How to plan and launch a merch drop

A drop compresses a normal retail quarter into about four hours. Demand arrives in a spike rather than a curve, which means every weakness in your operation gets tested at once — stock accuracy, sync speed, packing throughput, carrier capacity and your own ability to answer messages.

The failure modes are predictable, which is good news. Almost every drop that goes badly does so for one of five reasons: the product arrived late, the quantities were wrong, the store oversold, the orders shipped slowly, or nobody planned what happens the day after. Here is how to design each one out.

Work backwards from the launch date

Fix the launch date first, then schedule everything against it. Realistic lead times for a first-time drop:

That is roughly ten to sixteen weeks from brief to launch. The buffer is the line people delete first and regret most; production and freight are exactly where slippage happens, and a drop with a promoted date cannot move.

If a date is immovable and the timeline does not fit, cut scope — fewer items, simpler decoration — rather than cutting sampling or buffer.

Decide quantities with a method

Quantity is the hardest call in a drop, and both errors are expensive: sell out in minutes and you have left money and goodwill behind; overproduce and you own stock that costs storage every month until you discount it.

A workable approach:

• Anchor on a real number. Previous drop sell-through, email list size, engaged follower count — something observed rather than hoped.
• Apply a conversion assumption you can defend. For a warm, engaged audience, a small single-digit percentage of the reachable list buying is a reasonable planning figure. Be honest about how warm your audience really is.
• Split by SKU using known ratios. Size curves are not flat. Without your own data, apparel typically skews to the middle sizes.
• Deliberately under-produce a hero item. Scarcity is a legitimate part of the drop mechanic, and a genuine sell-out builds the next launch.
• Over-produce nothing. If in doubt on a non-hero item, order less and plan a reorder.

Write the assumption down. After the drop, compare it with what happened — that comparison is the most valuable asset the first drop produces.

Prevent overselling before it happens

A spike is precisely the condition under which inventory sync fails. Two people buying the last unit in the same second is a systems problem, not bad luck.

Before launch:

• Confirm inventory sync is event-driven, not polled. A 15-minute polling interval is an invitation to oversell during a spike.
• Confirm the warehouse is the source of truth for physical quantity.
• Zero out legacy stock locations so your platform is not adding phantom availability.
• Decide whether “continue selling when out of stock” is off — for a drop it almost always should be.
• Hold a small buffer off-sale to cover damages and mis-picks. Two percent is cheap insurance against telling a customer their order does not exist.

The technical detail behind this is covered in how stock sync actually works. For a drop, verify it rather than assume it — including the two-browser test where you try to buy the last unit twice.

Brief the warehouse like it is an event

A drop is not a normal Tuesday, and a fulfilment operation cannot absorb one silently. Tell them, in writing, at least two weeks out:

• The launch date and time, to the hour
• Your expected order volume in the first 24 hours and the first week
• Peak concentration — most drops take a large share of total orders in the first hour
• Whether packing is standard or special: inserts, tissue, stickers, numbered cards
• Whether any orders are held and shipped together on a set date
• Who to call, on both sides, if something goes wrong on the day

Then ask what they need from you and when. A provider who knows a spike is coming can staff for it; one who finds out from the order queue cannot.

The special-packing question deserves attention. Every additional manual step is priced per order and consumes time on the day. Design the unboxing so it is repeatable — a printed insert card achieves most of the effect of a handwritten note at a fraction of the throughput cost. This is why drops and custom packaging get specified together.

Plan launch day itself

Have a written runbook. On the day:

• Publish products as scheduled, not manually at the moment of launch
• Have someone watching orders per minute and stock levels
• Have someone watching support channels, because questions spike with orders
• Know the carrier collection cut-off and whether a same-day dispatch promise is realistic for hour-one orders
• Decide in advance what you do if you sell out in ten minutes — a waitlist captures demand that would otherwise leave
• Decide in advance what you do if you do not sell out — a planned second push beats a panicked discount

Say clearly what customers should expect. “Orders placed today ship within three working days” is far better than silence followed by an inbox full of “where is my order”.

The day after matters as much

The post-drop window is where a one-off becomes a programme:

1. Reconcile. Units sold against units received against units remaining. If the numbers do not agree, find out why now, not at year end.
2. Measure sell-through per SKU, not in aggregate. One item selling out while another sits is the most useful signal you will get.
3. Read the returns. Reason codes tell you whether a size ran small or a print failed — both fixable at product level rather than absorbed forever.
4. Decide reorder or retire per item, quickly. Momentum decays.
5. Write down what broke. Every drop reveals a bottleneck. Fixing one per launch compounds fast.
6. Deal with residual stock deliberately — bundle it, hold it for the next launch, or clear it. Stock with no plan becomes a write-off by default.

The five failure modes, and their fixes

• Product arrived late → build a two-week buffer and never delete it
• Quantities were wrong → anchor on observed data, plan a reorder path
• The store oversold → event-driven sync, one stock location, a held buffer
• Orders shipped slowly → brief the warehouse as an event, confirm cut-offs
• No plan for the day after → reconcile, measure sell-through, decide reorders

Running drops repeatably rather than heroically is the point of treating them as an operation — which is the model behind branded product lines and drops.

Frequently asked questions

How far in advance should I start planning a merch drop? Ten to sixteen weeks for a first drop, including a two-week buffer. Reorders of an already-approved product can be much faster because sampling is done.

How much stock should I produce for a drop? Anchor on observed data — previous sell-through or engaged audience size — and a conversion assumption you can defend, then split by SKU using realistic size ratios. Under-producing a hero item is a legitimate tactic; over-producing everything is not.

How do I stop my store overselling during a launch? Use event-driven inventory sync, make the warehouse the single source of truth, zero out unused stock locations, turn off overselling settings, and hold a small buffer back for damages.

Should I ship orders as they arrive or all together? Shipping as they arrive gets product to customers faster and spreads the packing load. Holding to a single dispatch date suits numbered or simultaneous-reveal drops, but tell customers clearly which you are doing.