I'm Comparing Two Candle Jar Suppliers — One Is Cheaper but Has More Breakage; How Do I Calculate the Real Cost per Usable Jar?

The formula in three layers

Price per piece answers "what does the supplier charge?". Cost per usable jar answers "what does each sellable candle's jar cost me?". The second number is higher whenever a single jar breaks, because some jars you pay for never reach a customer, and each of those takes your time and sometimes your wax and oil with it. Build it in three layers so you can see which layer is doing the damage.

Layer 1: price ÷ survival rate

If you buy N jars at price p and a share b of them break, you have N × (1 − b) usable jars for N × p rupees. Per usable jar that is p ÷ (1 − b). At 8% breakage, divide by 0.92; at 2%, by 0.98. A ₹60.00 jar at 8% becomes ₹65.22; a ₹70.00 jar at 2% becomes ₹71.43. This layer alone narrows a ₹10.00 sticker gap to ₹6.21.

Layer 2: + admin time per broken jar

Every broken jar takes time: separating it from good stock, photographing it, messaging the supplier, sweeping and wrapping glass for safe disposal, updating your count, and sometimes re-planning a production day around a shortfall. Put a number on it. In the worked example we use 12 minutes per broken jar and value your hour at ₹200.00, which is ₹40.00 per broken jar. Both are illustrative: use your own time and your own hourly value, and remember an hour spent on a claim is an hour not spent pouring or selling.

Layer 3: + materials lost in jars that fail after pouring

Some flaws are only found the hard way: a hairline crack or a thin spot that gives way when hot wax goes in, or a jar that cracks during a burn test. By then the jar has taken wax, oil and a wick with it. In our example one 150 g candle at 8% is 138 g of Luxury Soy Wax CSI 464 (₹70.02 at the 10 kg price), 12 g of Jasmine (₹47.08 at the 1 kg price) and one Eco C2 wick (₹9.44): ₹126.54 of materials, excluding labour, freight, packaging and GST. That is more than the price of either jar.

Freight, GST and the cost of waiting for a replacement belong in the same equation, and they are the subject of the landed-cost guide. We keep them out here deliberately, so the breakage effect is visible on its own. If one supplier charges freight and the other does not, you must add it before you decide.

Worked example: ₹60.00 at 8% against ₹70.00 at 2%

Two hypothetical suppliers, called A and B. We are not describing any real company, and CSI is neither of them; the point is the method. You are ordering 400 jars of a similar 200 ml size. A quotes ₹60.00 a jar; B quotes ₹70.00, 14.3% more. From your records, A's deliveries run at about 8% breakage, of which a quarter shows up only after pouring; B's run at about 2%, with the same quarter found late. Every input here is illustrative.

Read the bottom line carefully. A is still cheaper, by ₹1.44 per usable jar. But 86% of the ₹10.00 sticker advantage has gone, and it went in three places you would not see on an invoice: jars you paid for and threw away, your time, and wax and oil poured into glass that failed. If A's breakage creeps up to 10% (40 jars, 10 of them after pouring), A comes to ₹74.63 per usable jar, ₹1.74 more than B. A supplier whose breakage wanders between 6% and 10% is cheaper in some months and dearer in others, which is its own planning cost.

Where the jar breaks decides what it costs

"Breakage" is not one number. A jar found in pieces when you open the carton, a jar you reject for a chip at inspection, a jar that cracks when hot wax goes in, and a finished candle that arrives broken at your customer's door are four different costs. Log them separately; the formula only works if you know which bucket each one is in.

Stage 3 is where suppliers genuinely differ and where your own process can make things worse. Cracks are commonly linked to thermal shock (a cold jar meeting hot wax, a draught across a warm jar), to existing chips and flaws, and to uneven wall or base thickness. Heavier glass is not automatically safer. If one supplier's jars crack on pouring and the other's do not, test both at the same pour temperature before blaming either: a jar poured at 90°C in a cold room is not a fair test of anything. The Clear Glass Jar with Golden Lid (150 gram) page states "Safe for hot pour up to 85°C"; most CSI jar pages give no figure, and CSI 464's page gives a pour temperature of 80–90°C, so in that jar you pour CSI 464 at 80–85°C: the lower of the two stated upper limits governs.

Stage 4 is not a supplier comparison at all, but each loss there costs more than at any earlier stage, so keep it in the same tally. If your parcels break, no jar supplier will fix it; your carton, cushioning and the courier you choose will.

Admin and replacement time: the cost nobody invoices

Admin time is the easiest layer to leave out, because it does not arrive as a bill. It still costs you. Here is what 12 minutes per broken jar looks like in practice; time your own and replace our numbers.

Two things make admin time grow faster than breakage. First, claims are batched per delivery, not per jar, so a supplier with occasional heavy breakage costs less admin per jar than one with a steady trickle. Second, the real cost of a shortfall is often a missed dispatch date, not minutes. If 32 jars short means a corporate order ships late, the late-delivery cost dwarfs the jar cost. That is why buffers exist: how many extras to order for a first run of 100 and the buffer percentage by order type deal with it.

Replacement policy: the input that moves the answer most

If a supplier replaces jars that arrive broken, the jar cost of stage-1 breakage comes back to you, but only the jar cost. Your admin time remains, you wait for the replacement, and stage-3 failures after pouring are hard to claim, because by then nobody can easily tell a glass flaw from a process fault. Re-run the worked example with stage-1 breakage replaced free:

The replacement column matters more than the breakage gap in this example. That is why a replacement policy is worth getting in writing, with its conditions: what evidence is needed (photos, an unboxing video), within what time, whether replacements travel with your next order or separately, and whether freight on the replacement is charged. A policy that needs a video you did not take is not a policy you can use.

What about us? CSI does not publish a breakage rate or a replacement figure on its product pages, and we are not going to invent one for this guide. Ask us on WhatsApp what we need from you if jars arrive broken, before you order, and keep the reply. Ask the other supplier the same question in the same words. Then put both answers into the table above.

The break-even breakage rate, and reading the chart

The most useful single number in a supplier comparison is the breakage rate at which the cheaper jar stops being cheaper. Below it, buy the cheaper jar and manage the breakage. Above it, the dearer jar is the cheaper choice.

Price-only break-even

Set the two layer-1 costs equal: p_A ÷ (1 − b_A) = p_B ÷ (1 − b_B). Solve for A's rate: b_A = 1 − (p_A ÷ p_B) × (1 − b_B). With ₹60.00, ₹70.00 and B at 2%, that is 1 − (60 ÷ 70) × 0.98 = 16.0%. On price alone, a jar 14.3% cheaper can lose nearly one jar in six and still be cheaper. That is why the price-only view flatters cheap glass.

All-in break-even

Add admin and after-pour losses and the break-even falls sharply, to about 8.9% in the example. The chart at the top of this page shows why: A's all-in curve starts at the sticker price but climbs faster, because each broken jar now costs its price, 12 minutes and, one time in four, ₹126.54 of materials. The flat lines are B, whose breakage is fixed at 2% in the example.

Use the table as a quick screen: if a jar is only 5% cheaper, it has to break less than about 6.9% of the time on price alone, and less than that all-in. Small discounts rarely survive a breakage difference. Large discounts sometimes do. The all-in line depends on your recipe, your time and when your failures appear, so recompute it with your own m and a rather than reading ours.

Measuring breakage honestly: a tally sheet, not a feeling

It is easy to "know" which supplier breaks more without ever having counted. The formula is only as good as the b you put in it, and small samples mislead: in a pack of 10, a single broken jar is 10% breakage, so one test pack cannot tell a 2% supplier from an 8% one. Here is how to build a number you can trust.

The rows above reproduce the worked example (A: 18 + 6 + 8 = 32; B: 4 + 2 + 2 = 8), so you can check your spreadsheet against our numbers before you trust it with yours.

CSI jars to run the comparison on

You can use this formula on any two jars. If CSI is one side of your comparison, these are the listings to test, with the per-jar arithmetic already done. We do not publish breakage rates for any of them; the tally sheet above is how you would find ours.

Before the Diwali reorder, 8 November 2026

Diwali falls on 8 November 2026, 43 days after these prices were checked. If you are choosing between two jar sources for festival stock, you have time for one logged test delivery, not three. In that case, lean on the formula's structure rather than on a measured rate: take the supplier whose replacement terms you have in writing, order your festival buffer on top, and keep the tally sheet running so the post-Diwali decision rests on data. Dispatch is from Pune; confirm the dispatch date for your quantity before you promise customers a date.

Frequently asked questions