How Much Extra Inventory Should I Buy for Custom Wedding Materials That Cannot Easily Be Used After the Order?

Your order and its custom materials, as numbers

The title gives no figures, so here is a worked example to replace with your own. A wedding client has approved a sample: 750 candles of 150 g at 8% (12 g oil + 138 g wax, total-weight basis) in CSI’s Clear Glass Jar with Golden Lid (150 gram), coloured marigold with Orange Candle Dye Liquid at your recipe card’s 1.0 g per kg of wax, scented with Gilli Mitti, and labelled with the couple’s names from your own printer. You want 10 planned extras (client samples, retains, burn tests), so you need 760 good candles. Over your last three orders, 3% of the mix you made never ended up in a saleable candle (heels, pitcher residue, rejects you could not rescue). Your price is ₹520.00 a candle (₹3,90,000.00).

The loss rate is measured on mix, not on candles you threw away: mix made ÷ (saleable candles × fill) − 1. That one number already contains heels and residue, so it is not stacked with a separate heel allowance. If you have never measured it, measure it on your next two orders (counting pitcher heels and rejects shows how); until then a labelled 3–5% is a starting point, not a benchmark.

Visible or invisible: shrink what needs a buffer

Before any arithmetic, sort every material by one question: would the client notice if it changed? They will notice the colour, the scent and their names on the label. They will not notice whether the wax under the colour came from bag one or bag twelve, which wick sits in the middle, or what the outer carton is. Everything in the second group should be a standard material you would use anyway, bought and handled so that it never becomes client-specific.

This single decision does more than any formula. In the worked plan the client-specific lines come to ₹95,304.00, 40.0% of a ₹2,38,547.00 purchase; the other lines carry no leftover risk at all.

The steady losses: expected plus two spreads

Your 3% is an average. Some orders lose 2%, some 4.5%. If you buy exactly the average, any order that runs worse than average is short, and with custom materials short means a second purchase, possibly a different lot, or a client told there will be fewer candles. So the steady buffer has to cover a bad order, not an average one.

A simple, honest way to size it uses the two numbers you have. Treat each candle as having a 3% chance of being lost; across 760 candles the expected loss is 22.8 candles’ worth and the spread (standard deviation) is the square root of 760 × 0.03 × 0.97, which is 4.70. Expected plus about two spreads covers roughly 98 orders in 100. This is a normal approximation and a planning tool, not a promise; it assumes losses are independent, which is exactly why the next section exists.

Two spreads is the worked choice because the next step up buys four more candles’ worth of oil nobody else will use for a sliver of extra cover, and the step down leaves one order in ten short. If your client is unusually forgiving about a handful of missing candles, 1.3 spreads is defensible; if the count is sacred (one candle per guest, names on each), stay at two. Note what drives the figure: a lower loss rate shrinks both the expected loss and the spread, so measuring and cutting your losses is the most valuable buffer reduction there is.

The lumpy loss: one extra batch, of custom materials only

The confidence formula covers the small, scattered losses. It does not cover the day the oil is weighed for 50 kg instead of 5 kg, the batch is overheated and goes dark, or the colour comes out two shades too deep. Those failures take a whole batch at once, and when the materials are custom you cannot fix them with a quick order of something similar. The protection is blunt and effective: keep one extra batch of the client-specific materials, and only those.

Batch size is therefore a buffer decision as well as a production one: running the custom wax in 5 kg batches rather than 10 kg halves the allowance and saves ₹3,584.00 in oil on this order. A failed batch’s candles are melted out of the same jars, so the jars need no batch allowance; the wax for the remake is ordinary stock you hold anyway.

The purchase sheet and the leftovers

Custom lines carry the buffer; standard lines just carry the order. Every line still runs required → buffer → pack rounding → final order → leftover.

The oil line is the one to read slowly. The candles strictly need 9,120 g; the plan buys 9,950 g. If losses run exactly at the average and no batch fails, about 556.4 g (₹4,985.34) will be left over. That is the real price of being confident, and it is why the buffer belongs in the quote rather than in your hope that it gets used. If every gram is used, only 34 g remains.

Compare the common shortcut. "Twenty per cent extra on everything custom" turns 9,120 g into 10,944 g, bought as 10,950 g for ₹98,112.00: ₹8,960.00 more than the plan, with no deliberate allowance for a failed batch and more leftover client-specific oil when nothing goes wrong. And "exactly the order" buys ₹81,984.00 and leaves you exposed the first time a candle cracks.

Pricing the contingency into the quote

Everything in the custom buffer exists because of the client’s choices, so it is part of their price. Put it on the quote as its own line rather than spreading it invisibly, for two reasons: the client can see why custom choices cost more, and you can see the number when the next client asks for something unusual.

What happens to unused custom material afterwards should be decided now, in writing. The fair default is that a priced-in buffer is yours: if you can sell Gilli Mitti candles in the monsoon, the leftover becomes stock you were paid to hold. If the client wants the unused labels or oil, hand them over. What you should not do is charge for the buffer and then also charge for leftovers; once is enough (the same double-recovery trap 6351 covers for a fragrance change).

What you never skip, and what protects you

None of this replaces the approvals and tests that make the custom formula safe to produce. A buffer sized for 3% losses assumes the formula is proven; if it is not, the loss rate is unknown and the buffer is a guess.

If the calendar cannot hold those steps, the honest answer is to move the date or simplify the brief, not to buy a bigger buffer; why every critical material should be in hand before a date is promised is 6355.

The materials in this plan

Two client-specific materials and three standard ones. Nothing is ranked on performance (CSI publishes no burn data); the selection follows the worked brief.

Frequently asked questions