How Do I Build a Complete Candle Raw-Material Procurement System from Product BOM → Sales Forecast → Gross Requirement → Existing Stock → Production Buffer → Supplier MOQ → Pack-Size Selection → CSI Purchase Order?

Eight steps, each one answering a single question

A complete procurement system for a candle business is not software; it is eight questions asked in a fixed order, each feeding the next. What does one candle contain? (the BOM). How many of each will I make? (the forecast). What does that add up to, with allowances? (the gross requirement). What do I already have or have coming? (existing stock). How much cushion do I need while the next order is on its way? (the buffer). What is the smallest quantity I can actually buy? (the supplier minimum). Which real packs cover it most cheaply? (pack selection). And finally, what exactly do I order today, and what waits for a test? (the purchase order).

Each question has its own post in this series, and this page is where they join. It works one five-SKU business through every step with real CSI prices, so you can see what each step changes. The headline is simple: the gross requirement and the purchase order are very different numbers, and the difference is exactly what the middle steps — stock, buffer, minimum, packs — are for. Skip them and you either over-buy (ordering the gross with stock already on the shelf) or run out (ordering the net with no buffer for a slow delivery).

Steps 1 and 2: the BOM and the forecast

The example business sells five SKUs: Lavender, Dark Vanilla and Woody Oud as 200 g candles in the White Matte Glass Jar (the reader's weighed fill; CSI does not publish this jar's fill), and Lavender and Dark Vanilla as 85 g candles in the Golden Aluminium Tin (CSI's published 85 g fill). Everything is poured in Luxury Soy Wax Chunks at 8% (its page states a 13% maximum). Wicks are the reader's burn-test choices — Cotton Braided in the jar, Eco Thin C1 in the tin; CSI publishes no wick-size chart. Next month's forecast, the reader's own, is 400 candles.

The BOM is per candle and per SKU, never per batch; batches come later. Building it is the job of the exact-product cluster and turning a product idea into a complete BOM. The forecast is per SKU, because oils are per SKU — converting monthly sales into requirements and forecasting five SKUs separately cover it, and growing 20% a month shows how to raise it without over-buying.

Step 3: the gross requirement, with the allowance stack

Multiply each SKU's forecast by its BOM and add up by material. Shared materials pool across SKUs — the wax across all five, Lavender across a jar and a tin. Then apply the allowance stack in the series' fixed order: test candles as whole candles (none here — the formula is approved), a 2% reject allowance on consumables (÷ 0.98), 3% pour loss on wax and oil only, and 1% breakage on glass only. All three percentages are labelled working assumptions. The result is the gross requirement: 62.90 kg of wax, 2,590 g of Lavender, 1,871 g of Dark Vanilla, 1,009 g of Woody Oud, 279 jars, 133 tins, 276 Cotton Braided and 133 Eco Thin C1 wicks, and 409 stickers.

This is the step the 100-candle and 1,000-candle clusters work through in depth: 100 lavender candles in 8 cm jars is the hub for small runs, 500 wedding candles in two fragrances for large orders, and the 1,000-candle procurement sheet lays the step out as a table. Combining 20 SKUs' BOMs into one wax figure shows the pooling for a bigger range.

Steps 4 and 5: subtract stock, add the buffer

Step 4 nets off what you already have: stock on hand plus anything already on order. Step 5 adds back a safety stock — the cushion that covers a busy month arriving with a slow delivery. This page uses the series' max-minus-average form: daily use × (1.2 × worst lead time − usual lead time), with lead times of 4/7 days for wax, 7/12 for oils, 10/16 for jars and tins and 5/8 for wicks and stickers. Those lead times and the 1.2 peak factor are the reader's labelled examples; CSI publishes no lead times. The result: net requirement = gross + safety stock − on hand − on order. Where it is zero or negative, nothing is bought this month.

Read the net column. The wax gross of 62.9 kg becomes a net of 54.1 kg because 18 kg is already on the shelf, even after adding a 9.2 kg buffer. Woody Oud's 1,000 g on hand leaves a net of 258 g. The jars, with the longest lead time, carry the largest buffer in days. The buffer logic is the subject of safety stock by material role and separate reorder points per material; the stock columns are the ones the reorder sheet keeps up to date week by week.

Steps 6 and 7: supplier minimums and pack selection

Step 6 checks each net requirement against the supplier's effective minimum. CSI states no minimum order quantity; the smallest listed in-stock size is the practical minimum — 500 g for Chunks, 15 g for Lavender, Dark Vanilla and Woody Oud (production orders here start from 50 g), 5 White Matte jars, 12 Golden tins, 10 Cotton Braided, 20 Eco Thin C1 (its 10-pack is out of stock on the 4 October re-check), 40 stickers. Where a net requirement is below the minimum, the minimum is bought and the remainder is carried — the MOQ post covers how to account for it.

Step 7 rounds each line to the cheapest real combination of in-stock packs, using each product's own price ladder. Two CSI quirks matter here. For Chunks, two 5 kg bags (₹5,990.00) are cheaper than one 10 kg (₹5,999.00), so the wax line is 11 × 5 kg. For oils, the per-gram ladder differs by oil — Lavender is ₹5.40 a gram at 100 g, ₹5.24 at 500 g and ₹5.22 at 1 kg, so its line is 2 × 1 kg + 3 × 100 g + 1 × 50 g. The 13 kg versus 10 kg wax post and the 1.3 kg oil post work these choices in full.

Step 8: the CSI purchase order

The purchase order is the worksheet's priced lines, nothing more. Approved materials whose net requirement is positive go on it; anything still in testing does not. The CSI subtotal is ₹91,842.36, before labour, packaging, freight and GST. Non-CSI lines — lids for the White Matte jars, labels, inbound courier — are example lines you replace with your own quotes; they are never part of the CSI total.

Whether each line is bought monthly or in bulk is a policy you set once per material — the return-on-cash test decides it — and which lines are worth a volume conversation is decided by how many SKUs share them. Both feed back into step 7: a bulk line buys a quarter's net requirement; a monthly line buys one month's.

Where every post in the series fits

The series is built around these eight steps. Each cluster hub owns a stretch of the system, and each sibling answers one question inside a step. Use the map to jump straight to the post for the step you are stuck on.

Start-up budgets enter the system at step 8 from the other direction — a fixed sum rather than a fixed forecast. Dividing ₹10,000 between wax, oils, wicks and jars is the hub for that cluster; the arithmetic is the same, only the constraint changes. The first full run through the system is usually the first 1,000 candles, where the test kit and pilot stand in for a forecast you do not yet have.

Running it every month

The system becomes a routine once it is a loop. When the order arrives, it moves from on-order to on-hand. When the month ends, actual sales replace the forecast for that month, the next forecast is updated, and the buffer is recalculated if sales or lead times moved. Then the eight steps run again. After three or four months your own lead times replace the examples, your own peak factor replaces 1.2, and the purchase order becomes a number you trust rather than a number you hope for.

Two signals tell you the system needs attention. If you keep dipping into safety stock, the forecast is low or the lead time is optimistic. If a material's net requirement is negative month after month, it is being over-bought or its SKUs are slowing — check its commonality before it becomes dead stock. Both show up in the worksheet's columns without extra work.

Frequently asked questions