My Fragrance Supplier Sent a New Batch of My Bestselling Scent — What Should I Compare Against the Previous Batch Before Making Hundreds of Candles?

Why you need a retained sample

Most batch problems are never caught for a simple reason: by the time lot B arrives, lot A is gone. The maker opens the new bottle, sniffs it, decides it smells like Lavender, and pours. Smell memory is not reliable enough for that job. You have been living with the old lot for weeks, your nose adapts to scents you work with every day, and a difference that a customer notices in a warm living room can be invisible in a workshop at the bottle neck. The only fair comparison is side by side, on the same day, in the same wax.

That means keeping a retained sample of every lot you approve. Retained samples are standard practice in any manufacturing that depends on a raw material behaving the same way twice. For a candle maker the rule is simple: when a lot arrives, decant 50 g into a small, clean, full glass bottle, seal it, label it with the product name, the lot or invoice reference, the date received and the candle batches it went into, and store it cool and dark. At Lavender's 1 kg rate that sample costs ₹261.05. It is the cheapest insurance in your workshop.

A 5 g vial is enough to remind you of the smell. It is not enough to pour anything, and the whole point of this page is that the bottle is the least reliable place to judge a lot. Two oils that smell the same on a strip can behave differently in wax, and two that smell slightly different on a strip can give identical candles. Why bottle strength does not predict candle performance explains the mechanism. The retained sample has to be big enough to pour.

The six gates, cheapest first

The comparison is ordered by cost and by how close each check is to what your customer experiences. Gates 1 to 3 cost almost nothing and take a morning. They catch the obvious problems — a mislabelled bottle, a visibly different oil, a scent that has drifted. Gates 4 to 6 need wax and time, and they are the ones that decide whether lot B goes into 300 jars. Hold every other variable fixed, exactly as you would in a controlled 20-oil comparison: one wax lot, one wick size from one pack, one jar model, the same load, the same pour day and the same test room.

The table in the walkthrough deliberately does not include a density or refractive-index check. Those are real quality-control tools used by fragrance manufacturers, but the specifications for individual CSI oils are not published in our data, so a candle maker has nothing to compare a reading against. The retained sample does that job instead: it is the specification, in a bottle. If your business grows to the point where you want instrument checks, ask CSI on WhatsApp what lot information can be shared for the oils you buy.

The cost of checking

Here is the full materials bill for the six-gate comparison on a Lavender bestseller, with both lots costed at the 1 kg rate and wax from Luxury Soy Wax Chunks at ₹599.00 a kilo. The tins are Golden Aluminium Tins with a published 85 g fill weight, at ₹360.00 for 12. Cold-throw tins need no wick.

The table prices four candles — two per lot — because a single candle per lot cannot tell you whether a difference is the oil or the candle. If you are confident in your process, one candle per lot halves the jar, wick and candle-wax cost. Either way the number to hold it against is the run. 300 Lavender candles at 8% use 4,800 g of oil and cost ₹270.45 each in oil, wax, wick and jar, which is ₹81,135.60 of materials before labour, packaging, freight and GST. The test is 1.89% of that. A single returned order from a customer who noticed the difference would cost more.

There is a second calculation that most makers miss. If you buy exactly five 1 kg packs for a 300-candle run, keep 50 g as the retained sample and use 45.6 g for the test, you have 4,904.4 g left. That is only 2.17% over the 4,800 g the candles need — under the 3–5% extra we suggest planning for pour loss and test burns (a working assumption). At a 3% allowance that oil covers 297 candles, not 300. Budget the retained sample and the test into the order, or the reference you keep for next time quietly shortens this run. I Need Enough Fragrance Oil for 500 × 200 g Candles at 8% Load — How Much Should I Order After Including Formulation Trials and Production Wastage? and I'm Producing 1,000 × 200 g Candles at 10% Fragrance Load — How Many Kilograms of Fragrance Oil Should I Purchase so I Don't Run Short During Production? work through the order-size maths in full.

Release, hold or ask

At the end of the test you have a filled comparison sheet and one of three outcomes. The rules below are working decision rules, not industry standards. Write your own version down before you test, so a result does not tempt you into a decision you would not have made in advance.

The last row is where makers go wrong. If lot B burns weaker, the instinct is to add oil — 8% becomes 9% — and carry on. That turns a supplier question into a formulation change, raises your cost per candle and may overload the wick. At Lavender's 1 kg rate, one extra percentage point in a 200 g candle is 2 g, or ₹10.44 a candle — ₹3,132.00 across 300. Before changing the formula, check that the difference is the oil and not your process. My Small Test Batch Has Excellent Hot Throw but the 20 Kg Production Batch Smells Weaker — What Production Differences Should I Investigate Before Blaming the Fragrance Oil? gives the production-side checklist, and My New Fragrance Oil Smells Almost Identical to the Old One but Has a Different Composition — Can I Use My Existing Wick and Fragrance Load Without Retesting? covers what to do when a supplier confirms a reformulation.

Buying so lots stay manageable

Every new lot costs you a test. Buying your bestseller in larger quantities reduces how often you run one. On CSI's listed catalogue, larger packs mostly buy convenience rather than a lower rate: of 91 oils listing both 100 g and 1 kg, 52 are within 1% per gram. Lavender is typical — ₹5.40 a gram at 100 g and ₹5.22 at 1 kg, a saving of 3.3%. The better reason to buy the kilo for a bestseller is that it lets you run one acceptance test per kilo instead of one per 100 g bottle. I'm Launching 10 Candle Fragrances — How Do I Decide Which Scents Deserve 1 Kg Fragrance Oil Purchases and Which Should Remain in Smaller Trial Sizes? covers which scents deserve kilo purchases in the first place.

Two habits make lot control almost automatic. First, record the lot on the batch sheet for every production pour — the oil lot or invoice reference, the wax bag, the wick pack. When a customer says a candle smells different, you can trace which lot it came from. How Do I Document a Custom Candle Fragrance Formula so My Production Team Can Reproduce the Same Scent and Strength Every Time? shows a batch record that includes these fields. Second, do not mix lots in one pour unless both have passed. Topping up the last 300 g of lot A with lot B produces candles that match neither retained sample.

Frequently asked questions