Blog/Logística

Ecommerce logistics for demand spikes: a step-by-step plan

Ecommerce logistics for demand spikes: forecast orders per day, protect stock, scale picking and split deliveries without errors or delays.

Ecommerce logistics for demand spikes: a step-by-step plan

A sales spike can be great news on Shopify and a real headache in the warehouse. If 500 orders come in where you'd normally see 80, it's not enough to just ask the team to move faster. Ecommerce logistics for demand spikes relies on three things: capacity planned by day, stock validated before the campaign, and processes that don't lose accuracy when volume multiplies.

In short: forecast how many orders will come in each day (not just the total), check stock on the SKUs driving the campaign, set up wave picking with scanning, split deliveries across several couriers, and leave capacity for the returns that follow. If your team can't absorb the spike without hiring and training people every few months, consider a logistics operator that can scale for you.

Black Friday, an influencer campaign, a product launch, sales season or an email to your whole list all have the same effect: they concentrate orders into a few days that need to be picked, packed and delivered without losing control. The customer doesn't care about your internal effort; they care whether the order arrives on time, complete and well presented.

Why spikes break internal logistics

Problems don't show up at order number 20. They show up when order 200 comes in at the same time as customer queries, a restock, three address changes and a collection the courier hasn't completed. If inventory lives in a spreadsheet, picking depends on one key person, or labels are created by hand, every extra order multiplies the risk of error.

The cost isn't limited to hiring temps either. There are training hours, extra space, shifts, consumables, courier coordination and reshipments for badly prepared orders. A campaign that's profitable on acquisition can lose its margin at the last mile if the operation wasn't designed to absorb it.

The goal isn't to size your structure for the worst day of the year. It's to have an operation that grows during the campaign and returns to its normal size once it's over, paying for the capacity you use rather than turning every spike into a permanent expansion.

Ecommerce logistics for demand spikes: what to plan

Preparing doesn't mean stockpiling boxes or doubling inventory on a hunch. It means making data-driven decisions and turning them into a plan with dates, owners and clear limits.

1. Forecast orders per day, not just total sales

A monthly revenue forecast is of little use if it doesn't say when the orders will land. Review previous campaigns by day and by time slot, and break them down by channel, country, SKU, bundles and delivery type. Early access for returning customers creates a very different curve to an open promotion backed by heavy ad spend.

Work with three scenarios: baseline, likely and maximum. The maximum isn't there to scare you, but to know what happens if the campaign performs better than expected. Define how many orders a day you can ship without delays, and at what point you need to bring in more resources or scale back the promotion.

Don't forget the tail. The Monday after a big campaign brings a wave of changes, cancellations, returns and queries. Capacity needs to cover those days too, not just the moment checkout spikes.

2. Protect stock on the SKUs driving the campaign

During a spike, a stock error turns into hundreds of orders sold against units that don't exist. Store inventory and physical stock need to be synced in real time, with control over inbound stock, reservations and outbound stock by SKU. If you sell across several channels, check out how to sync stock across channels to avoid overselling before launching anything.

Classify your catalogue by turnover. Campaign products should sit in accessible picking locations, close to packing, with restocking planned in. Slow movers can move to secondary positions during those weeks. This relocation is one of the biggest time-savers, and one of the most overlooked.

Safety stock depends on replenishment lead time, how variable demand is, and how critical the product is. There's no universal percentage. An item made nearby with weekly restocking needs less of a buffer than an imported one with several weeks in transit, which also needs a cut-off date for the last order to the supplier.

Before launching the campaign, confirm pending deliveries have been received, resolve any discrepancies, and check that each bundle is correctly set up: if you sell a bundle of three items, the system should deduct each component. If you're giving away a gift with purchase, that rule needs to reach the warehouse without handwritten notes.

3. Design picking that gains speed without losing accuracy

Preparing more orders isn't about walking faster. Sustainable speed comes from grouping tasks: wave picking collects several orders in a single run, especially when they share SKUs. In campaigns built around a hero product, many orders are single-SKU and can be batch-prepared.

Double-checking is a must for products with sizes, colours or high value: barcode scanning at picking and a second check before sealing the box. It adds seconds, but prevents a return, a reshipment and an unhappy customer. If you want to build the full method, here's how to avoid errors in order preparation.

Also define exceptions before they happen: incomplete address, damaged product, partial stock shortage, last-minute changes or customisations. Every case needs a written rule. If the team has to ask what to do with each one, the flow stops right when volume is highest.

4. Prepare consumables and packaging with a margin

A classic mistake: there's product, there's staff, and a specific box size runs out on day two. Calculate boxes, mailers, filling, tape, labels and cards based on the maximum scenario, not the likely one. If you use branded packaging from an external supplier, order weeks in advance, since print lead times stretch out during peak season.

Simplify during the spike too. If your packaging has five manual steps, consider cutting back the ones that add less value (for example, one insert instead of three) so packing doesn't become the bottleneck.

5. Spread the risk across couriers

Relying on a single courier during the highest-volume week is an unnecessary gamble. Delays come from route saturation, postcode restrictions or incomplete collections. Working with several lets you choose the service based on destination, weight and urgency; this article on multi-courier shipping covers when it pays off and how to automate it.

It's not about always choosing the cheapest option. For low-value orders, an economy service protects your margin; for gifts with a deadline or premium products, speed and tracking come first. Also check daily collection capacity and cut-off times: an order ready at 6pm loses its value if it misses the collection and waits another day. Every cut-off needs to match what you promise on the store.

6. Reserve capacity for returns

Returns follow the spike, and in fashion or gifting they can be substantial. If no one's reviewing them, sellable stock sits in a corner while the store shows it as out of stock. Assign people and space to receive, inspect and return sellable items to stock within one or two days.

Timeline: what to do and when

  • 6 to 8 weeks before: daily forecast with three scenarios, supplier orders, packaging and consumables orders, and a conversation with couriers about volumes and collections.
  • 3 to 4 weeks before: receiving campaign stock, relocating fast-moving products, setting up bundles and gifts, and written exception rules.
  • 1 week before: stocktake of key SKUs, a trial day with simulated orders through picking and labelling, and training for temporary staff.
  • During the spike: check pending orders, stock-outs and collections twice a day. If a SKU drops below the minimum, hide it or switch it to pre-order before overselling.
  • Afterwards: return items to stock within one or two days, analyse errors and delays, and write notes for the next campaign.

If your main spike is in November, you'll find the detail in the guide to preparing Black Friday logistics.

Example with rough figures

Picture a cosmetics store that ships around 15 orders a day. For Black Friday it forecasts a likely scenario of 120 orders a day for four days, and a maximum of 200.

As a rough benchmark, one person with a well-organised warehouse and simple orders can pick and pack between 15 and 30 orders an hour, depending on the product and packaging. With two people, 120 orders seem to fit within a working day. But in practice you have to subtract time spent printing labels, restocking locations, handling issues and preparing the collection. And if the maximum scenario hits, a day's delay builds up and drags into the rest of the week.

In cases like this, the bottleneck is rarely hands on deck. It's usually poorly located stock, labels generated one by one, or a single daily collection. Fixing those three things adds more capacity than bringing in a third untrained person.

The metrics that show whether you can scale

Order count alone isn't enough. During a spike, track quality and speed indicators together:

  • Percentage of orders shipped within 24 hours: your responsiveness.
  • Picking accuracy: catches errors before they turn into support tickets.
  • Difference between physical stock and system stock: tells you whether your inventory is still reliable.
  • On-time deliveries by courier and area: if one performs well in the city but struggles in certain regions, adjust the assignment rule.
  • Logistics cost per order, separate from advertising: picking, packaging, shipping, issues and returns.

With a clear cost per order, you can compare campaigns, see which products can sustain free shipping, and decide whether to raise the minimum order threshold during the promotion.

Mistakes seen every peak season

  • Launching a promotion with stock still pending receipt, selling units that are still on the lorry.
  • Bundles set up as their own SKU that don't deduct their components.
  • Promising
Back to blog