Trade term
The reference sample, and the lot behind it
How many packages you open is decided by how big the lot is, and a lot is defined by shared characteristics rather than by the pallet or the delivery. Same packer, same origin, same species, same class, same size, same variety, same packaging: change any of those and you are looking at a second lot that needs a second sample of its own.
The figures
| What a lot is | A quantity of produce which at the time of inspection has similar characteristics, including packer, country of origin, species, class, size, variety and type of packaging |
|---|---|
| Primary sample | A package taken at random from the lot |
| Bulk sample | Several primary samples, enough in total to assess the lot against every criterion |
| Secondary sample | An equal quantity of produce taken at random from the primary sample |
| Reduced sample | A quantity taken from the bulk sample, restricted to the minimum needed to assess certain individual criteria |
| Packages to draw, lot of up to 100 packages | 5 |
| Packages to draw, lot of 101 to 300 packages | 7 |
| Packages to draw, lot of 301 to 500 packages | 9 |
| Packages to draw, lot of 501 to 1,000 packages | 10 |
| Packages to draw, lot of more than 1,000 packages | 15 at least |
| Where the method lives | Annex V of the conformity check regulation, the methods of inspection |
| Our own procedure for retained samples | Not agreed with the grower |
What it changes on an order
You opened three boxes and formed a view
The number of packages is fixed before anyone looks at the fruit. Drawing fewer than the method asks for produces a reading nobody on the other side has to accept, however bad what you found was.
Two classes arrive on one truck
That truck is carrying two lots, and each of them needs its own draw. Mixing the boxes into a single assessment produces a percentage that describes a quantity of produce which does not exist.
You want the count to scale with the size of the problem
It does not. The table steps with the size of the lot alone, from a handful of packages on a small one up to fifteen or more above a thousand, and a worse looking load does not earn a bigger sample. What it earns is a second lot, if the characteristics actually differ.
You want fruit kept back for later
Official sampling exists to judge a lot on the spot, and it gives you no procedure for storing fruit to argue about in a fortnight. Who keeps it, at what temperature and for how long are things to write down in advance.
What this note does not settle
- This method belongs to official conformity checks carried out by inspection bodies. A commercial buyer is neither obliged to follow it nor prevented from borrowing it.
- The tables live with the conformity check rules and get reissued along with them. This wording came through the last recast untouched, which is a reason to check the version in force rather than to stop checking.
- Sampling establishes whether a lot conforms to the standard. It does not establish who caused what it found, or when.
- We have no retained sample arrangement, no agreed inspection point and no third party appointed. None of that exists to describe today.
Source: Commission Implementing Regulation (EU) 2023/2430, Annex V, methods of inspection
The lot decides, and not the delivery
Almost everybody who draws a sample starts from what arrived. The method starts from somewhere else entirely. It asks you to sort what arrived into groups sharing a packer, an origin, a species, a class, a size, a variety and a kind of packaging, and to count only afterwards. One delivery can hold several such groups. One group can run across deliveries that came on different days.
Which is why an assessment written against a truck convinces nobody. The percentage it produces belongs to a mixture the standard has no name for, and the other side can say so without even looking at your fruit. Split the arrival into its groups first, roughly if that is all the dock allows, and the figures afterwards describe something recognisable.
What a sample carries, and what it drops
A properly drawn sample settles one thing and settles it well: whether a quantity of produce conformed at the moment somebody looked at it. It carries nothing about who caused a defect or on which day it appeared. Those are separate questions, answered with dates, photographs and temperature records, and a larger sample never brings them any closer.
Fruit held back for later is a different arrangement again, and official sampling offers no template for it. Who keeps it, at what temperature, for how long, and who is allowed to open it are all terms somebody has to write down. None of that is agreed with our grower, so ask us what it would look like instead of assuming a procedure is already there.
Talk to the grower
Tell us the variety, the sizes and the delivery week you need. You get a delivered price and what we can commit to in writing, from the person who picks the fruit.
The grower handles enquiries. Response times are still to be confirmed.