“We can’t make this” is almost never one sentence. It is five different sentences that sound identical, and which one you are hearing determines whether the project is dead, delayed by ten weeks, or ten minutes away from a yes. Sometimes it means the tooling does not exist. Sometimes it means the volume is too small to justify starting the machine. Sometimes it means the bottle will run, but not reliably enough to ship. And occasionally it is the most useful answer a supplier can give you: an honest no, from someone who knows their own process.
The five sentences hiding inside “we can’t”
Before you can respond to a refusal, you have to know which of these you received.
1. The tool does not exist. Nobody has a mold for this shape, and building one is a capital decision.
2. The volume is below what the process needs. The machine will run, but the setup and changeover costs cannot be spread across a small order.
3. It will run, but not reliably. The design pushes against what the process can hold consistently, so yield drops and scrap rises.
4. The bottle is fine — the closure is the problem. The neck finish or the pump you want is not compatible with how the bottle is made.
5. This is genuinely outside our process. They do not own the technology, and saying so early is doing you a favour.
Only the last one is a real dead end. The other four all have a workable path, and three of them are usually cheaper to solve than buyers expect.
“The tool doesn’t exist” — the mold question
This is the most common meaning, and it is not a refusal so much as an invoice. A new mold is a real capital cost with a real lead time, and a factory will not quietly absorb it into a unit price on a 5,000-piece order.
The questions to ask are: how close is your existing tooling; what does a new tool cost; and what order quantity would justify it. In our case the first question matters more than people expect — with more than 2,000 molds already built, a very large share of “custom” requests turn out to be 80% solved by a tool that already exists, plus a different shoulder or a modified neck. That is a fraction of the cost and a fraction of the lead time of starting from zero.
If a new tool genuinely is needed, ask what volume makes it worthwhile. Often the honest answer is that the mold pays for itself somewhere in the tens of thousands of pieces, and the buyer’s real decision is whether to commit to that volume or to adapt an existing shape.
“Your volume is too low” — the economics of starting a machine
Every production run carries fixed costs regardless of length: material changeover, colour changeover, machine setup, tooling fitting, first-piece approval and the scrap generated getting to a stable process. On a long run those costs disappear into the unit price. On a short run they dominate it.
Colour makes this sharper. Each shade needs its own concentrate and its own purge of the machine between runs, which is why a minimum per colour exists separately from a minimum per order. A factory quoting a 5,000-piece minimum is usually saying 5,000 of one specification — not 5,000 split across five colours.
If your volume is genuinely small, ask whether your colour can be matched to a stock shade, or whether a standard natural or white bottle with decoration would get you the shelf presence you need. Decoration often buys the differentiation that a custom colour was meant to buy, at a much lower minimum.
“It will run, but not reliably” — the yield problem
This one is a quality warning disguised as a capability problem, and it is the answer buyers most often push back on — and most often regret pushing back on.
Certain geometries are simply hard to hold: very sharp internal corners that thin the material as it stretches, extremely wide flat panels that want to warp or sink, ultra-thin necks that distort, and heavy bases paired with very light walls. A prototype can be coaxed out once. Running it at 200,000 pieces a day without a rising defect rate is a different question entirely.
The right response here is not to insist. It is to ask which feature is causing the problem and what the minimum change would be. Often a 1 mm radius, a slightly thicker shoulder or a ribbed panel solves it, and the bottle looks the same on shelf.
“The bottle is fine — the closure is the problem”
This is worth separating out because it is so frequently misread as a bottle limitation. The neck finish determines which closures, pumps, sprayers and droppers fit. If you want a specific pump with a specific collar diameter, the bottle neck has to be built to suit it — and if the bottle is stretch-blow-molded PET, the neck is formed in the preform stage and has to be designed in from the start.
The practical fix is to choose the closure first and let it dictate the neck, rather than falling in love with a bottle shape and then hunting for a pump that fits. Buyers who sequence it this way almost never hit this refusal.
“This is genuinely outside our process” — the honest no
Sometimes the answer really is no, and it is worth recognising it when you hear it, because a factory that admits a limit is giving you more useful information than one that says yes to everything.
A supplier who tells you early that a job belongs with a different process saves you a mold cost and three months. The warning sign is the opposite: an enthusiastic yes to everything with no questions and no conditions attached.
Three questions that tell you which one you are hearing
Ask these in order, and the answer usually reveals the category immediately.
“Is this a tooling issue, a quantity issue, or a process limitation?” Most suppliers will simply tell you if asked directly, and the question signals that you understand the difference — which changes the quality of the answer you get.
“What is the closest you already make?” This is where existing tooling shows up, and it is the fastest route from no to yes.
“What would need to change for this to be possible?” A specific answer — a radius, a neck, a quantity, a colour — means you are negotiating. A vague answer usually means you are hearing category five.
Five redesign moves that turn a no into a yes
These come up again and again, and none of them compromise how the product looks on shelf.
Soften sharp internal corners. Move a heavy base slightly thinner, or add a recessed panel. Standardise the neck to match an available closure. Choose a stock colour and differentiate with decoration instead. Accept a 5–10% change in proportion so the shape can use an existing tool. In most cases the customer never notices, and the difference in cost and lead time is substantial.
What we do when we say no
When we turn something down, we try to say which of the five it is, because a bare no wastes your time. Since 2003 we have run more than 30 production lines across 15,000 square metres, exporting to more than 30 countries, and the overwhelming majority of “impossible” requests resolve into one of the first four categories — usually the tooling one, and usually solvable by adapting something already in the tool room.
The version buyers should be most suspicious of is not the no. It is the immediate, unqualified yes.
Questions buyers ask us most
Is “we can’t” just a way of saying the order is too small?
Often, but not always — which is why you should ask which one it is. If it is a volume issue, the fix is usually a standard colour, an existing mold or a higher quantity, not a different supplier.
Why won’t a factory just absorb the mold cost into the unit price?
Because the mold is paid for once and the unit price is paid for every order. If the tooling is absorbed into a small first order and you never reorder, the factory has funded your tooling. That risk gets priced in, one way or another.
If one factory says yes and another says no, who is right?
Ask both what assumption they are making. A yes based on “we will try it and see how yield goes” is not the same as a yes based on “we run this geometry every week”. The second is worth more than the first.
How do I know if a refusal is about quality risk?
If the conversation turns to wall thickness, sharp corners, warping or yield, you are hearing a yield warning. Treat it as engineering feedback rather than a sales objection, and ask which specific feature is the problem.
Can I send my design to several factories and compare?
Yes, and you should — but send the same information to each, including quantity and closure choice. Otherwise you will be comparing answers to slightly different questions, which is how buyers end up choosing the wrong supplier for the right reasons.