Shelf-life testing tells you how long your product really lasts, and batch records tell you what went into each batch and where it went. Both cost something: jars you do not sell because you keep them as samples, lab fees and minutes of writing things down. Spread over what you sell, it is usually little per unit. In return, you stop guessing the date on the label and, if a batch goes wrong, you pull only that batch rather than everything you have out in shops.
We do not cover what the law requires for your product here. Check with your local food safety authority which registrations, tests and records apply to you.
Why test shelf life
Without testing, the best-before date is a guess, and it usually goes wrong one of two ways:
- Too short. You put two months "to be safe". Shops pull your jars early, send them back or ask you to swap them, and you make smaller batches than you could. The guide on shelf life and the cost of unsold stock shows what that waste costs.
- Too long. You put six months because "jam keeps". If mould shows up at four months or a lid starts to dome, the problem reaches your customer and your name.
With testing, the date comes from what you saw in your own product, made with your recipe, your jars and your filling method.
What tests you can do
Your own checks with retained samples
From every batch, keep a few jars in the same conditions they are sold in: on a shelf at room temperature, or in the fridge if that is how you sell. On dates you choose (every two weeks, for example), open one and note:
- Appearance: colour, separation, bubbles, mould.
- Smell and taste.
- Texture.
- The lid: whether it is domed or air escapes when you open it.
When you see a change you would not accept in a jar you sold, you know how far your product goes in those conditions. Your date should land before that point, with room to spare.
This does not replace a lab analysis. It is your own information to decide with, and something concrete to show if anyone asks.
Lab analysis
A food lab can test your product at different points in its life. Ask which tests make sense for a jam, a chutney or a sauce, and which your authority expects. Get a written quote. We do not give prices here because they depend on the lab and the test.
What to put in your batch records
Batch records are a sheet (paper or phone) with one line per batch. At minimum:
- Batch code and production date. The same code goes on the jar label.
- Ingredients and supplier, and if you can, the lot number or purchase date of each.
- Quantities of each ingredient and yield: how many jars came out.
- Process: cooking time and temperature, how you filled, who made it.
- Samples kept: how many and where.
- Where it went: which shop, which market or which orders got jars from that batch.
- Notes: anything unusual, from fruit that was riper than usual to a lid that did not seal well.
The guide on food safety basics for small kitchens covers the kitchen practices worth recording alongside this.
Worked example: strawberry jam in jars
You sell strawberry jam in 340 g jars. You sell 500 jars a month in four batches of 125. Your loaded cost per jar is $3.60.
Retained samples. You decide to keep 2 jars per batch: one to open at your checks and one spare in case of a complaint.
- Jars kept per month: 2 × 4 = 8
- Cost: 8 × 3.60 = $28.80 a month
- Per jar sold: 28.80 ÷ 500 = $0.058
Batch records. You timed it: writing up a full batch takes 10 minutes. You value your hour at $15.
- Time per month: 10 × 4 = 40 minutes
- Cost: 40 ÷ 60 × 15 = $10.00 a month
- Per jar sold: 10.00 ÷ 500 = $0.020
Total before lab tests: 0.058 + 0.020 = $0.08 per jar.
Lab analysis. Get a quote and spread it over the jars you will sell with that recipe. If the recipe stays the same for a year, that is 500 × 12 = 6,000 jars. Every $100 of lab fees adds 100 ÷ 6,000 = $0.017 per jar. Multiply that by your quote, in hundreds.
What the records are worth
Now imagine a shop tells you a customer found mould in a jar. You have three batches out, 375 jars in total.
- Without batch records: you cannot tell which batch it came from, so you pull all 375. If you cannot sell them, you lose 375 × 3.60 = $1,350 in cost alone.
- With batch records: you read the batch code on the label, check your retained sample from that batch, and pull only those 125 jars. You lose 125 × 3.60 = $450.
That is $900 in one incident, before the trip to collect the jars, your time and the shop's trust. Against $10 a month of record-keeping, the sum is clear.
And if your tests show the jam lasts longer than the date you were using, you can put a longer date with something behind it. If 25 of every 500 jars used to come back near their date, that was 25 × 3.60 = $90 a month of waste.
How to bring it into your cost and price
Samples and records are part of the cost of a packaged food product, just like the jar and the label. Add them as a cost per unit: in the example, $0.08 plus your share of lab fees. To see how much your margin moves with those cents, use the markup and margin tool.
If you change the recipe, the jar or the way you fill, the shelf life can change too. Test again and start a new run of batch codes. And if the batch code is printed or stickered, it is one more part of the label; the guide on label costs per unit shows how to add it in.
Do it with the calculator
In the food cost calculator, add a line for "samples and batch records" with your cost per jar ($0.08 in the example, plus lab fees spread out), and you will see your full loaded cost before setting prices for shops and markets.
Frequently asked questions
Can I put a best-before date on without testing?
You can estimate one, but it is a guess. Without testing you tend to set a date that is too short, which brings returns, or too long, which puts customers at risk. What the rules require for your product, check with your local food safety authority.
How many samples should I keep from each batch?
Enough to open one at every check you have planned until the date on the label, plus one spare in case of a complaint. Set the number by how many checks you want to do.
What is a batch code and how do I make one?
It is a code that identifies everything you made together in one run. It can be as simple as the production date plus a letter, such as 181026-A, as long as it also goes in your records.
Do I need a lab to test shelf life?
Your own observations give you information, but they do not replace a lab analysis. Which tests you need, and whether they are required for your product, is for your local food safety authority to say.