Most private-label projects fail slowly rather than dramatically. The product is fine, the supplier is competent, and yet the launch slips twice, the unit cost lands higher than the model assumed, and the second order is somehow harder than the first. Almost always, the cause traces back to a handful of questions that were left open at the start. Here are the twelve worth settling before anything is produced.
The product questions
1. What exactly is the product, in writing? Not "dried apricots" but the grade, calibre, moisture content and treatment. A specification that fits in one sentence is a specification that will be interpreted differently by two people.
2. Which of these do you actually need to control? Some parameters drive your product's identity and some do not. Knowing which is which stops you paying for precision that no customer will notice, and stops you economising on the one variable that defines the product.
3. What is the sample approval process? Who signs off, against what criteria, and how many rounds are budgeted. Sample rounds are where timelines quietly disappear.
The packaging questions
4. What format, decided alongside the product rather than after it? Format drives filling method, shelf life, print lead time and unit cost at the same time. Choosing it late forces a compromise on the product or a delay to the launch.
5. What are the one-off costs, and across how many units? Plate, die and design preparation cost the same for a small run as a large one. On a first order they land visibly per unit; across a larger run they thin out. A quotation that looks expensive at low volume is frequently a one-off cost being amortised across few units — which is a different problem from an expensive product.
6. What is the print lead time? It is nearly always longer than the production lead time and it is the item most often forgotten in a launch schedule.
The compliance questions
7. Who is responsible for label compliance? The brand owner, in the destination market. A supplier can prepare artwork, supply ingredient and allergen data and flag known requirements — but the legal responsibility sits with the party whose name is on the pack. Agree this explicitly.
8. What does the destination market require that you have not checked yet? Language, allergen declaration format, nutrition presentation, importer details. Each is routine on its own and each has delayed launches.
9. What batch documentation comes with each delivery, and is it enough for your channel? A marketplace listing and a supermarket listing ask for different evidence.
The commercial questions
10. How small can the first order be while still teaching you something? Small enough to learn from, large enough to be worth producing. The purpose of a first order is to validate product, packaging and channel and to generate real sell-through data. Committing to volume before that evidence exists turns a learning exercise into an inventory problem.
11. What happens if the first run has a defect? Agree the remedy before you need it. This conversation is straightforward in advance and adversarial afterwards.
12. What does the second order look like? If reordering requires repeating the whole process, something in the first order was structured wrong. A well-run first order leaves behind a written specification, approved artwork and a documentation template that make the second one routine.
Where we sit in this
TeraVella distinguishes plainly between what we produce and what we source. Tea bags and blends, dried fruit, natural soap, rose water and essential oils come from our own production network. Honey, tahini and confectionery, spices, nuts, olive oil and beverage powders are sourced through verified supplier partners with batch-level documentation. In both cases you deal with one point of contact for specification, packaging and shipment, and prices and minimum quantities are confirmed at quotation rather than published — because on a private-label project the honest answer depends on format, volume and destination, and a fixed number published on a page would be a fiction for most of the people reading it.
