Look — a 3PL's margin doesn't live in the warehouse fee. It lives in the rate differential: the spread between what you negotiated with carriers and what you bill the customer (S28, the industry's worst-kept secret and its least-managed number). And September is the month that spread gets decided for the whole year, because September is when you publish peak cutoffs and lock capacity commitments. Most 3PLs treat that cutoff sheet like an ops memo. It's a P&L document, and it should be priced like one.
This one's for the 3PL executive, the person who has to explain Q4 margin in February.
The cutoff sheet, priced honestly
Every line on a peak cutoff sheet is a financial commitment wearing an operations costume. "Orders received by 2 p.m. ship same day" is a promise that your dock, your sort, your carrier pickups, and — if your network touches the region — the intermodal handoff upstream of your dock all perform at peak load. When the promise breaks, the cost doesn't land evenly. It lands on whichever customer screams loudest, in the form of expedites you eat, and that expedite comes straight out of the rate differential. The spread that looked like four points of margin in your blended view can be negative on your noisiest account by Thanksgiving, and your blended view will never tell you.
That's the structural problem with how most 3PLs run shipping economics: blended. One negotiated rate stack, one ops standard, one peak plan, margin computed at the company level. But peak doesn't hit you at the company level. It hits you one customer at a time — the apparel account whose volume triples, the electronics account whose oversize mix triggers demand fees, the subscription account whose injection schedule can't slip. Per-customer P&L isn't a finance nicety. It's the only view where peak commitments can be priced before you make them.
The September math, gateway edition
If your buildings sit anywhere from Chicagoland to Gary — and for national 3PLs, plenty do — your cutoffs have an upstream dependency most cutoff sheets ignore: ramp dwell. Inventory that feeds your peak waves arrives through the Joliet and Elwood ramps, on networks where six Class I railroads converge and chassis pools tighten exactly when your customers' freight surges. A cutoff sheet that assumes inventory arrives on schedule in November is a fiction document. The honest version models dwell at the gateway as an input: current baseline, peak-season degradation from prior years, and the replenishment buffer per customer that follows from it. [S-cite: Chicago gateway peak-season dwell vs. annual baseline]. Build the buffer into the commitment, or build the expedite into your margin forecast. Pick one on purpose.
And while you're modeling dwell, model the equipment underneath it, because dwell and chassis are the same shortage wearing two hats. When ramps back up, boxes sit on chassis; when boxes sit on chassis, the pool tightens; when the pool tightens, street dwell gets longer and per-diem clocks start running on everything you've got out. I've explained chassis-pool economics to more VPs than I can count, and the summary is always the same: by the time the shortage is visible on your dock, it's been building at the ramp for three weeks. Watch the pool numbers in October, not the dock in November.
The onboarding ceiling
Second September reality: the customer your sales team signs this month is a January customer. Cloud-era multicarrier implementations run eight to twelve weeks for the mid-market (S26, the consensus benchmark — and that's when integrations behave). A 3PL that promises a November go-live on a September signature is volunteering to run its newest, least-understood account through its most dangerous quarter on a half-tested integration. Onboarding time is the growth ceiling for a 3PL — not sales pipeline, not square footage. The fix isn't heroics in October. It's an onboarding pipeline measured and engineered like the operation it is, in the quarters when nobody's watching.
What the software vendors get half-right
The enterprise shipping vendors court this market hard, and they're not wrong about the need. ProShip's positioning leans on high-volume execution speed; Pierbridge's Transtream has long pitched the 3PL multi-tenant story. Fine products solving real problems — but notice the frame: both descend from shipper-side TMS thinking, scaled up. A shipper needs one P&L and fast labels. A 3PL needs N P&Ls, one per customer, with rate differential, surcharge exposure, exception cost, and onboarding status visible per account, per lane, per week. If your shipping stack can't tell you which three customers will be unprofitable in December under announced demand fees, you don't have a 3PL system. You have a shipper system with extra logins.
Tradeoff, acknowledged: per-customer P&L takes invoice-level data discipline and rating logic that most operations haven't built, and September is late to start building it for this peak. But it is not too late to do the triage version.
Here's the concrete ask. Take your top ten accounts by Q4 volume. We'll run a per-customer peak exposure worksheet with you — negotiated rates versus billed rates, announced surcharge layers applied to each account's actual profile, dwell-adjusted replenishment risk flagged per lane. Two weeks, your data, no charge for the first pass. You'll publish the same cutoff sheet either way. The difference is whether you know what it costs.