I Approved a Fragrance Sample Months Ago and the Supplier’s Current Batch Smells Different — Should I Treat the New Material as a Fresh Qualification?

First question: which one changed, the oil or the reference?

When a current batch smells different from an oil you approved months ago, the natural assumption is that the new batch has changed. Sometimes it has. But three other things have also had months to drift: the retained sample, if you kept one; your memory of how the approved candles smelled; and your recipe, which may now use a different wax bag, a new box of wicks or a slightly different pour temperature.

A retained fragrance sample that has been opened repeatedly, stored in a half-empty bottle or kept somewhere warm can lose top notes and darken. Compared with that, a fresh batch may seem brighter or sharper even when it is closer to the original than the reference is. Memory is less reliable still. It is very easy to remember an approval candle as richer than it was.

So check the reference before judging the new oil. Note how the retained sample was stored and how often it was opened. If you kept a finished candle from the approval, bring it out; a candle stored away from light and heat is often a steadier reference than an opened oil. Where the two references disagree with each other, you have learned something about your retention method before you have learned anything about the supplier.

Ask the supplier, then test regardless

The supplier can tell you something you cannot learn yourself: whether the formulation of that oil has been changed since your approval date. For CSI oils, ask on WhatsApp at +91 7397976926, giving the product name and when you approved it, and write their answer into your record word for word. CSI product pages publish some details, such as flashpoints and, for some oils, vanillin content, but they do not publish lot numbers, certificates of analysis or change notices, so a direct question is the only way to ask.

Treat the answer as information that sets your starting tier, not as a test result. If the supplier confirms a change, go straight to a full re-qualification. If they say nothing has changed, still run your comparison, because raw materials inside any blend can vary naturally, and because your own recipe may be the thing that moved. A supplier's assurance and your own evidence answer different questions.

Four tiers of response

The table below is a suggested starting point for sizing the work. It is not a standard and not CSI policy; adjust it to your products and the risk you are prepared to carry.

For the blind strip step, code the strips so you cannot tell which is which, and use an odd-one-out arrangement where two strips come from one oil and one from the other. If you cannot pick the odd strip reliably, the difference is unlikely to matter in a candle. If you can, move to the confirmation candle.

A note on customer-facing changes

Tier 2 and Tier 3 results raise a question beyond testing: whether your product description still fits. If a candle you sell as warm and resinous now opens brighter and spicier, the honest move is to update the description, or to keep the batch out of that product line, rather than hope no one notices.

Worked example: Royal Oud, six months on

A maker approved Royal Oud six months ago and now receives a 500 g pack at ₹3,300.00, which is ₹3,300.00 ÷ 500 = ₹6.60 per gram. The new oil seems smokier on opening. The retained sample from the approval was opened several times and is in a bottle a third full, so the maker also brings out a finished approval candle kept in a dark cupboard.

On coded strips, the assessor picks the new oil reliably against the retained oil. Held beside the cold approval candle, though, the new oil's character seems close, while the retained oil smells flatter than both. That pattern points to an aged reference rather than a changed batch, so the candle becomes the reference for the next step. The supplier, asked on WhatsApp, says the formulation is unchanged. That sets a Tier 1 starting point: a confirmation candle.

Two confirmation candles at 8% in the approved 200 g jar use 2 × 16 g = 32 g of oil, costing 32 × ₹6.60 = ₹211.20. The product page recommends 72–96 hours of cure because oud takes longer to express, so the comparison with the approval candle cannot happen for at least three to four days. If the confirmation candle matches, the oil is released and the cost of the question is ₹211.20 of oil plus the wait.

If it does not match, the maker moves to a partial re-qualification. A small grid of three loads within the page's advice, 6%, 8% and 10%, with two wick sizes, gives six 200 g candles. The oil for those is 2 × (12 + 16 + 20) = 96 g, costing 96 × ₹6.60 = ₹633.60. The wax is 2 × (188 + 184 + 180) = 1,104 g; in CSI 464 at ₹507.40 per kg that is 1.104 × ₹507.40 = ₹560.17. Oil and wax together come to ₹633.60 + ₹560.17 = ₹1,193.77, before jars and wicks, and the full 72–96 hour cure applies again.

The value of the tiers is that the maker pays for Tier 2 only when Tier 1 has failed, and for Tier 3 only when the evidence shows the oil is genuinely different. Skipping straight to Tier 3 on a first sniff would have spent a week and several hundred rupees answering a question two candles could settle.

Decide, record, and protect the next comparison

Whatever the tier, a batch that fails a fair comparison stays out of production whoever supplied it, including us. And if the new batch passes, do not hold it back out of habit; releasing good material is as much a part of QC as rejecting bad material.

One practical point: record which tier you reached each time the oil changes batch. Over a year, that history tells you whether this oil is stable enough for a light goods-in routine or whether it needs a confirmation candle on every delivery. That is a better basis for your QC effort than a single memorable scare.

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