Peak season no longer starts in November. Adobe expects $95.8 billion in US online spending in October alone, up 8% from last year, with $9.9 billion of it during the October 6 to 7 Prime Day event. The full November and December season is forecast at a record $275.1 billion, with $47.5 billion of it packed into the five days from Thanksgiving through Cyber Monday. Adobe’s own summary is that pre-season shopping “has changed the shape of the holiday season.”
Carriers are planning for that volume in their own way. UPS said on October 7 that it plans to hire 100,000 seasonal workers this year, according to The Epoch Times, about 9% fewer than last year, as more of its network runs through automated facilities.
The other half of the capacity question sits with brands and their fulfillment partners. Every warehouse and 3PL has a ceiling: a number of orders it can pick, pack and hand to carriers in a day. Most of the year, brands never get close to it. In peak season, a single promotion can push volume past it, and once that happens every order behind the line ships late.
What capacity actually means at a 3PL
Capacity at a fulfillment center is a set of limits. Labor is the most obvious: how many pickers and packers are on the floor, and how quickly seasonal staff get up to speed. Then there is the dock: how many trailers can be loaded before the last carrier pickup. Receiving capacity matters too, because inventory that sits on a trailer waiting to be checked in cannot be sold. And storage sets a limit on how much peak inventory can be on hand at once.
Most 3PL relationships account for this through forecasts. Brands share expected order volumes ahead of peak, and the 3PL staffs and plans to them. When actual volume runs well above forecast, the 3PL has to choose which orders to work first, and service levels such as same-day shipping for orders placed before a cutoff start to slip.
None of this reflects badly on the 3PL. It is how shared fulfillment capacity works. The brands that come through peak cleanly are the ones that plan for the day their volume exceeds what one building can ship.
The signs you are reaching the ceiling
Capacity problems show up in order data before they show up in customer complaints. Order age is the earliest signal: the time between an order being placed and the 3PL picking it. When the oldest unshipped order is a day old instead of a few hours, a backlog is forming.
Same-day ship rate is the next signal. If the share of orders shipped on the day they were placed starts to fall, cutoffs and delivery promises shown to customers are already wrong. Receiving backlog is a third: inbound inventory waiting to be checked in means sellable stock is sitting in the yard while the storefront shows it as unavailable. And “where is my order” contacts rise last, after customers have already noticed.
Brands that watch these numbers daily from mid-October onward get several days of warning. Brands that wait for customer complaints get none.
Overflow options before you need them
Overflow fulfillment means sending some orders to a second location when the primary location is at capacity. There are several ways to set it up, and many brands combine them.
A second 3PL is the most common. Brands that already work with more than one fulfillment partner can shift volume between them, and brands with one partner can add a smaller overflow partner for peak. A brand-run warehouse or a returns facility can also serve as overflow for a narrow set of fast-moving SKUs. Retail stores that ship online orders add capacity spread across many locations, close to customers. Marketplace fulfillment programs can take marketplace orders off the primary 3PL. And some products can ship directly from a supplier or manufacturer.
Each option has a cost. An overflow location may charge more per order, ship from farther away, or hold only part of the catalog. The question for planning is which orders should go where when the primary location is full, and that is a routing question.
Routing rules for overflow
The simplest overflow rule is a daily volume cap. The primary 3PL takes orders up to the number it can ship that day, and every order above the cap routes to the overflow location. This keeps the primary location at its sustainable pace and stops a backlog from forming in the first place.
Caps work better with conditions attached. Route to overflow only for SKUs the overflow location actually holds. Route by region, so overflow orders go to the location closest to the customer. Route by channel, for example sending marketplace orders to marketplace fulfillment and keeping direct-to-consumer orders at the primary 3PL. Or route by order age, sending orders to overflow when the primary location’s oldest open order passes a threshold.
For more on how routing engines make these decisions at peak, see Achieve Peak Fulfillment with Pipe17’s AI Order Routing Engine.
The rules need to be set by the people who run operations and changeable in minutes. On Cyber Monday, the right cap is the one that matches what the building is actually shipping that afternoon, which may differ from what the forecast said in September.
A worked example
Picture a brand whose primary 3PL can ship about 5,000 orders a day at peak staffing, and which expects promotional days well above that. Before peak, the brand adds a smaller overflow 3PL and stocks it with its 50 best-selling SKUs, which cover most single-item orders.
On a normal November day, every order goes to the primary 3PL. On the morning of a big promotion, the routing layer sends orders to the primary 3PL until it reaches its cap, then sends any order made up entirely of the 50 overflow SKUs to the overflow partner. Orders that include other SKUs stay with the primary 3PL and are worked in sequence. By the end of the day, the primary 3PL has shipped what it could at a sustainable pace, the overflow partner has absorbed a large share of the spike, and the backlog that would have rolled into the next three days never forms.
The numbers here are illustrative, and every brand’s catalog and network will differ. The structure is what matters: a cap, an overflow location stocked for the orders it will receive, and rules that switch on automatically.
Inventory has to be in the right place first
Overflow only works if the overflow location has the stock. That makes inventory positioning the planning decision that comes before routing. Decide in October which SKUs the overflow location will carry, how many units, and when transfers need to land so they are received and sellable before the first big promotion.
Splitting inventory across more locations also raises the risk of overselling. If the storefront and marketplaces see one blended number while stock sits in two buildings, an order can be accepted against units that are physically in the wrong place. Per-location availability, updated in close to real time and shared with every channel, is what lets a brand add an overflow location without creating a new source of cancellations.
Keep delivery promises honest during overflow
Overflow changes where orders ship from, which changes how long they take to arrive. An order that would have shipped from a primary 3PL a day’s drive from the customer might ship from an overflow location across the country. If the storefront and marketplaces keep showing the primary location’s delivery estimate, customers get a promise the overflow location cannot keep.
The fix is to calculate delivery estimates from the location that will actually fill the order, as Pipe17’s Order Promising does, and to update marketplace handling times when overflow is running. Brands that tie their delivery promises to the same routing rules that send orders to overflow keep the estimate accurate automatically.
What overflow costs, and when it is worth it
Overflow is rarely free. An overflow partner may charge a higher rate per order, require minimum volumes or a setup fee, and ship from a location that costs more to reach most customers. Inventory placed at the overflow location also has to be paid for twice: the transfer to get it there, and either a transfer back or a slower sell-through afterwards.
Against that, the cost of not having overflow shows up as late orders, cancellations, marketplace performance penalties, extra support contacts and customers who do not come back. For most brands, the useful comparison is the overflow cost per order against the cost of a late or cancelled order during the most valuable weeks of the year. Capping overflow to the days and SKUs where it is needed keeps the extra cost contained.
Connect the overflow partner before peak
Bringing a new fulfillment location online takes work: order and inventory integration, receiving and advance shipping notices, packing and labeling requirements, carrier accounts and test orders. All of it is manageable in October. Very little of it is manageable during the week of Black Friday.
Brands that plan to use overflow should have the location connected, stocked and tested with live orders before November. Then turning overflow on is a routing change, and turning it off in January is another.
Measure it daily through peak
Four numbers tell an operations team whether its capacity plan is working: orders routed to each location against that location’s cap, the age of the oldest unshipped order at each location, the share of orders shipped within the promised window, and the cost per order at the overflow location compared with the primary. Reviewed daily from the start of the October promotions through mid-December, they show when to raise or lower caps and when to move more inventory to overflow.
Where Pipe17 fits
Pipe17 is the Order Operations Platform. It connects sales channels, fulfillment and back-office systems, and orchestrates every order from checkout to delivery, with rules the business team sets. For peak capacity, that means one place to see per-location inventory across 3PLs, warehouses and stores, and to route each order with rules such as daily caps, SKU eligibility, region and channel. Adding an overflow 3PL becomes a connection and a routing rule, and every channel keeps selling against accurate availability while volume moves between locations.
The takeaway
Peak season now starts in October and peaks in a handful of days in late November. Every 3PL has a ceiling, and the brands that ship on time are the ones that plan for crossing it: an overflow location connected and tested before November, inventory positioned for the orders it will take, daily caps and routing rules the operations team can change in minutes, and the daily numbers to know when to use them.
