If I Save ₹25 per Candle and Produce 5,000 Candles a Month, How Do I Determine Whether the Reformulation Effort Is Financially Worthwhile?

₹15,00,000.00 a year, and what sits under it

Your two figures, ₹25.00 a candle and 5,000 candles a month, are yours; we use them exactly. Multiplied out: ₹1,25,000.00 a month, ₹15,00,000.00 a year, ₹28,846.15 a week, ₹4,109.59 for every calendar day. For scale, at 2,500 a month the same saving is ₹7,50,000.00 a year and at 10,000 it is ₹30,00,000.00. The volume does the heavy lifting: the same ₹25.00 that is a pleasant extra for a small maker is a salary-sized number here.

The title does not say where the ₹25.00 comes from, so here is one illustrative way a 200 g candle could find it. Every unit price on the left is the reader's own supplier's figure (illustrative, not a CSI price and not a claim that anyone is overpriced); the right-hand wax is CSI's live price. The candle size is our labelled default, because the title gives none.

Two separate changes make up the ₹25.00. The load step (9% → 8% on the reader's own wax) removes 2 g of oil worth ₹8.52; 2 g of wax at ₹520.00 replaces it for ₹1.04, so it nets ₹7.48. The wax step (184 g of reader wax to 184 g of Natural Soy Pearl Wax) saves ₹17.52. Together: ₹25.00, which is 9.4% of the before-COGS. Keeping them apart matters later: they can pass or fail independently, and a business case that lumps them together cannot bank half a result.

What the ₹25.00 becomes after the test

A reformulation is a new candle until proven otherwise. A new wax can need a different wick, can refuse to throw at the lower load, or can produce more frosting or rough tops that you scrap. Each outcome changes the per-candle figure, and at 5,000 a month even small changes move the annual number by lakhs. These are the scenarios worth writing down before you pour a single test candle:

Three things stand out. First, even the weakest passing outcome, the load cut alone at ₹7.48 a candle, is worth ₹4,48,800.00 a year: there is no scenario short of outright failure in which this test is not worth running. Second, the wick matters more than its price suggests, because it is the line most likely to change with a new wax (standardising wick SKUs without forcing an unsuitable wick covers the risk of one wick everywhere). Third, rejects are where a "cheaper" wax quietly gives its saving back: 2 extra points of rejects costs ₹4.92 per saleable candle. If rough tops need a heat gun, add the labour too; the cheaper-wax-that-cost-more post works that case through.

The minimum worth chasing is easy to compute: divide the one-off by your volume. With the worked-example one-off of ₹55,963.44, any validated saving above ₹11.19 a candle pays the whole test back within one month, and anything above ₹0.93 within a year. At 5,000 a month almost every real saving clears that bar. The discipline is not finding savings big enough to test; it is refusing to bank any of them before the test says so.

The one-off bill: matrix, labour, labels, dead stock

Our worked example assumes 6 SKUs share the reformulation (use your own count). Each SKU gets four arms of three candles: A the current candle as control (reader wax, 9%); B the load cut only (reader wax, 8%); C the new wax at 8% with your current wick; D the new wax at 8% with Eco C2 as a second wick rung. Comparing A with B isolates the load, B with C the wax, C with D the wick: one variable per comparison, all poured in the same week. Why that discipline matters is why one variable at a time matters.

Test materials come to ₹16,397.44; labour is the largest single line at ₹15,000.00, because burning 72 candles to the end is supervised time. Spread over six SKUs the whole bill is ₹9,327.24 per SKU. The 10 kg Pearl bag (₹4,248.00) would also cover the test, but 7 kg in one 5 kg and two 1 kg bags costs ₹3,024.00 and leaves 376 g spare for a pitcher heel: buy the test quantity, not the production quantity. How to cost a much larger reformulation, twenty existing candles, in detail is the cost of reformulating 20 candles; this page needs only the order of magnitude, and the order of magnitude is days of savings.

Dead stock of the old material

Old wax is a cost only if you scrap it. Suppose you hold 600 kg of the old wax (worked example), ₹3,12,000.00 of stock. At 910 kg a month on the current formula it covers about 20.1 days of production, well inside a six-week test. Time your last order of old wax to run out as the test ends and the dead-stock line is nothing. The risk is the opposite: a large old-wax order placed "because the price was good" just before you approve a switch. Keep buying the old material monthly until the new formula passes.

Relabelling: a cost only if the label must change

If your label says "soy wax", a different soy wax may need no change. If it says "100% soy", check the new wax's wording before you test: the Natural Soy Pearl page says "Made from a blend of soy wax", while the Eco Soy CSI 400 page lists "Wax Type 100% Soy Wax". The ₹8,000.00 line is a worked example for new plates and a print run; it is nothing if the wording holds, and a reason to ask us first if it might not.

Time and risk: the costs that actually decide it

Time. A credible test is pour, cure, then burn to the end, and a pilot batch before full rollout. Our worked example allows 6 weeks (judgement; cure practice and burn length vary by candle). During those weeks you give up ₹28,846.15 a week: ₹1,73,076.92 in total, 3.1 times the entire one-off bill. That is the honest price of testing, and the diagram shows it: by week 16 the lucky rushed route is ₹2,29,040.36 ahead of the test-first route.

Risk. The unlucky rushed route ships the new formula untested and finds out from customers. If a month of output has to be withdrawn or replaced, the materials alone are 5,000 × ₹241.32 = ₹12,06,600.00: 9.7 months of the full ₹25.00 saving, before courier, refunds, marketplace penalties and the repeat customers who quietly stop buying. Divide the time cost by the materials at risk and you get the break-even probability: rushing only pays if the chance of a failed month is below 14.3%. With six SKUs, a new wax and a wick that has not been bracketed, that is not a bet a careful maker takes.

The worked-example risk reserve. Even a passing formula can behave differently in thousands of homes. The ₹16,566.00 line assumes returns run 0.5 percentage points higher for two months, each return costing the new COGS plus a ₹90.00 reship (worked example; use your own rates). It is deliberately pessimistic and still small. If returns do not rise, it is money you keep.

How to shorten the wait without skipping anything

Pour every SKU and every arm in the same week, not one SKU at a time. Bank the arm that passes first: if B (load only) passes and C (new wax) fails, you still roll out ₹7.48 a candle, ₹4,48,800.00 a year, while the wax question goes back for another wick rung. Approve SKU by SKU rather than waiting for the slowest. And start the clock now: a test poured this week can be through cure and well into its burns before Diwali on 8 November 2026, 35 days from 4 October 2026, but do not let the festive rush turn into a rollout deadline (see the strip below).

Go or no-go: six criteria

Pass every one and you roll out; fail any one and the change waits. This page owns the money; the paperwork belongs to the change-control form, and the full testing approach for cost-downs on a bestseller to testing cheaper alternatives without risking a bestseller.

Buying for 5,000 a month: what bigger packs don't save

At this volume it is tempting to assume the biggest pack hides an extra discount. On CSI's list it mostly does not, and we would rather say so than let you build it into the business case.

Three honest points. Wax: 920 kg a month is 18.4 of our 50 kg bags; one bag covers about 1.7 days of production. Natural Soy Pearl's price falls 5.6% from 1 kg to 5 kg and then stays at ₹424.80 a kilo through 10 kg and 50 kg, and CSI lists no 25 kg or 100 kg packs. Whether a commercial quantity could do better is a conversation, not a listed price: ask on WhatsApp, and do not put a discount into the business case until it is quoted. Wicks: Eco C2 is ₹9.44 a wick in the 1,000-pack and the 10-pack alike; ₹47,200.00 a month either way. Fragrance: 1 kg is the largest oil pack CSI lists. And if your oil costs ₹4,260.00 a kilo, do not swap it for ours to save money: our Woody Oud at the 1 kg price is ₹72.50 for 16 g, ₹4.34 more than yours, and Lavender is ₹83.54. A cheaper CSI oil exists (Citrus Lemon is ₹2,832.00 a kilo), but a different oil is a different scent, not a cost tweak.

Cash is the last check. One month of Pearl is ₹3,90,816.00; at a 1.5% monthly cost of money (worked example) holding it costs ₹5,862.24 a month, small against ₹1,25,000.00 of savings. The danger is not holding a month; it is buying three months of a wax that has not passed yet. When bigger orders really do earn their cash, the three-months-of-stock post does the comparison; the cost of switching supplier rather than formula is the 15% cheaper supplier post.

What to buy for the test, and what to skip

Four items the worked-example matrix uses, chosen on page statements, price and stock on 4 October 2026. CSI publishes no throw or burn data, so nothing here is ranked on performance; your arms decide that. Per-candle figures use 184 g of wax (200 g at 8%).

Skip for now: the 50 kg Pearl bag and the Eco C2 1,000-pack until a formula passes; both cost the same per unit as the test packs. And skip any production quantity of anything the matrix has not approved.

Frequently asked questions