What the record must carry
A batch record is the written account of one made quantity of product, and it has to be complete enough for someone else to repeat the batch. It starts with the date, the product name and a unique batch code, then the batch size in grams. Every material is listed with its INCI name, the supplier, and the lot number if one is printed. The target grams and the actual grams sit side by side, so a difference is visible without arithmetic. The phase, the temperature at combining, the time mixed, the final pH and its temperature, the maker's name and the number and size of jars filled are all part of the same page.
The record also carries the things that went wrong. A dropped spatula, a second pH adjustment, a longer heat than planned, a substitution because a material ran out: each is written at the time, with the time. A record with no deviations is not necessarily a good record; it may be a record that was written from memory at the end. The point of the deviation column is not blame. It is that the next maker can see what changed and decide whether the change mattered. A batch that was made twice, once with a deviation and once without, may not behave the same in a stability trial.
Attach the calculation sheet and the pH readings to the record rather than copying only the final numbers. The calculation shows how each gram was derived, and the pH readings show whether the number was stable or drifting. If a premix was used, record its strength, its lot and the grams taken. If water was weighed, record the grade and the container lot. These details look excessive until a client reports a reaction or a jar changes colour, and then they are the only way to find out which material, which lot and which day are involved. A copied pH value without the temperature is half a measurement, because pH changes with temperature and a reading at 25 C cannot be compared with one at 40 C.
Sign and date the record, and keep it with the retention sample from the same batch. The retention sample is the physical evidence; the record is the description. One without the other is much less useful. Keep both for at least the shelf life claimed plus a margin, and longer if the product is still on a client's shelf. Store the record where it can be found by batch code, not in a pile ordered by date, because the first question in an investigation is always which batch, not which day. File by code, with the retention sample beside the record, and keep the file dry and away from sunlight so the ink and the sample both survive.
- 1. Enter the date, product name and a unique batch code.
- 2. Write the batch size in grams and the phase list.
- 3. List every material by INCI name with supplier and lot.
- 4. Record target grams and actual grams on the same line.
- 5. Record temperatures, mixing times, final pH and the pH temperature.
- 6. Record the jar count, the pack type and the fill mass.
- 7. Note every deviation with the time it happened.
- 8. Sign, date, and attach the calculation sheet and pH readings.
Bench example
A 200 g test batch of an oil-in-water lotion is recorded on one page. The batch code is 20261025-01, the product is 'O/W lotion, test', and the batch size is 200.00 g. Phase A lists aqua 140.00 g, glycerin 6.00 g and a chelating agent 0.10 g, with the supplier and lot for each. Phase B lists the oil, the emulsifier and the thickener, with target and actual grams. The phases are combined at 72 degrees Celsius, recorded at 08:40, and mixed until 08:55. The cool-down is recorded at 38 degrees, and the preservative is added at that point, 1.00 g actual against a 1.00 g target. The final pH reads 5.21 at 25 degrees Celsius. Four 50 g jars are filled, and the note says 'no split at 24 h; pH 5.18 at 24 h'. The maker signs, and the calculation sheet is stapled behind the page. A second maker can read the page and repeat the batch without asking a question, which is the test the record is meant to pass. The page is dated at the start and signed at the end, and the difference between those two times is the batch.
Caution: A record written after the jars are filled is a reconstruction, not a record. Write each value at the bench, while the beaker is still on the scale.
Write during the batch, not after
The record is written during the batch, in ink, on the sheet that belongs to that batch. Values are entered as they are read, not collected in memory and transferred later. Memory is good at the shape of an event and poor at the third decimal place, and a pH recalled an hour later is a guess. Keep the record at the edge of the bench, out of splash range but within reach, and put the pen with it. A record that lives in a drawer is a record that gets written at the end, and a record written at the end always looks cleaner than the batch was.
Every batch gets a unique code. Date plus a sequence number is enough: 20261025-01, 20261025-02. Do not reuse a code because the product is the same, and do not code by product name alone, because there will be several batches with the same name. The code goes on the record, on every jar, on the retention sample and on any lab sample. If a code is written incorrectly, do not overwrite it; strike it once, write the correct code beside it, and initial the change. An overwritten number is unreadable and looks like a concealment even when it is only a slip.
Record the actual mass even when it is wrong. If a target of 10.00 g became 10.02 g, write both and note the difference. If the difference is outside the limit set for that material, write what was decided: continue, adjust another line, keep as a lab sample, or discard. The decision and its reason are part of the record. A record that shows only the target is not a record of the batch; it is a record of the plan. The plan and the batch diverge at the bench, and the divergence is the information that matters later. The deviation note should be short enough to write in the moment and specific enough to be understood a year later, with the time and the decision.
Scale changes need the same discipline. If a formula is made at half size or double size, the record still carries the percent and the grams for that batch. Do not copy an old record and change only the batch size, because the other numbers will not have been recalculated. Write the new gram column from the percentages and check that it sums to the new batch size. A copied record with a changed date is the most common way a small error becomes a repeated error, and it is entirely avoidable with a fresh calculation sheet. Write the new batch size at the top and recalculate every line, even the water, because a copied water figure is the most common error in a scale change.
- 1. Keep the record and a pen at the bench before the first weighing.
- 2. Write each value as it is read, not from memory afterwards.
- 3. Assign a unique batch code before the batch starts.
- 4. Strike through an error once, write the correction, and initial it.
- 5. Record the actual mass beside every target.
- 6. Write each deviation, the time, and the decision taken.
- 7. For a scale change, recalculate every line from the percentages.
- 8. Sign and date at the end, and file the record with the retention sample.
Bench example
A maker starts a 500 g body cream at 09:00. The record is on a clipboard at the end of the bench, and the batch code is 20261025-02. At 09:20 the thickener is weighed at 4.02 g against a target of 4.00 g, and both numbers are written. At 09:35 the water bath overshoots to 78 degrees, and the note says 'held 3 min at 78, cooled to 73 before combine'. At 10:05 the pH reads 5.60 against a target of 5.30, and a citric acid solution is added in three steps, with each mass and each pH reading on the sheet. At 10:30 the batch is filled into ten 50 g jars, and the maker writes the jar count and the fill mass. At the end the maker signs and attaches the calculation sheet. Two weeks later a jar shows a slight separation, and the record shows the heat overshoot and the three pH adjustments, which is exactly the sequence the investigation needs to see. The record is filed with the retention sample the same afternoon, and the maker initials the file entry so the chain from jar to page is complete.
Caution: Do not rewrite a record to make it look tidy. A crossed-out number with an initial is honest; a clean copy made later is not a record of what happened.
A weak record and how to fix it
A weak record has recognisable features. It says 'water, oils, emulsifier, preservative' with no INCI names and no grams. It has a target list and no actual weights. It says 'pH ok' with no number and no temperature. It reuses a batch code, so two jars cannot be told apart. It says 'made more' or 'added a little' without a mass. It has no name, no date and no signature, and it was clearly written after the jars were filled. Each of these is fixable, and the fix is the same: write the missing value at the bench, for the next batch, and treat the existing batch as less well documented than it is.
The test of a record is simple. Give it to another maker, with the same raw materials, and ask them to repeat the batch. If they can weigh the same masses, reach the same temperature, and check the same pH without asking a question, the record is finished. If they have to ask what 'a little' means, or which preservative, or whether the water was distilled, the record is a draft. Run this test on your own record before a client or an inspector does, and fix the gaps while the batch is still fresh in memory. A record that passes this test is worth more than a longer record that nobody can follow, because the test measures whether the information is usable.
Retention matters as much as completeness. Keep a sample of every batch in the same pack, with the same lid, at the same storage condition, for at least the shelf life claimed plus a margin. Write the make date and the discard date on the retention jar. If the product is a water-containing formula that has not passed a challenge test, the retention period is short, often 7 to 14 days, and the record should say so. A retention sample that has been opened, used, or stored on a sunny shelf is not a retention sample. Write the storage condition on the record as well, because a sample kept in a warm cupboard tells a different story from one kept cool and dark.
When a complaint arrives, the record is how the batch is found. The batch code on the jar leads to the record, the record leads to the raw material lots, and the lots lead to the supplier. Without the code, the investigation starts from nothing. With the code, it starts from a page. This is why the code is written on the jar as well as the record, and why the record is filed by code. The paperwork is not for the inspector; it is for the day a client says the cream changed, and the answer has to be more precise than 'it was fine when it left'.
- 1. Read the last record and list every missing value.
- 2. Replace vague words with masses, temperatures and times.
- 3. Give every batch a unique code and write it on every jar.
- 4. Add the INCI names, supplier names and lot numbers.
- 5. Add the actual masses and the pH with its temperature.
- 6. Attach the calculation sheet and the pH readings.
- 7. Keep a retention sample in the same pack with a discard date.
- 8. Test the record by asking another maker to repeat the batch from it.
Bench example
A salon's folder holds twelve records for the same hand cream. Six say 'pH ok', four have no batch code, and two list 'preservative' without a name or a mass. A client reports that a jar from July developed a sour smell. The jar carries no code, so the record cannot be matched; the folder has three July batches, and all three records are incomplete. The salon cannot tell whether the fault was the preservative lot, the water, or the jar. The next batch starts a new sheet: code 20261025-03, every material by INCI name, supplier and lot, actual masses, pH 5.44 at 24 degrees Celsius after 24 hours, ten jars, retention sample with a discard date of 20261108. The record is tested by asking a second maker to repeat a 100 g trial from it; the trial matches within 0.02 g on every line. The old records are kept, not discarded, but the new sheet is the standard from that day. The old records stay in the file with a note that they are historical, so nobody mistakes them for a template to copy.
Caution: A record with no actual masses cannot answer a complaint. If the jar and the record cannot be matched by code, the batch is untraceable however good the product was.
This is a Nandu Nai textbook. It is a private brand book, not a government licence, and it does not give the right to give medical treatment.