Every warehouse handles its average day fine. It’s the day that isn’t average where the trouble starts.
The most common error we see is sizing dock doors, staging area and pick paths around average daily volume, then assuming labour flex and overtime will absorb whatever comes on top. That works right up until peak volume exceeds what the physical building can move. At that point, adding headcount stops helping. You can’t put more people into a staging area that’s already full, or push more forklifts through a dock that only has six doors.
A retailer’s peak is a sharp seasonal spike with a hard calendar date attached to it. Christmas doesn’t move. Chinese New Year doesn’t move. The volume is outbound-heavy, concentrated in a narrow window, and everyone in the business already knows it’s coming months in advance.
An FMCG manufacturer’s peak usually looks nothing like that. It’s promotional, driven by trade activity that operations may only find out about a few weeks ahead of a major retail push. That peak lands on inbound and staging as much as outbound, because the volume has to arrive, get checked in and get staged before it ever reaches a dock door going out.
A pharmaceutical distributor’s peak may not be seasonal at all. It can be regulatory or tender-driven, arriving as a large, low-frequency block of volume tied to a government contract or a product recall. The rest of the year might run at a fraction of that volume.
The multiple, whether peak is 1.5 times average or 3 times average, matters less than where that peak volume actually lands in the building. A layout that handles a 2x outbound spike well can still fail on a 1.3x peak that hits inbound and staging at the same time, because those functions were never modeled to run under stress together.
Good design models peak-day throughput as its own scenario, with its own bottleneck analysis, not as a percentage uplift on the average-day model. Where does dock capacity run out first? Where does staging fill up? Where does the pick path start to congest because two functions are competing for the same aisle at the same time?
When that modeling doesn’t happen, the failure shows up in predictable places. Trucks queue outside because there aren’t enough doors to turn them around fast enough. Staging areas fill with pallets that have nowhere to go, so operators start using aisles as overflow storage, which slows down picking. Pick paths that worked fine at average volume become gridlocked because the building was never asked to run inbound, staging and outbound at full tilt at the same time.
The commercial cost isn’t abstract. A distribution centre that misses its peak-day service commitment on the one week that matters most, Christmas fulfilment, a major promotional launch, a tender delivery deadline, does more damage to a customer relationship than a dozen smooth average days ever earns back.
Experienced operators don’t ask what average daily volume looks like. They ask what the worst realistic day looks like, and where the building breaks first.
Designing for peak doesn’t mean building a facility sized for the worst day of the year and running it half empty for the other 350. That’s the opposite failure, and it’s just as expensive in a different way.
The practical approach is to model peak-day throughput as its own scenario early in the design process, identify exactly where the bottlenecks sit, and solve those specific bottlenecks rather than scaling the whole building up. Sometimes that means an extra two dock doors and a slightly larger staging buffer, not a bigger footprint. Sometimes it means redesigning the pick path so inbound and outbound aren’t competing for the same aisle during the peak window. Sometimes it means banking land or shell space that can flex into extra staging or cross-dock capacity for the six weeks a year it’s actually needed, rather than paying for that capacity permanently.
Automation decisions should follow the same logic. A conveyor or sortation system sized for average volume can become the single hardest constraint in the building during peak, because unlike a person, it can’t work faster under pressure. Automation should be sized, or built with bypass and manual overflow options, around the peak scenario, not the average one.
Design a building around peak-day throughput and average-day operations run themselves, because average is always inside the envelope you’ve already tested. Design around average and peak will find every weak point in the building, the week you can least afford it.
Talk to LCA Savills. We have been designing, justifying and delivering distribution centres across Asia since 1999, as operators, not report writers. Start with a Warehousing Health Check.
Average-day design sizes dock doors, staging and pick paths around typical daily volume, assuming overtime and extra labour will cover spikes. Peak-day design models the worst realistic day as its own scenario, identifying where dock capacity, staging space or pick paths run out under maximum load. The difference matters because labour flex can absorb volume increases only until the physical building runs out of room, at which point headcount no longer helps.
Retail peaks are outbound-heavy, calendar-driven spikes tied to fixed dates like Christmas or Chinese New Year. FMCG peaks are promotional, often known only weeks in advance, and land on inbound and staging as much as outbound. Pharmaceutical peaks are frequently regulatory or tender-driven, arriving as large, low-frequency volume blocks rather than seasonal patterns. Each shape stresses a different part of the building, so the design response has to match the sector’s actual peak behaviour.
There’s no fixed multiple that applies across sectors. What matters more is identifying where peak volume lands, inbound, staging or outbound, and sizing that specific function to handle it. Overbuilding the whole facility for a peak that occurs a few weeks a year wastes capital. The more reliable approach is targeted capacity at the actual bottleneck, sometimes just extra dock doors or staging buffer, rather than a larger overall footprint.
Common signs include trucks queuing outside because dock doors can’t turn around fast enough during busy periods, staging areas overflowing into aisles, pick paths becoming congested when inbound and outbound run simultaneously, and automation or conveyor systems becoming the hardest constraint precisely when volume rises. If problems only appear during promotional periods, tenders or seasonal peaks, the facility was almost certainly modeled on average throughput.
Only up to a point. Overtime and casual labour can increase throughput when the building has spare physical capacity, such as unused dock doors or staging space. Once the building itself is full, whether that’s staging area, aisle space or dock capacity, additional headcount has nowhere to work. At that point, service failures during peak periods reflect a design gap, not a labour shortage.
