Off-the-Clock Work & Compensable Time

Hotels run on handoffs: the agent who counts a cash bank before the system lets them log in, the housekeeper who collects a radio and key packet before riding the service elevator to her floor, the engineer who answers a 2 a.m. call about a flooded suite. Each of those minutes is either compensable working time or it is not — and the gap between what the timeclock captures and what the law treats as "work" is the most reliable class-action generator in hospitality. The minutes are small; the math, multiplied across a workforce and a multi-year lookback, is not.

The continuous workday and the Portal-to-Portal Act#

Start with the federal architecture. The FLSA never defined "work," so the Supreme Court did: exertion or time spent predominantly for the employer's benefit. Congress trimmed that in the Portal-to-Portal Act of 1947 (29 U.S.C. §254), which excludes travel to and from the place of performance and activities that are "preliminary or postliminary" to principal activities. The regulations then supply the doctrine that decides most hotel cases: the continuous workday rule (29 CFR 790.6). Once an employee performs the first principal activity of the day, the workday has begun, and everything until the last principal activity — walking, waiting, gathering supplies — is compensable, bona fide meal periods aside. IBP v. Alvarez (2005) confirmed that walking time after the first principal activity counts, and that an activity is principal if it is integral and indispensable to the job, even if it happens before the scheduled shift.

The practical question for an operator is therefore not "was the employee on the schedule?" but "what was the first thing we required them to do, and when did they do it?"

The hotel pre- and post-shift inventory#

Audit each department against this list — these are the activities that show up in hotel complaints:

• Cash-bank counting at the front desk. An agent required to count and verify a bank before signing into the PMS is working from the first bill counted. If the till count happens before the punch — because the timeclock lives in the back office, or because managers tell agents to "be ready at shift start" — every count is off the clock. The same applies to end-of-shift drops and over/short reconciliation after punch-out.
• Pass-down briefings. Stand-up meetings, shift-overlap handoffs, and reading the pass-down log are integral to front desk, bell, security, and housekeeping work. A 10-minute pre-shift huddle that starts before the punch is ten compensable minutes for every attendee, every day.
• Uniform changing (donning and doffing). When changing into required uniforms or protective gear at the workplace is mandatory — kitchen whites in a facility that prohibits wearing them home, culinary safety gear, engineering PPE — the changing time is integral and indispensable and starts the workday. Steiner v. Mitchell (1956) is the anchor: clothes-changing required by the nature of the work is compensable. Sandifer v. U.S. Steel (2014) adds the union wrinkle: FLSA §3(o) lets a collective bargaining agreement exclude time spent "changing clothes," a clause many unionized hotels rely on. Where employees genuinely have the option to dress at home and most do, changing at work is generally not a principal activity under federal law — but the option must be real, not theoretical (a tuxedo uniform that may go home is different from chef's whites that must stay on property for sanitation).
• Equipment, radio, and key pickup. Housekeepers who must queue at the housekeeping office for boards, master keys, radios, and amenity carts have started work at the front of that line. The walk from the office to the assigned floor is continuous-workday time. The same is true for valets collecting key pads and security officers drawing radios.
• Walking to assigned floors or outlets. Walking before the first principal activity (from the employee entrance to the time clock) is excluded; walking after it (from the supply room to the 14th floor) is not. Place timeclocks before the first required task, not after it.

Hotel trap The most expensive version of this mistake is structural: a single timeclock in the basement, a required uniform room next to it, and a rule that housekeepers "clock in on their floors." Every step of that design converts compensable time into unpaid time. Map the physical sequence — entrance, changing, equipment, punch, first task — and put the punch as early in the chain as any mandatory step.

Screening and donning time: pay everywhere or split by state?#

Many hotels screen departing employees — bag checks at the security office, loss-prevention searches after banquet teardown — and require on-site changing for some roles. Both raise the same operator question: one rule across the portfolio, or a rule that bends at the California line? The standards diverge. Under the FLSA, Integrity Staffing Solutions v. Busk (2014) holds mandatory post-shift security screenings noncompensable postliminary activity — not the principal activity the employee is employed to perform, and not integral and indispensable to it. California reaches the opposite result: in Frlekin v. Apple (2020) the California Supreme Court held required exit searches are "hours worked" because the employee is subject to the employer's control — a test that does not turn on the integral-and-indispensable framework at all. California's control test is the stricter rule, and the highest applicable floor controls in every California property.

That leaves two clean policy options for a multi-state group.

• Pay everywhere. One brand-standard rule — screening and required on-site changing are on the clock — applied identically in Dallas, Miami, and San Diego. Slightly more expensive, but uniform, trainable, and audit-proof.
• Split by state. Unpaid screening where federal law permits (Texas, Florida), paid screening in California — which in practice means moving the search before punch-out, i.e. paying for it anyway. The "savings" branch survives only in the non-California states.

Price the cost shape before choosing. The quantity is minutes × headcount × shifts per year × the regular rate. A 3-minute screening per departing employee across a 300-employee property working 250 shifts a year at a $20/hour blended rate: 0.05 hr × 300 × 250 = 3,750 paid hours/year, about $75,000 annually — the entire price of the pay-everywhere rule, a known and budgetable cost.

The administrative-simplicity case is decisive in most portfolios. A split rule multiplies policy versions, timekeeping configurations, and audit surfaces by jurisdiction, and one misrouted brand standard — "all properties, bag check after punch-out" — converts the California subset into a class action that dwarfs the premium of paying everywhere. Recommendation, conditioned on geography: if the portfolio touches California — or any control-test state — pay everywhere; the uniform cost is trivial against one mis-synced policy. A portfolio entirely within federal-floor states can defensibly leave brief, genuinely post-shift screening unpaid — but most national flags should simply pay it.

Time rounding's collapse#

For decades employers rounded punches to the nearest quarter hour under 29 CFR 785.48(b) and, in California, under See's Candy Shops v. Superior Court (2012), which blessed neutral rounding policies that do not systematically underpay over time. That doctrine has collapsed in stages:

• Donohue v. AMN Services (2021): rounding can never be applied to meal-period punches in California — a 28-minute lunch rounded to 30 fails the meal mandate, full stop. Donohue also created the records presumption that now drives discovery: time records showing a missed, short, or late meal period raise a rebuttable presumption of a violation (see meal & rest breaks).
• Camp v. Home Depot (Court of Appeal 2022): the Sixth District held that an employer that can and does capture the exact minutes worked must pay for all of them — neutral rounding is no defense when the system already knows the true time. The California Supreme Court granted review in 2023; as of June 2026 it has issued no decision, and the question remains formally open. It does not matter operationally: every plausible outcome leaves exact-time payment as the only safe practice.

Operative advice Stop rounding. Everywhere, not just California. Your timeclock already captures exact minutes — rounding is a manual choice to pay something other than the recorded truth, it saves trivial amounts in either direction by design (that is what "neutral" means), and it hands plaintiffs a clean, certifiable, company-wide policy to attack. There is no business case for keeping it.

De minimis minutes: shrinking federally, gone in California#

The federal de minimis doctrine traces to Anderson v. Mt. Clemens Pottery (1946) and is codified at 29 CFR 785.47: insubstantial, irregular periods of a few seconds or minutes that cannot practically be recorded may be disregarded. Note what it requires — irregular and administratively impossible to capture. A recurring task fails both prongs. California rejected the doctrine outright for that scenario in Troester v. Starbucks (2018): a closer who spends four to ten minutes every shift on post-punch closing duties — setting the alarm, locking the door, walking out stragglers — must be paid for them. The hotel analogues are everywhere: the night auditor who walks the lobby after clocking out, the outlet supervisor who arms the restaurant alarm, the agent who powers down the desk after the punch. If a task recurs, schedule it on the clock or capture it through an exception code; "it's only a couple of minutes" is a damages stipulation, not a defense.

On-call and waiting time#

The regulations draw the classic line at 29 CFR 785.14–.17: an employee engaged to wait (waiting is part of the job — a bell captain between arrivals, a banquet houseman waiting for a function to break) is working; an employee waiting to be engaged (free to use the time for their own purposes) is not. On-premises waiting is almost always compensable. Off-premises on-call time turns on how tightly the employer's rules constrain the employee's life: required response times measured in minutes, geographic radius rules, no-alcohol rules, and high call frequency push on-call time toward compensability; a pager and a reasonable response window do not.

The hotel pressure point is the overnight on-call engineer. An engineer sleeping at home who must merely answer the phone is generally not working until called (and then the call and any travel to the property are compensable). But a "responsible to respond" policy that demands a 20-minute arrival, prohibits trading coverage, and generates several calls a night converts the whole night into working time under both federal predominant-benefit analysis and California's control test. If your engineering rotation effectively confines someone, pay them; if you want unpaid on-call, write response windows a person could actually live inside, allow swaps, and track call frequency.

California adds a separate layer: reporting-time pay under the wage orders — half the scheduled shift, minimum two and maximum four hours, when an employee reports and is sent home early — and Ward v. Tilly's (2019) extended "reporting" to telephone call-in scheduling, where employees must call two hours before a shift to learn whether to come in. A call-in system that costs nothing in Nevada owes two to four hours of pay per cancelled shift in California. The full treatment, including the fair-workweek ordinances that reach hotels, is in the scheduling brief.

Travel between sister properties and mandatory training#

The ordinary home-to-work commute is excluded by the Portal-to-Portal Act and 29 CFR 785.35 — even when the employee drives to a different property than usual within the same metro area. But travel between properties during the workday is compensable, period (29 CFR 785.38): the spa therapist who covers the sister resort across town after her morning shift, the engineer dispatched between a cluster's three buildings, the banquet team shuttled to an off-site event. Pay the travel time and remember it counts toward overtime.

Mandatory training and meetings follow the four-factor test of 29 CFR 785.27. Training time is unpaid only if all four hold: attendance is outside regular hours, attendance is genuinely voluntary, the program is not directly related to the employee's current job, and no productive work is performed. Brand-standard service training, anti-harassment compliance modules, new-PMS classes, and pre-opening orientations all flunk at least one factor — usually two. The recurring hotel mistake is the "voluntary" e-learning module assigned with a deadline and a manager follow-up: that is not voluntary, and minutes completed on a phone at home are still compensable.

Biometric timeclocks and BIPA#

One pay-adjacent warning. If any Illinois property uses fingerprint or facial-scan timeclocks, the Biometric Information Privacy Act (740 ILCS 14) requires written notice, a written release before collection, and a published retention-and-destruction policy — and it carries a private right of action with statutory damages of $1,000 per negligent and $5,000 per reckless or intentional violation. After Cothron v. White Castle (2023) held that claims accrue with each scan, exposure briefly reached existential scale; a 2024 amendment pared that back to a single recovery per person for repeated collections of the same biometric by the same method, but per-employee statutory damages across a hotel workforce remain enormous, and hotel operators and their timekeeping vendors have been frequent defendants. Before activating a biometric clock in Illinois — or in the growing list of states with similar bills — route the consent paperwork through counsel and confirm your vendor's data-retention contract matches your published policy.

Timeclock hygiene: the practices that win these cases#

Off-the-clock claims are ultimately records cases. The canonical home for the burden machine — who must prove what, in what order — is recordkeeping; what matters here is the application. When the pre-shift cash count or the post-punch alarm-set was never captured, there is no punch to contradict the employee: under Anderson v. Mt. Clemens Pottery the worker proves the amount by a "just and reasonable inference" from credible testimony alone — "I counted my bank for about ten minutes before every shift" — and the burden shifts to the employer to produce precise contrary evidence it never created. The perverse consequence drives these settlements: the very gap the missing punch left is what relieves the plaintiff's burden, so failing to capture the minutes does not bury the claim — it strengthens it. The defensible program looks like this:

• Punch-out attestations. A daily prompt — "Did you record all time worked today, including before and after your shift?" — with a "no" answer routed to payroll for same-week correction. Attestations are not bulletproof, but they convert silent drift into a documented exception process.
• Exception reporting. A standing report of punches outside scheduled windows, missed punches, and edits, reviewed by someone outside the department whose labor budget is at stake.
• Manager edit audit trails. Every timecard edit should carry the editor's identity, a reason code, and the employee's acknowledgment. Unexplained manager edits that always shave time are the single most damaging exhibit in an off-the-clock trial.
• Open punches for real work. If pre-shift tasks exist, schedule them: build the cash count and pass-down into paid shift time rather than pretending they happen instantly at the scheduled start.

Exposure anatomy: small minutes, large class#

These claims settle on arithmetic, not novelty. A few unpaid minutes is beneath any individual's notice; the same minutes annualized across a full property, through the FLSA's lookback and liquidated-damages frame, is a seven-figure number. Hold the minutes constant and scale the headcount — the move plaintiffs' counsel makes on day one.

How this compounds#

Off-the-clock minutes rarely arrive as a single claim. The same uncaptured time transmits into adjacent theories, and the failure that creates the exposure is usually the same failure that loses the related case.

• To recordkeeping — the proof flows backward. The missing punch that hides the unpaid minutes is the same gap that triggers the Mt. Clemens inference and shifts the burden to the employer. The off-the-clock claim and the recordkeeping defense run on one artifact, the unedited punch trail — build it once (recordkeeping).
• To meal and rest breaks — the same minutes, a second theory. An engineer who answers a radio call during an off-duty meal, or a housekeeper pulled mid-lunch to re-key a room, both worked unpaid minutes (this brief) and had a meal period that was not genuine relief from duty under Brinker. The identical interruption is recoverable twice — as off-the-clock time and as a §226.7 meal premium (meal and rest breaks).
• To scheduling — off-platform staffing is dual-purpose evidence. A manager who runs shift coverage through a group text creates unpaid compensable work (employees monitoring and answering off the clock) and a discoverable record of last-minute schedule changes that feeds reporting-time and fair-workweek claims. One chat log proves both (scheduling).
• To the California stack — the multiplier, not the minutes. Every California off-the-clock dollar is also a wage that should have appeared on the statement and in final pay, pulling in §226, §203, and the PAGA per-pay-period penalty. The wage is the small number; the stack is the exposure (PAGA and class actions).

Compliance checklist#

• Map the first and last required task for every position (use the front desk checklist as the template); relocate punches so no mandatory task precedes clock-in.
• Cash counts, pass-downs, equipment pickup, and required on-site uniform changes scheduled on the clock; §3(o) language reviewed in any CBA before excluding changing time.
• Security/bag checks paid in California (and ideally everywhere); no off-the-clock searches.
• Rounding disabled in every timekeeping system; exact-minute pay confirmed in payroll exports; never round meal punches anywhere.
• Recurring post-punch tasks (alarm setting, lobby walks, system shutdowns) eliminated or captured — no reliance on de minimis, especially for California properties.
• On-call policies audited for response-time, geography, and call-frequency restrictions; engineer call-backs paid from call receipt; day travel between properties paid.
• Training assigned with deadlines treated as paid time, including mobile e-learning.
• Daily attestations, exception reports, and manager-edit audit trails live in every property; housekeeping equipment-pickup and floor-travel sequences reviewed annually.
• BIPA consents, release forms, and retention policies in place before any biometric clock operates in Illinois.

Key authorities#

• Portal-to-Portal Act, 29 U.S.C. §254; 29 CFR 790.6 (continuous workday); 29 CFR 785.14–.17 (waiting/on-call), 785.27 (training), 785.35 & 785.38 (travel), 785.47 (de minimis), 785.48 (rounding).
• FLSA remedies: 29 U.S.C. §255(a) (2-year, 3-year-willful lookback); 29 U.S.C. §§216(b), 260 (liquidated damages equal to the unpaid wages, attorney's fees, and the good-faith defense).
• Anderson v. Mt. Clemens Pottery Co., 328 U.S. 680 (1946); Steiner v. Mitchell, 350 U.S. 247 (1956); IBP, Inc. v. Alvarez, 546 U.S. 21 (2005); Sandifer v. United States Steel Corp., 571 U.S. 220 (2014); Integrity Staffing Solutions, Inc. v. Busk, 574 U.S. 27 (2014).
• Frlekin v. Apple Inc., 8 Cal.5th 1038 (2020); Troester v. Starbucks Corp., 5 Cal.5th 829 (2018); Donohue v. AMN Services, LLC, 11 Cal.5th 58 (2021).
• See's Candy Shops, Inc. v. Superior Court, 210 Cal.App.4th 889 (2012); Camp v. Home Depot U.S.A., Inc., 84 Cal.App.5th 638 (2022), review granted (Cal. S277518) — no decision as of June 2026.
• Ward v. Tilly's, Inc., 31 Cal.App.5th 1167 (2019); IWC Wage Order 5-2001 §5 (reporting time).
• Illinois Biometric Information Privacy Act, 740 ILCS 14; Cothron v. White Castle System, Inc., 2023 IL 128004, as modified by the 2024 amendment.