Overfilling Is Not Playing It Safe: Why Food Factories Should Control Pack Weight, Not Pad It

By Isaac Muthui, CEO, Tamalaki Business Network Ltd

Ask a packing supervisor how they make sure a 2 kg bag of flour is really 2 kg, and you will often hear the same answer: "We put a little extra."

It sounds safe. It is not a control. A little extra on every pack is product you are giving away all day, and it still does not prove that no pack is short. One worn filler head, one new roll of heavier film or one rushed operator, and you can be overfilling most bags while a few go out light.

This article is for production leads, QA officers, accountants and owners of small and mid-sized food and agro factories in Kenya and East Africa: millers, packers of sugar, rice, pulses and spices, dairies, juice and edible-oil processors, snack makers and feed millers. It gives you an eight-step net quantity check routine you can run every hour on a packing line. It is operations guidance, not legal advice; confirm specifics for your products with your county Weights and Measures office.

What the rules actually say

Kenya's pack-weight rules sit under the Weights and Measures Act (Cap. 513) and the Weights and Measures (Sale and Labelling of Goods) Rules. A few points matter most on the factory floor:

• Net, not gross. Goods sold by weight or measure must be sold by net weight or measure (rule 4), and net weight excludes wrappers and other packing materials (rules 2 and 8(4)). There is a narrow exception for small, individually wrapped items.
• No "approximately". A quantity declaration must not use words or marks implying an approximation, or anything that exaggerates the quantity (rule 8(1)).
• The tolerance test. Under rule 14, a declaration is deemed correct if, when tested, no pack is deficient by more than 10% for a bottle of 250 ml or 250 g or less, or 3% of the marked quantity in any other case, and the average quantity across the sample is equal to or more than the marked quantity.
• How inspectors sample. The Ninth Schedule sets sample sizes (for example, 30 packages for a non-destructive test on a batch of 30 to 500) and requires samples to be picked at random. For packs on or at the end of a packing line, a "batch" is defined as the line's maximum hourly output (rule 2). The Tenth Schedule describes working out tare by weighing ten empty containers and using their average.
• Penalties. Offences under the Rules carry a fine of up to KSh 20,000, imprisonment of up to three years, or both (rule 15). Under section 42 of the Act, it is an offence to have pre-packed goods for sale when their quantity is found to be less than stated.
• Your defence is your records. Section 47 of the Act gives a defence to a person who proves they took all reasonable precautions and exercised all due diligence. Section 48 covers deficiencies that arise after packing, but only where reasonable allowance was made for them when the goods were made up and marked.

The State Department for Trade explains that county Weights and Measures officers verify trade equipment and inspect pre-packed goods, and it lists trader obligations, including keeping a valid stamp of verification on every piece of equipment used for trade and taking any equipment you suspect is inaccurate out of use. The same page notes that Kenya has been a member of the International Organization of Legal Metrology (OIML) since 1982. OIML's international recommendation on prepackages, OIML R 87:2016, uses the same two ideas: the average must be at least the nominal quantity, and individual packs may only fall short within set limits. If you export, ask your buyer which rules apply in their market.

Put together, those rules suggest a simple floor rule: the average must be on or above the label, no pack may fall outside the tolerance, and you must be able to show the checks you ran.

Why "put a little extra" is not a control

Overfill does two things you do not want:

1. It costs money every hour. Product in the bag beyond the label is product you paid for and do not get paid for.
2. It hides variation. If your filler is erratic, a generous target can lift the average while individual packs still fall below the tolerance.

Illustration (not real customer data): a mill packs 2 kg bags of maize flour and aims "a bit high". The scale log shows the average sitting about 30 g above the label. That is 1.5% of every bag given away. On 1,000 bags a shift, it adds up to roughly 30 kg of flour per shift leaving the factory unpaid, and the scale log may still show the occasional light bag.

The goal is a fill target low enough to stop giving product away and high enough that both legal tests pass every hour, backed by records.

The 8-step net quantity checklist

1. Write down the rule for every SKU

For each pack you run, record on one sheet:

• Declared quantity and unit (grams below 1 kg, kilograms from 1 kg upward; millilitres below 1 litre, litres from 1 litre upward, per rule 9).
• Whether the product is sold by weight or by measure of capacity. The Rules list goods in schedules: for example, milk up to 5 litres and edible oils up to 1 litre are sold by measure of capacity, while maize grain and spices may be sold by weight or capacity.
• Whether a standard pack size applies. The Sixth Schedule sets standard quantities for some goods; for maize flour, for example, 1 kg then steps of 1 kg up to 10 kg, then 20 kg, 50 kg and 90 kg.
• The tolerance that applies (10% for bottles of 250 ml or 250 g or less, 3% otherwise) worked out in grams or millilitres. A 500 g pack at 3% means no pack below 485 g.

If the label team and the packing team are working from different sheets, start here. The label control checklist covers the artwork side.

2. Use a scale you can trust, and prove it every shift

• Use a scale with a valid stamp of verification, as the State Department for Trade's trader obligations require for equipment used for trade.
• Check it at the start of every shift with a known test weight close to the pack weight, and record the reading.
• Choose resolution that matches the tolerance. A scale that reads in 10 g steps tells you very little about a 3 g tolerance on a 100 g sachet.
• If a check fails, take the scale out of use and escalate. Do not "adjust and carry on".

Scale checks belong in the same routine as the rest of your preventive maintenance.

3. Get tare right, and re-tare when packaging changes

Most short-weight surprises are tare surprises.

• Weigh at least ten empty packs (bag, bottle, cap, label, liner) and use the average as your tare, the same approach the Tenth Schedule describes.
• Re-do it when the film supplier, bag grade, bottle mould or label stock changes.
• Write the tare and its date on the line sheet so everyone uses the same figure.

A heavier batch of bags with an old tare figure means every "net" weight you record is wrong in the same direction.

4. Set the fill target from data, not habit

• Run the filler at the current setting and weigh a run of packs (30 is a practical start) to see the average and the spread.
• Set the target so that the average stays on or above the label and the lightest packs stay well inside the tolerance.
• The steadier the filler, the closer to the label you can safely run. Reducing variation is what earns you back the overfill.
• Agree the target and who may change it. Record every change with the time and the reason.

5. Sample every hour, at random, from every filling head

Because a line "batch" is an hour's output under the Rules, an hourly check lines up with how an inspector would look at your line.

• Take a set number of packs each hour, at random times, not just the first packs after a break.
• On multi-head fillers, sample every head. One bad head can hide inside a good average.
• Weigh gross, subtract the agreed tare, and record each net weight against the line, SKU, lot and time.

6. Apply both tests, every time

For each hourly sample, record a pass or fail on two questions:

1. Is the average net quantity equal to or more than the declared quantity?
2. Is every pack within the tolerance (no deficiency greater than 3%, or 10% for small bottles)?

Both must pass. A good average with one pack outside the tolerance is a fail. So is a sample where every pack is inside the tolerance but the average is below the label.

7. When a check fails: stop, fix, hold and sort

• Stop the line or the affected head and fix the cause: setting, tare, product flow, worn parts or operator.
• Put every pack filled since the last passing check on hold. Treat it like any other QA hold.
• Check-weigh the held packs one by one; top up, rework or downgrade any that fail.
• Restart only after a fresh sample passes, and record what was done.

Packs that failed must not reach the truck. Your finished goods dispatch checklist should block held lots.

8. Keep the records and track give-away

• File scale checks, tare records, hourly samples, failures and actions by line, SKU and date. If an inspector ever tests your packs, these sheets are how you show the precautions and due diligence that section 47 of the Act talks about.
• Track average overfill by SKU, line and shift. It is a cost line, not a rounding error.
• Feed it into batch costing and yield tracking. Unexplained yield loss is often overfill nobody measured, and it shows up again as a gap at your next stock count reconciliation.
• If your product can lose weight after packing (some dry or baked goods lose moisture), measure it and set your allowance deliberately. Section 48 only helps where reasonable allowance was made.

Illustration (not real customer data): a spice packer samples 100 g sachets hourly. For three days, the afternoon shift's average runs a few grams higher than the morning's. The cause turns out to be a different film roll with a heavier tare that nobody re-measured, so operators were raising the setting by eye. Re-taring fixes the records and the give-away at the same time.

Habits that make pack-weight control stick

• One sheet per SKU. Declared quantity, tolerance, tare and target in one place.
• Check the scale before you check the pack. A wrong scale makes every record wrong.
• Hourly, random, every head. Not just at start-up.
• Two tests, both recorded. Average and individual packs.
• Overfill is a number. Review it weekly with production and finance.

Where an ERP helps

A clipboard and a verified scale can run this routine on one line. It gets hard with many SKUs, several fillers, tare changes and holds to trace.

FactoraERP tracks production batches and lots, stock status including QA holds, non-conformance records, and batch costs and yields. That means a failed pack-weight check can be tied to the batch it affects, the affected output can be held, and the cost of overfill shows up in your yield figures instead of disappearing. See factoraerp.com.

Free download: eTIMS readiness checklist for factories

Pack sizes, stock and invoices increasingly have to tell the same story for Kenyan manufacturers. If you want a free eTIMS readiness checklist for factories, sign up for the FactoraERP Insights newsletter at https://isaacmuthui.substack.com. You will get practical operations notes for food and agro manufacturers. No spam; unsubscribe anytime.

FAQ

What is a net quantity check?
It is a routine test of packed goods to confirm the quantity of product, excluding packaging, matches what the label declares. In practice that means weighing a sample, subtracting an agreed tare, and checking both the average and every individual pack against the rules.

What tolerance applies to pre-packed goods in Kenya?
Under rule 14 of the Weights and Measures (Sale and Labelling of Goods) Rules, a declaration is deemed correct if no tested pack is deficient by more than 10% for bottles of 250 ml or 250 g or less, or 3% of the marked quantity otherwise, and the sample average is at least the marked quantity. Confirm how it applies to your products with your county Weights and Measures office.

Is overfilling safer than underfilling?
It lowers the risk of short packs only if your filler is steady, and it costs product on every pack. Reducing variation and setting a data-based target is usually safer and cheaper than a generous margin.

How often should we check pack weights?
At least every hour on each line, at random times, from every filling head. The Rules define a line batch as the line's maximum hourly output, so an hourly check matches how a batch is defined.

Does the packaging count in the net weight?
No. Net weight excludes wrappers and other packing materials, with a narrow exception in rule 8(4) for small, individually wrapped items.

Can a small factory do this without software?
Yes. A verified scale, a test weight, a tare record, an hourly sample sheet with the two tests and a hold rule cover the basics. Software helps when SKUs, lines and lots multiply.

---

Isaac Muthui, CEO, Tamalaki Business Network Ltd

---

Originally published on the FactoraERP blog: https://factoraerp.com/en/blog/net-quantity-pack-weight-checks-food-factories-kenya