My New Fragrance Batch Smells Slightly Different From My Retained Sample — How Do I Determine Whether the Variation Is Commercially Acceptable?

"Slightly different" needs a test, not an opinion

The new batch smells a little different from your retained sample. Maybe brighter, maybe a touch flatter, maybe you are not sure. The trouble is that the person who knows the fragrance best — you — is also the person most likely to imagine a difference, or to talk yourself out of a real one. You know which bottle is new. You have been sniffing both for ten minutes. Your nose has adapted. And you would quite like the answer to be "fine" because the kilogram is already paid for.

"Commercially acceptable" breaks into three separate questions, and each needs its own check:

Is there a real, detectable difference? That is a sensory test done blind. Would a customer notice it in the candle? That is a cold-throw and hot-throw comparison, not a bottle sniff. Is the difference a normal variation or a sign of a problem? That is about the character of the change — and a conversation with the supplier.

Answer them in that order. If the first answer is no, you can stop.

Step one: a blind triangle test on blotter strips

The triangle test is a standard sensory method, and it suits small businesses because it needs no equipment beyond strips, labels and a helper. Each assessor is given three strips, two dipped in one oil and one in the other, coded so they cannot tell which is which, and asked only one question: which one is the odd one out? Guessing gets the right answer one time in three, which is what makes the result meaningful.

Choose assessors with care. The person who ordered the oil, or who has been handling it all week, is the least useful nose in the business. Family members, a packer who rarely works with that scent, or a friend who buys candles make better assessors precisely because they carry no expectations. Nobody should be wearing perfume or have just used scented hand cream, and the room should be free of burning candles and open fragrance bottles. Run the test before lunch rather than after, when cooking smells linger.

A practical design for a small team is three assessors doing two rounds each, giving six answers. Here is what chance alone would produce:

The working for the 5-or-more row: exactly five correct is 6 × (1/3)⁵ × (2/3) = 12/729, exactly six is 1/729, so together 13/729. The cut-off lines are a suggested starting point, not a published standard; formal sensory standards exist if you later need a more rigorous plan. If the panel cannot reliably pick the odd strip, accept the lot on odour and record the result.

Step two: does the difference survive into the candle?

A difference on a blotter strip is not automatically a difference a customer will notice. Wax holds some materials back and releases others, and the burning candle adds heat, air movement and a room full of other smells. So if the triangle test says the difference is real, take it to the product.

Make two candles on the same day from the same wax bag, same jar, same wick, same load and same pour temperature — one with the retained oil, one with the new batch. Cure them together for at least the stated minimum on the product page for your wax type (soy 48–72 hours on most CSI oil pages), and then run the same blind triangle idea on cold throw using three lidded jars — two of one oil, one of the other — with lids lifted one at a time. For hot throw, burn each in turn in the same closed room for the same time, with a fresh-air break between, and have assessors describe what they smell at the door without knowing which candle is lit. Never leave a test burn unattended.

Step three: what kind of difference is it?

Once you know a difference is detectable, its character tells you what to do with it. Use the assessors' free-text notes.

Then add your own commercial context. A signature scent sold to repeat customers or to a retailer who reorders by name has less room for drift than a seasonal one-off. A fragrance that sits in a gift set beside candles from the previous batch has less room still, because customers compare them side by side. That context changes where you draw the line — not the test.

Worked example: a jasmine signature scent

You sell a jasmine candle as your signature line and make around 120 of them a month in 200 g jars at 8%. That is 16 g of oil per candle, so a 1 kg bottle of Jasmine at ₹3,923.00 puts 16 ÷ 1,000 × ₹3,923.00 = ₹62.77 of oil in each candle, and 120 candles a month use 1,920 g.

The new kilogram smells "greener" to you. You run the strip triangle with two helpers and yourself, two rounds each: four correct out of six. Borderline, so you repeat the next morning with fresh noses: three out of six. Across both sessions, seven correct out of twelve. The difference, if it exists, is small and hard to find even blind.

You still make the paired candles, because this is a signature line. After the page's stated minimum cure and a few more days, the cold-throw triangle scores two out of six, and nobody can tell the lit candles apart at the door. Decision: accept. You record both sessions, keep 20–30 ml of the new batch as the next retained reference, and write the invoice number and date on the bottle, since CSI product pages do not publish lot numbers.

Had the candle test scored five or six, the answer would have been different: quarantine, a message to the supplier with your notes, and no blending of the new oil into old stock to soften the change. A batch that fails your comparison goes back in quarantine whoever supplied it — including us. For CSI, send the question on WhatsApp +91 7397976926 and ask whether anything in the formula or its materials has changed; IFRA and MSDS documents can be requested there too.

Where this page fits

CSI already has guides that cover parts of this question. This page covers the goods-in check for the question above; these go deeper on the rest:

Three things to buy for this job

Frequently asked questions