Where production actually stalls.
Packaging is the most common one. Bottles, caps, sleeves, cartons and labels all have their own lead times and minimums, and they are frequently longer than the production run they are waiting for. A product can be made and sit for weeks because a cap is late.
Artwork is the second. Label copy that has not been checked against the destination market gets sent back, and a reprint is measured in weeks. Getting the label reviewed before it is printed is the cheapest week you will ever spend.
The third is testing. Certificates and analyses take time, and if nobody has said which ones this product needs, they get started after the batch is made instead of alongside it.
Decide the market before the product.
Ingredients, doses, claims, label format and required documents all follow from where you intend to sell. The US and the EU differ on all five, and a product built for one is not automatically sellable in the other.
This is the single decision that removes the most rework later. It is also the one most often deferred, because it feels like a commercial question rather than a production one.
If you genuinely intend to sell in both, say so at the start. There are formulas and label layouts that work in both, and they are much easier to design than to retrofit.
Small first, then scale.
A smaller first batch costs more per unit and less in total, and it buys you the right to be wrong about flavour, pack size or positioning while being wrong is still cheap.
It also gets you a real product to sell, photograph and send to people months earlier, and a product in hand changes conversations with retailers and partners in a way a rendering does not.
Scale when the thing you are scaling has been confirmed by orders. Larger runs are a reward for knowing what sells, not a way of finding out.
