Is It Better to Save ₹10 Each on Wax, Fragrance, Jar and Packaging or Make One Larger Change to the Most Expensive Component?

The worked candle, and what ₹10.00 needs on each line

Your title gives two figures: ₹10.00 on each of four lines, which is ₹40.00 in total. Everything else here is a worked example, labelled as such and priced at the reader's own suppliers unless a CSI product replaces a line: a 160 g candle in a frosted 200 ml jar at 9% fragrance, selling at ₹549.00, 300 a month.

The jar is the most expensive line at 29.0%, so it is the "one larger change" candidate. Now ask what ₹10.00 actually requires on each of the four lines in the title:

The fragrance ₹10.00 is nearly there, but only by dropping a point and a half of load, which only your blind burn test can approve; if 8% is as low as it holds, the lever is worth ₹6.32. The jar ₹10.00 is the easiest if your supplier gives a volume price on the same jar: no test, no customer change. The box is a real ₹9.58, but visible. So the "small" path is four different kinds of change, not four identical ones.

The scoring table: rupees, tests, visibility, reversibility

Each change gets the same four questions: what it saves on real prices, how many test candles it needs before rollout, whether the customer can see it, and how easily you can undo it. Test candles are priced at the current candle's full material cost, ₹244.88 each (a conservative worked example), plus labour at ₹150.00 an hour.

Read the table across, not down. Path S spreads its saving over four changes, two of which need full candle tests and one of which (the wax) reopens almost everything about the candle. Path B concentrates the saving in one change with one test, but on the line the customer meets first. Neither path is free: Path S costs ₹6,357.84 to validate, Path B ₹3,403.92.

Why four small changes are more than four times the work

The two candle changes in Path S interact. A lower load can change how a wick burns, and a new wax can change how much fragrance it holds and releases. Run them together and a failure tells you nothing about which change caused it. Run them one at a time, each test on top of the change before it, and you need two test cycles in sequence. That is the real cost of the small path: not the candles, the calendar. How to keep a multi-change trial from losing track of what changed is the one-change-at-a-time post.

Why one big change is not one-fifth of the risk

The jar change needs a single burn test, because a new diameter needs its own wick check, but it is the change most likely to show up in your reviews. The clear Salsa looks different from a frosted jar: clear glass, a golden lug cap, and an upcycled jar by the page's own description. Whether buyers will notice or mind is not a burn-test question; it is a perception test with real customers, which is the jar-perception test post. What must be retested when the jar changes, even with the same wax and oil, is the jar-change retest post.

Time is money: when each saving starts

Each change starts saving only once it has passed. Using judgement timings (a candle test cycle of pour, cure and full burn is taken as about three weeks; a box panel or a supplier quote about two), Path S's box and jar-price changes go live at half a month, the load at three-quarters, and the wax at a month and a half, because it is tested on top of the new load. Path B goes live at three-quarters of a month. At 300 candles a month:

On these numbers Path B is worth more in every column: it is larger per candle, starts almost as soon, and costs less to validate. That is the arithmetic case for the one big change, and it is why "fewest validated changes" is the default rule. But the arithmetic assumes sales stay the same. That is the one assumption Path B puts at risk.

The visibility break-even

Path B saves ₹1,467.60 a month more than Path S at run-rate. Each candle you no longer sell because buyers dislike the new jar costs you its contribution: price minus materials, ₹549.00 − ₹203.38 = ₹345.62 (before fees, courier and overheads, which would make it smaller still). Divide one by the other: Path B stops being worth more than Path S once it costs about 4.2 sales a month, 1.4% of 300. That is a small number. A jar change that makes even one buyer in seventy hesitate is enough to reverse the decision.

You cannot know that number in advance, but you can test it cheaply: list both versions side by side, show photos to a buyer panel, or switch one sales channel first and watch conversion and returns for a month. Label every such result as yours; nobody, including us, can promise how your customers will react. If you are tempted to move the jar change to a new, cheaper product line instead of the existing candle, that removes most of the risk: existing buyers never see their candle change.

Test gate for either path

The method belongs to the change-control series; this page sets the order and the budget. For the load step, which percentages to compare is the lower-load test post and whether a lower load really performs is the 7% versus 10% test. For the wax step, the money-and-method view of a cheaper wax is the cheaper-wax revalidation post. For testing cheaper alternatives without risking a bestseller, use the 10% cost-reduction testing post. For the jar's wick check, three candles per rung is the three-candle wick approval post.

Whether every cost change needs the full validation of a new candle, or some (like a same-jar price) need almost none, is the validation-depth post.

What to buy for each path

Five items behind the two paths, chosen on page statements, price and stock on 4 October 2026; CSI publishes no throw or burn data, so nothing is ranked on performance. One is here to tell you not to buy it.

Skip for this decision: the coloured Salsa (₹62.00) and CSI's 200 ml matte and frosted jars (₹70.80, flat at 5 and 10): as a one-change route they save ₹9.00 and almost nothing respectively, and they are still a visible change. And skip any 10 kg wax bag bought for a discount: CSI's soy waxes are flat or nearly flat per kilo from 1 kg.

Frequently asked questions