My Soy Wax Supplier Changed the Product Slightly Without Me Changing My Candle Formula — How Can I Determine Whether the Raw Material Caused the New Problem?

Change one thing at a time and let the candles tell you

Your formula has not changed. Same recipe card, same load, same wick, same jar. But a problem has appeared that was not there before — rougher tops, wet spots on the glass, a weaker throw, a candle that tunnels where it used to burn clean — and it seems to have arrived with the latest wax. The natural conclusion is that the supplier changed the product. It might be true. But a candle has many inputs, and several of them change without anyone touching the recipe card: the oil lot, the wick pack, the jar batch, the weather, the size of the melt, the person pouring.

The way to settle it is the same way any manufacturer settles it: hold every input still except one, and see whether the problem follows that input. For wax, that means pouring candles from a retained sample of the old delivery beside candles from the new one, with one oil lot, one wick pack, one jar batch, on the same day, in the same way. If the new-wax candles show the problem and the old-wax candles do not, the wax delivery is the most likely cause. If both show it, the wax is probably not the cause. If neither shows it, the cause is in production, not in the materials.

This page builds that test step by step: how to define the problem so you can score it, what else might have changed, the nine-candle swap design that fits a normal retained sample, how to read the result, and what to do with it. One thing it cannot rely on is the supplier telling you about a change in advance. CSI publishes no change notices, lot numbers or certificates of analysis for its waxes; you can and should ask us on WhatsApp whether anything changed, but the evidence that settles it is your own.

Step one: define the problem so you can score it

A vague problem cannot be tested. "The candles aren't right" will produce a vague answer. Before pouring anything, write the problem as something you could score on a test candle at a stated checkpoint: what, where, when, how often, and how bad.

The "which SKUs" line is often the most revealing. A problem that appears across every fragrance points towards something shared — the wax, the wick, the jar or the process. A problem confined to one oil points towards that oil or its pairing with the wax. Why a wax can suit one oil and not another covers the oil-specific case.

Step two: list everything else that changed

Before you blame the wax, check the timeline. Write down every input and condition, and when each last changed, against the date the problem first appeared. Your batch sheets and delivery records are the source — the delivery-record post sets out what they should hold, and the lot-label post shows how lot codes make this table possible.

Anything that changed at the same time as the wax is a rival explanation. Anything the swap test cannot vary — the melter, the season, the team — is checked by the result itself: if none of the bench cells reproduces the fault, the cause sits in one of those rows. That is a useful answer too, because it stops you from changing a supplier to fix a process problem.

Step three: the nine-candle swap test

The design fits the retained sample most makers actually have. A 580 g-or-larger retained sample of the old delivery makes three 200 g candles at 8% (184 g of wax each, with a 5% allowance). That gives cell A. The new delivery, which you have plenty of, gives cells B and C. Cell C uses your previous oil lot — the end of the last bottle, which is why it is worth keeping 50.4 g or more of every oil lot until the next one is proven.

Reading the result: what each pattern means

Read the pattern across cells, not single candles. Three candles per cell is small; one odd candle is a flag, a consistent cell is a finding. If the result is borderline, repeat the test rather than stretching the conclusion.

The fourth and fifth rows are the ones makers find hardest to accept, because they mean the supplier did not cause the problem. They are also the most valuable: they stop an expensive and pointless supplier change. If the fault only appears in production and never on the bench, validating at sample, pilot and production scale explains why scale matters, and testing under your actual workshop temperature covers conditions.

The priced swap-test kit

The new-delivery wax for cells B and C (1,159.2 g with allowance) comes from your stock; at ₹507.40 a kilo it is ₹588.18 of wax. The old-delivery wax comes from your retained sample. The current oil lot for cells A and B needs 100.8 g with allowance; cell C uses 50.4 g from your retained bottle of the previous lot.

If you are buying the current oil specifically for the test, note that Cinnamon Vanilla at 100 g + 15 g (₹590.00) is the cheapest way to 100.8 g. If you have an unopened bag of the old delivery and want cell D for a full 2 × 2, add 579.6 g of old wax and another 50.4 g of the previous oil lot, plus three more jars and wicks.

Against those, the swap test is a few hundred rupees of oil, jars and wicks and about two weeks of cure and burn. Material cost here is ₹242.94 per candle and excludes labour, packaging, freight and GST.

Materials in the worked example

The decision rule after the swap test

For the example maker, a cleared wax means carrying on with Luxury Soy Wax CSI 464 in 10 kg bags (₹5,074.00 each) and fixing the oil or the process. An implicated wax means holding the lot, messaging the supplier with evidence, and running the next delivery through incoming QC before it touches production — and, if the supplier cannot explain the change, starting the backup approval in the backup-supplier post.

Making the next investigation shorter

This investigation was possible only because some records and samples existed. Every one of them is cheap. The next time a problem appears, the same habits make the swap test a one-day decision instead of a week of searching.

Frequently asked questions