How Do I Qualify a Second Soy Wax Supplier Without Accidentally Changing the Performance of My Bestselling Candles?

Approve the second wax per bestseller, against that bestseller's own control

The fear behind this question is specific. Your bestselling candles have a performance your customers recognise: the way the pool opens, how the room smells by the second hour, how the top looks when the box is opened. You want a second wax on the shelf so a stock-out cannot stop production, but you do not want the act of adding it to quietly change the candles that pay the bills. That fear is correct, and the way to respect it is to change what you mean by "approved".

The common mistake is to qualify the second wax in general — one test candle, one favourite oil, a nice result — and then declare it interchangeable across the range. That is how a bestseller drifts. The fix is SKU-level approval. Each bestseller is a fixed recipe: wax, oil, load, wick, jar, fill weight and process. The second wax enters that recipe as the only change, the bestseller's current version is the control candle, and the second wax is approved for that SKU only when it is no worse than the control on every line of the scoring sheet. A SKU that passes becomes dual-sourced; a SKU that does not stays single-sourced on the incumbent wax until it has its own test.

That sounds slow, and for a whole catalogue it would be. But you do not need the whole catalogue. You need the few candles whose stock-out would hurt — your bestsellers — and you need them first. Everything else can wait for its turn, or simply stay on the incumbent wax. Whether to approve a backup supplier at all is settled in the previous post; this one assumes you have decided yes and need to do it without damaging what already sells.

Freeze the bestseller recipe before you test anything

Before a single candle is poured, write down each bestseller exactly as it is made today. This is the reference the second wax has to match, and it has to be complete, because anything you leave vague becomes a second variable. In the worked example the business has three bestsellers, each a 200 g candle in a White Matte Glass Jar with an Eco Candle Wicks Thin C1 wick at 8% fragrance load — 16 g of oil and 184 g of wax per candle on CSI's finished-weight convention. The oils are British Rose, Sandalwood and Dark Vanilla. The incumbent wax comes from the reader's current supplier; the two candidates for the second source are CSI's Luxury Soy Wax Chunks and Luxury Soy Wax CSI 464.

One line on that card is allowed to move: process temperatures. A candidate wax is entitled to its own maker's instructions, and forcing it through the incumbent's temperatures would be an unfair test. For the two CSI candidates, CSI's product page states a maximum fragrance load of 13%, fragrance addition at 80–90°C, pour at 75–80°C in its spec table (its step list also shows 70–80°C), and a cure of at least 48 hours, ideally 1–2 weeks; no melt point is published. For CSI 464, CSI's product page states fragrance load up to 10%, a melt/pour working range of 80–90°C in its spec table, a shelf life of 60 months in cool, dry storage, and describes it as single-pour with glass adhesion and low frosting — the maker's description, not CSI test data. Treat every adjective on either page as a claim to test, not a result. How to verify a supplier's "direct replacement" claim goes through that distinction line by line.

The bench: every bestseller in three waxes, side by side

For each bestseller you pour three variants on the same day: the incumbent wax (control) and each candidate. Each variant gets 3 candles — two for burning, one for the shelf — which is a labelled working practice, not a standard. That is 9 candles per SKU and 27 in all. If you only want to qualify one candidate, drop a column and the bench falls to 18 candles. Two burn candles per variant is fewer than a full single-change study would use; it is acceptable here because the control is poured in the same session and the pilot stage below adds three more burns per SKU before anything is released.

Equivalence, not improvement: the scoring sheet

Most wax comparisons ask which wax is better. This one asks something narrower and stricter: is the candidate indistinguishable from the bestseller where customers would notice, and no worse where safety is concerned? A candidate that throws noticeably stronger is not automatically a pass. Your customers bought a particular candle; a stronger one is a different product, and a stronger hot throw that arrives with a deeper pool or a hotter jar is a safety flag, not a feature. Set the tolerance for every line before the first burn.

Two scoring habits make the sheet trustworthy. First, the blind throw question is a triangle test in spirit: give the scorer the bestseller and the candidate unlabelled and ask whether they are the same. If they cannot tell reliably, that is the best outcome for a second source. Second, write every tolerance in the "rule" column before the first match is struck. A tolerance written after you see the numbers is a rationalisation. Whether to use the same oil and wick or optimise each wax explains why this equivalence bench deliberately does not optimise the candidate: optimising it would change the bestseller.

The priced bench kit from CSI

Grams follow the CSI convention: oil = candle weight × load. Each candidate wax needs 9 candles × 184 g = 1,656 g, or 1,738.8 g with a 5% pour-loss allowance (a working assumption). Each oil serves three variants of its own SKU — 9 candles × 16 g = 144 g, 151.2 g with the allowance. The control candles use incumbent wax you already hold, so it is not priced. Wax prices are CSI live pricing, re-checked 1 October 2026; oils CSI live pricing, pulled 24 September 2026; wicks & jars verified 10 September 2026.

Look at the two wax lines. 1,738.8 g of Chunks costs ₹1,196.00 as 4 × 500 g, because the 500 g bag is ₹299.00 — ₹598.00 a kilo, a rupee under the 1 kg bag. CSI 464 costs ₹1,014.80 as 2 × 1 kg: there the 1 kg bag at ₹507.40 beats two 500 g bags at ₹520.00. These are small differences, but they are typical of the listed catalogue — bigger is not automatically cheaper per kilo, and the only soy wax on CSI's list with a real bulk break is Natural Soy Pearl, ₹450.00 a kilo at 1 kg falling to ₹424.80 from 5 kg.

From bench to pilot: proving the second wax at your batch size

A SKU that passes the bench gets one pilot batch in the candidate wax at your normal batch size. In the example that is 24 candles per SKU (a labelled example — use your own batch size). Pull three candles from the pilot — first, middle and last poured — and burn them against three control candles from your current production. Scale changes how long wax sits hot, how evenly oil is stirred through a larger melt, and how the last jars cool, so the pilot tests the process with the new wax, not just the wax. How many candles to make from a new lot before release sets out the logic of the pulled-sample count.

For the three-SKU example on CSI 464, the pilot needs 3 × 24 × 184 g = 13,248 g of wax, 13,910.4 g with the allowance, which rounds to 1 × 10 kg + 4 × 1 kg at ₹7,103.60. Each oil needs 403.2 g per SKU. The pilot candles are saleable if they pass, so this is mostly production you would have made anyway — the real cost is the three pulled candles per SKU and the time.

Running two waxes without blending them by accident

Approval is half the job. The other half is making sure the second wax only ever goes where it was approved. In a small workshop the risk is not a dramatic failure; it is a half-empty bag of Supplier A topped up with Supplier B on a busy day, or a dual-approved British Rose batch followed by a Dark Vanilla batch from the same melter. Write these rules on the wall above the melter.

What qualification costs against what an unqualified stock-out costs

Material cost per candle, at the 5 kg wax rate and the 1 kg oil rate, with the ₹5.90 wick and ₹70.80 jar, excluding labour, packaging, freight and GST, differs between the two candidates by the wax line alone. At 184 g per candle the gap is ₹16.85 a candle. That is worth knowing, but it is not the reason for the exercise; the reason is not losing a month of bestseller sales.

Now the comparison that justifies the work. Take an example of 300 bestseller candles a month across the three SKUs at an illustrative ₹799.00 each — replace both with your own figures. That is ₹2,39,700.00 of monthly bestseller revenue, consuming 55,200 g (55.20 kg) of wax. Against that, the full two-candidate bench costs ₹7,338.10 in materials: 3.1% of one month's bestseller revenue in this example. A two-week stock-out with no qualified second source puts roughly half that month's revenue at risk. With a qualified source, the cost of the same stock-out is the price of a delivery from Supplier B.

The decision rule and the first buffer purchase

Decide per SKU, record per SKU, and only then think about buying. The rule below is written so that two people reading the same sheets reach the same outcome.

If CSI 464 earns DUAL status on your bestsellers, the first purchase is a buffer, not a switch. In the example, 300 candles a month use 55.20 kg of wax; two weeks of cover is 25.48 kg, which rounds up in 5 kg and 10 kg bags to 3 × 10 kg of CSI 464 at ₹15,222.00. If Chunks earned it instead, the same cover is 6 × 5 kg at ₹17,970.00 — two 5 kg bags of Chunks cost ₹9.00 less than one 10 kg bag, so the smaller bags win here. Hold it in cool, dry storage — CSI 464's page states a 60-month shelf life under those conditions — and rotate it into a dual-approved SKU every few weeks so the buffer is always a lot you have actually burned. The backup supplier post covers how large that buffer should be for your lead times.

Where this sits in the supplier-selection cluster

This post is the SKU-by-SKU method for adding a second source. The siblings take the decisions on either side of it. Choosing between two suppliers for regular 100 kg orders is the head-to-head before either is approved; a supplier that is ₹60 a kilo cheaper is the switch rather than the backup; a new wax that looks identical explains why the burn test cannot be skipped. On the far side, the standard test candle is the reference used to screen new suppliers before any bestseller is risked, and the full selection and approval system places this step inside the whole process. If your second wax is a wider move across the range, switching suppliers without reformulating the collection handles the full catalogue.

Frequently asked questions