 2026. Verified live 2026-10-25.
- “New Research on Frontline Workers Finds Widening Expectation Gap in Schedule Flexibility, Pay Clarity and Digital Tools.” Workday and DailyPay (compiling primary research from The Harris Poll). Published September 17, 2026. Verified live 2026-10-25.
- “Absences from work of employed full-time wage and salary workers by occupation and industry.” US Bureau of Labor Statistics, Current Population Survey, 2025 annual averages. Published undated. Verified live 2026-10-25.
- “The Impact of Overtime and Long Work Hours on Occupational Injuries and Illnesses: New Evidence from the United States.” Dembe, Erickson, Delbos and Banks, Occupational and Environmental Medicine 62(9), via CDC Stacks. Published September 2005. Verified live 2026-10-25.
- “Staffing Index Edges Up in September.” American Staffing Association. Published September 22, 2026. Verified live 2026-10-25.
- “Travelers Releases 2025 Injury Impact Report.” Travelers. Published June 3, 2025. Verified live 2026-10-25.
- “How many hours per day or per week can an employee work?” US Department of Labor, elaws FLSA Advisor. Published undated. Verified live 2026-10-25.