What you have to write down
Updated
The three
Three records cover almost everything anybody will ever ask you for, and all three are things you already know at the time and will not be able to reconstruct afterwards. That asymmetry is the whole argument for writing them down as you go rather than at the end of the day.
- A lot code on the jar, which has to appear on the label and resolve to one batch.
- The batch record, with weights as used, the stop reading, the process and the jar count.
- A shelf-life statement, which for a correctly made high-acid preserve is long, and which you should be able to say how you arrived at.
What it buys you
Records are not paperwork for its own sake. They are the difference between diagnosing a problem and guessing at it, and between a question from an inspector and a problem with one.
Containment. A complaint traced to one lot is one day's production withdrawn. The same complaint with no traceability is everything you have sold, which on a seasonal product is the whole year.
What also has to be recorded
Where the batch went. Two of the published failures are storage failures, so the question "is this the recipe or the route" can only be answered if the destinations were written down.
The minimum that works
If all of this sounds like more than you will sustain, there is a shorter version that keeps most of the value. Doing the short version consistently beats doing the full version for three weeks. A log with gaps in it is worse than a short log kept faithfully, because the gaps are always in the weeks that turned out to matter.
A notebook with one line per batch, and the lot code as the date. It does not need a system. It needs to exist and to be the same every time, which is what the batch log is for.