How Do I Take a Candle From Rough, Frosted or Cracked Appearance to a Validated Premium Finish — Identify Wax Versus Process Causes, Optimise Pouring and Cooling, Calculate Finishing Labour, Compare CSI Wax Samples, Pilot Production and Buy in Bulk?

Stage 1 · Describe the defect precisely, because four observations point at four different variables

Write down what you are looking at, precisely enough that a photograph would settle an argument. A rough or crystalline top is grainy or pitted under raking light — crystallisation and contraction at the air interface. Frosting is the whitish feathery bloom across a coloured surface or up the glass, worst where you least want it: dark dye in clear glass. Cracking is mechanical. And the fourth question is no defect description at all — “it looked right when I poured it and it does not now” is a stability observation.

Each points at a different part of the chain. Rough tops and frosting sit downstream of crystallisation, so pour temperature, cooling and anything interfering with it — dye, fragrance, an additive — are candidates. Cracking is contraction and release, which is why it shows up most in pillar and moulded work. Instability over weeks says the candle you approved and the one your customer unboxes are not the same object.

Which gives the only honest position on “is this normal?”. With no appearance specification published, there is no standard against which your top is normal or faulty. We will not reassure you and we will not call it a defect. The standard is one you create: pour a batch, choose the unit you are willing to sell, photograph it under fixed lighting, write the formula and process on the sheet, and keep it unburned as the retained reference.

Stage 2 · Wax versus process: the four-jar single-pour test

The question underneath this series is whether the appearance you dislike is intrinsic to the wax or produced by what you do to it — answerable in an afternoon, with wax you own, at no cost in new material. Melt one batch (one wax, one fragrance, one load, one dye level, one wick), pour four identical jars, change exactly one thing. A single melt removes every batch-to-batch variable at a stroke.

Run it as rounds, in the cost order below: pour temperature; then jar temperature, two at room temperature and two pre-warmed with the temperature recorded rather than called “warm”; then cooling and ambient; then load; then the oil; then dye in measured drops. Only after all six fail does “this wax will not give me this finish” become a conclusion.

Every round is judged at the same four points against the retained reference: at pour, at 24 hours, at one week and at one month. Appearance is not a day-one property, and CSI publishes no ageing or appearance-stability data for any wax. Add one temperature cycle between a warm room and an air-conditioned one.

Stage 3 · Optimise pouring and cooling — one published variable, two unpublished ones

Three of the sixteen waxes publish a pour temperature and one a fragrance-addition temperature. That is the complete set. CSI 464 states 80–90°C; CSI 468 states 58–60°C with a melting point of 50–53°C; Luxury Soy Chunks gives two conflicting ranges on one listing, so only the 75–80°C overlap is defensible. Eco Soy CSI 400 publishes none — its 85°C (185°F) is a fragrance-addition instruction.

Jar temperature and cooling being unpublished makes them your two highest-value experiments — unpublished does not mean unimportant, it means nobody has done the work for you. Record real numbers: jar temperature at pour, room temperature, whether a fan or air conditioner was running. “Cooled slowly” is not data. The same discipline answers the cracking question and the sinkhole question.

One finishing tool exists: the Hot Air Gun at ₹531.00, showing low stock: 1 unit. There is no automated finisher and no production surfacing machine at any price — so remediation is hand work, which is what you are about to price. Whether a heat gun suits your top is a separate question from what it costs on every unit.

Stage 4 · The cheap interventions, in order, before the expensive one

Dye before anything, because it is most often over-applied and has no published guidance at all: no dye dose, no dosage chart, no drops-per-kilogram figure and no shade card exists anywhere on this store. There is also no powder dye and no pigment product — the colourants are liquid dye, dye chips and mica, and mica is a suspended particulate, not a dye. Dye chips publish no weight, so no per-gram figure can be stated. Run a ladder in measured drops instead. Entry is ₹236.00 for 15 ml on every shade — and black and brown each exist in two pricing families at different prices: compare one black listing with the other.

Then load and the specific oil, one at a time — both yours, because no appearance effect is published for any oil in any wax. The only published constraint is the ceiling, on three waxes: Luxury Soy Chunks 13% of total wax weight, CSI 464 10%, CSI 468 13%. Dye and load interact, so test them separately first.

One comparison we will make, one we will not. Vybar at ₹2.87–₹5.74 per 200 g candle is comfortably cheaper than any wax change, because one rung up the real ladder is ₹10–₹25 a candle at 184 g. What we will not claim is that it beats your own finishing labour — no labour, throughput or energy figure is published anywhere here.

Stage 5 · Calculate your finishing labour, and convert it into ₹/kg of wax

The formula is the only honest form of this answer: (your seconds per candle ÷ 3,600) × your hourly labour cost × your units per month. CSI publishes no finishing time, no wage, no electricity cost, no throughput and no reject rate, so if you have read “heat-gunning takes two minutes” anywhere — including in your own head — treat it as your figure. Time ten real units; take the median.

Read the last column slowly. If your finishing labour is your 45 seconds at your ₹150 an hour, that is ₹1.88 a candle — so a wax change removing finishing entirely only repays itself if the new wax costs no more than ₹10.19 per kilogram extra. Now read the ladder: Natural Soy Pearl to CSI 464 is ₹82.60/kg, ₹15.20 a candle; CSI 464 to Luxury Soy Chunks is ₹91.60/kg, ₹16.85. Eight or nine times the labour it would save. It is arguable only at 120 seconds and ₹250 an hour, which buys ₹45.29/kg — enough for the ₹37.00/kg step to CSI 468.

Two closing rules. The right unit is cost per saleable candle, not cost per kilogram: wax plus rework labour, over units that pass your own standard — a denominator only your pilot produces, because CSI publishes no reject or rework rate. And 45 seconds at ₹150 an hour is ₹937.50 across 500 units a month, ₹3,750.00 across 2,000. We will not project sales, growth or payback.

Stage 6 · Comparing CSI wax samples — the trial, priced and re-added

If stages 1 to 5 have run and the surface still fails your standard, a wax comparison is justified. 500 g packs exist on Luxury Soy Chunks, CSI 464, Soy Pillar, Premium Coconut Soy, CSI 468, Coconut Wax, Soy Butter Wax, Fully Refined Paraffin and Microcrystalline — but not on Natural Soy Pearl, Eco Soy CSI 400, Semi Refined Paraffin, Beeswax or Gel Wax. So a five-candidate trial buys three 500 g packs and two 1 kg packs. At 184 g per 200 g candle, 1 kg is five test candles and 500 g is two.

The Clear Glass Jar with Golden Lid publishes a 150 g fill weight, which makes a per-candle figure arithmetic rather than assumption: at 8% of finished weight a 150 g candle is 12 g of fragrance and 138 g of wax, so a 100 g bottle scents eight and 1 kg of wax pours seven. The matte, frosted and glossy family at a flat ₹70.80 publishes no gram capacity, so no per-candle cost may be built on it. Other published capacities: 70 gram votive ₹31.86, 500 gm Apothecary ₹212.40, 600 g Borosilicate ₹200.00.

Judge all five on the same four points, the same cycle, the same reference and the same wick variant. Because CSI publishes no wick sizing data of any kind, a wax that wins on appearance has won nothing yet; “Thin (C1)” and “Thick (C2)” is the whole published range. If one wax wins on tops and another on throw you have a trade-off, not a winner.

Stage 7 · The pilot — and the one number none of this can be decided without

A 5–10 kg pilot is cheap to specify, because 5 kg and 10 kg packs are real on most waxes. Say CSI 464 came through stage 6: its 5 kg is ₹2,537.00 and its 1 kg ₹507.40, so 6 kg is ₹3,044.40. At 138 g per 150 g candle that is 43 units — but jars come in packs of 10 and 20, so two packs of 20 at ₹1,888.00 gives 40 jars for ₹3,776.00. The binding constraint is rarely what you were thinking about.

Now the number the page has been building towards. Photograph all 40 units at all four points, run the temperature cycle, and count how many still meet your standard. That count is the denominator that turns a price into a cost — and CSI publishes no reject rate, no rework rate and no defect-rate figure, so nobody can supply it but you. With figures that are explicitly hypothetical and explicitly yours: reject 3 of 40 and material cost per saleable candle rises from ₹251.19 to ₹271.55; reject 8 and it is ₹313.98. That ₹62.79 swing dwarfs every step in the stage-5 ladder.

Add the stage-5 finishing labour on top and you have a per-saleable-candle figure comparable across candidates. Do it for the incumbent too — a pilot that measures only the new wax has no baseline. Keep one unburned unit as the new reference, because an incoming QC specification with no physical reference attached is a document, not a control.

Stage 8 · Buy, lock and monitor — and the bulk tier that does not exist

The published wax ladder is 500 g / 1 kg / 5 kg / 10 kg, and Natural Soy Pearl’s 50 kg at ₹21,240.00 is the only pack above 10 kg in the entire catalogue. A “25 kg purchase” is two 10 kg plus one 5 kg — on CSI 464, ₹5,074.00 + ₹5,074.00 + ₹2,537.00 = ₹12,685.00, which is ₹507.40/kg, the same rate as one kilogram; a “100 kg order” is ten 10 kg packs, ₹50,740.00 at that same rate. And no MOQ, slab, rate card or quantity discount is published anywhere.

What gets locked after approval is a short list, specific enough that another person could reproduce the candle: the product name and handle, the exact variant, the pour window validated inside the published range, the load with its denominator, the dye count in drops from a named bottle, the wick variant, the jar with its measured fill weight, the cooling conditions, and a receiving code per incoming wax lot.

Monitoring later lots is the same four-jar test, run small and often, photographed against the reference. A lot that differs is then a conversation you can have with a number and a photograph instead of an impression. Note the limits: shelf life exists on two waxes only, CSI 464 and CSI 468 at “60 months, cool dry storage”, while no wax publishes a heat-resistance figure, a softening point or a storage-temperature limit.

Frequently asked questions