Packaging "Presale and Batch Delivery": 3 Real Boundaries — Print First and Wait / Print Once and Collect in Batches / Channel-Specific Color Schemes

In April 2026, a customer launching an emerging skincare brand approached us.

"We want to start with 100,000 shopping bags and boxes," the customer said. "Launch them first, and reorder if they sell."

We asked: "100,000? What if they don't sell?"

The customer replied casually: "If they don't sell, we can slowly work through the inventory."

Later, we calculated the full economics of the order and found that the printer had assumed all the market risk, while the customer had assumed none. This was not a cooperation structure that "gave the customer an opportunity"; it was a structure with a completely unbalanced allocation of risk.

That does not mean the customer’s request was unreasonable—new brands do need initial market coverage, and printers do not want to miss new customers. The problem is that in the phrase "Make 100,000 first, and reorder if they sell," the customer front-loads the costs and leaves the risks with the printer, while the reorder usually never happens in the packaging industry.

Today, we explain the three real cost structures involved.

Boundary 1: Print First and Wait—All Three Major Costs Are Incurred the Moment Production Begins

Packaging has a cost structure unlike that of most manufacturing: most of its expenditure occurs before production and is irreversible.

We broke down the costs of this order:

The key point is the second row: the materials are purchased for 100,000 units and match only those 100,000 units. If only 30,000 units are eventually sold, the paper corresponding to the remaining 70,000 units becomes slow-moving inventory in our warehouse. Selling it to another customer would constitute a breach of contract because doing so would compromise the customer’s exclusivity, while keeping it ties up capital.

This is what we internally call the core risk of "printing first and waiting": the issue is not that unsold units are difficult to work through slowly; it is that the printer must absorb dedicated materials for 100,000 units.

There is another often-overlooked cost: the locking in of colors and processes. If the customer uses a particular Pantone color for the initial run, the 100,000-unit order locks in that color. If the reorder uses a different production batch, paper, or printing method, the color may shift—a disaster for a skincare brand, because packaging from different batches displayed together on the same shelf will reveal obvious color differences.

Therefore, in the "Make 100,000 first, and reorder if they sell" model, the act of reordering itself carries a cost far higher than the customer may imagine. A reorder is not simply "printing the same thing again"; it means "creating a new plate, buying more paper, and matching the color again."

Boundary 2: Print Once and Collect in Batches—Shipping Is Saved, but Capital and Risk Are Exposed

The second boundary is considerably less severe than the first. Its logic is straightforward: print once and collect the order in batches.

The customer’s usual request is: "We will collect the goods weekly and take part of each shipment." The benefits are clear: the customer does not need to stock as much inventory at once, reducing the pressure of market rollout.

However, the cost difference between "print once and collect in batches" and "print once and collect once" is much larger than the customer may expect:

First, batch picking and shipping. Each collection requires separate picking, counting, packing, and shipment. Sending one full container and sending the same volume in 10 courier shipments differ not only in freight but also in packaging consumables and labor.

Second, working-capital turnover. If the customer pays in full before collecting the goods, the capital burden remains with the printer. If the customer pays after delivery, the cycle from warehousing to payment must be borne separately for each batch. Collecting the order in 5 batches requires 5 capital turnover cycles.

Third, inventory management. The goods must be separately identified and stacked apart from the customer’s own inventory, with both parties counting and confirming each collection. These tasks themselves require labor.

Our approach is therefore: "Print once and collect in batches" is feasible, but the cost of batch shipment must be explicitly shown in the quotation—allocate the cost according to the customer’s planned number of collections rather than assuming it is included. Once the customer sees the figure, they can adjust the collection schedule—for example, combining 5 collections into 3—instead of discovering only during execution that every collection costs more money.

One further point must be clarified in advance: batch collection has a "batch period" tail. If the customer’s first collection occurs 3 months later and the final collection occurs 10 months later, those 100,000 boxes remain in storage for 10 months, not 3. Some materials may deteriorate during this period—paper may absorb moisture and warp, colors may shift slightly through light exposure, and edges may acquire compression marks. These are not printing quality problems but storage problems. Batch collection agreements must therefore specify the maximum age of a batch still eligible for delivery—for example, the printer may deliver batches older than 6 months at the costs prevailing at that time.

Boundary 3: Channel-Specific Color Schemes—One Design Printed as Two Different Colors

The third boundary is the most technical and most frequently overlooked in presale models: using different color schemes for different channels and producing them in separate batches.

A typical scenario: the e-commerce channel uses Version A, which emphasizes the main visual, while offline channels use Version B, which gives greater prominence to product information. The customer wants to "print the e-commerce version first, and print the offline version if sales are good."

The approach itself is not a problem, but it has three real boundaries:

Boundary 1: Materials cannot be shared between different versions. Versions A and B require two different die-cutting plates and two separate material sets. The paper can be shared (the same batch of white cardboard), but the plate fees must be paid twice, and press startup waste must also be borne twice. If the designs differ enough to require different papers to maintain channel positioning, even the paper must be purchased in two separate batches.

Boundary 2: Whether color consistency can be guaranteed depends on the production method selected.

Therefore, if the customer requires packaging for two channels to look indistinguishable when placed together, the determining factors are the material batch and the adjacency of production, not merely the printing process. During quotation, we tell the customer directly: to ensure consistency, both versions must use the same paper batch, be scheduled in adjacent press slots, and have no intervening jobs—which means the two versions are best printed together rather than in separate stages.

Boundary 3: "Print one channel first and wait for feedback on the other" eliminates production-scheduling flexibility.

This sounds like testing before scaling up, but it has a side effect: if 100,000 units have been printed for the first channel while the second channel remains unprinted, the timing, paper, and paper price for the second channel are all locked in—because the dedicated materials for those 100,000 units already occupy production capacity and capital. In effect, you have conducted a "test the market response" experiment using an order of 100,000 units.

If the first channel performs well, this model benefits you because the cost of opening a new plate is avoided and volume increases. If the response is poor, you are locked into a specific volume and a specific point in time.

How Should a Printer Negotiate an Order for "Make 100,000 First, and Reorder If They Sell"?

We generally do not reject such requests. Instead, we turn them from a vague promise into an executable agreement. Four actions are required:

First, replace "reorder if they sell" with specific quantities and deadlines. "Make 100,000 first" should be divided into, for example, "an initial batch of 30,000, with a second batch of 70,000 to be ordered within X months." If the customer cannot even define its replenishment plan, proceed with an "initial batch of 30,000 plus a reserved production capacity allocation." We reserve materials and a press slot for the second batch, but the customer must decide within the agreed period; otherwise, the allocation is released to other customers.

Second, change the plate fee from a one-time charge to an allocable charge. If the customer ultimately produces only 30,000 units, the plate fee should be amortized against the actual 30,000 units rather than charged in full. This lowers the customer’s trial-and-error cost while also reducing the printer’s risk—the adjustment we are most willing to negotiate because it benefits both parties.

Third, specify how the unsold portion will be handled. Choose one of three options: the customer consumes it within the agreed period, with the printer continuing delivery at the original unit price; the customer accepts the remaining units at cost, effectively purchasing them; or the printer disposes of or sells them at an agreed price with the customer’s written consent. Without this clause, "slowly work through it if it does not sell" becomes a courteous remark with no legal effect.

Fourth, if multiple channels and batches are involved, document the conditions required for color consistency. The same paper batch, adjacent scheduling, and no intervening jobs—the three conditions must be included in the production plan; otherwise, consistency depends on luck.

A Message to Customers

In the phrase "Make 100,000 first, and reorder if they sell," the customer understands the cost of the "first" part and assumes the "reorder" part is inexpensive.

In reality, the reorder is not inexpensive: creating another plate, buying more paper, and matching the color again can make the unit cost higher than the initial order when the initial and subsequent volumes differ significantly, because the scale advantage of the first order is no longer available. In other words, this model makes the customer pay more in the “good sales” scenario and leaves the risk with the printer in the “poor sales” scenario.

Our recommendation is: change the model to an "initial batch plus reserved production capacity plus plate fees allocated by actual volume." This reduces the customer’s initial exposure from 100,000 units to 30,000, allows successful products to scale smoothly with more controllable costs, and prevents a large amount of dedicated material from remaining in someone else’s warehouse if sales disappoint.

Ultimately, this is not a case of printers being unwilling to assume risk; risk should be placed with the party capable of bearing it. Customers should bear market risk, while printers should bear production risk—the current formulation merely disguises market risk as production risk.

Further Reading

• The 3 Real Boundaries of Customer "Consignment Inventory": Untransferred Ownership / Slow-Moving Inventory Occupying Space / Who Bears the Cost After a Customer Cancels—presales and consignment are two forms of the same risk
• The 3 Real Cost Boundaries of "One Order vs. Split Orders" in Packaging Procurement: Minimum Order Quantity / Inventory Holding / Version Freezing—the other side of split ordering is version freezing
• 3 Real Solutions for Rush Orders During Peak Season: Expedited Scheduling / Process Changes / Order Transfer—reserved production capacity needs an enforcement mechanism; this article explains scheduling
• 3 Real Methods for Creating Digital Printing "Proof Records": Physical Sealed Samples / Color Cards / Proof Photos—the basis for color matching across batches
• What to Do When a Gift Box Order Is Cut or Reduced? 4 Emergency Solutions for Printing Company Owners—how to handle the unsold side of the equation