Should I Negotiate Raw-Material Prices Based on Each Order Quantity or Commit to Monthly Purchase Volume for Better Commercial Terms?

Four ways to buy the same wax

Whichever structure you choose, the volume itself must already be real. When your consumption first justifies asking at all is the when-to-ask post: largest listed pack more than once a month, around ₹25,000.00 a month on one material, and three stable months.

What the wax line is worth on one candle

The title gives no candle, so this uses the series default: 200 g at 8% fragrance, 184 g of wax and 16 g of oil (total-weight basis). Wax and oil at CSI list prices; jar and label are worked figures.

Wax is 39.3% of this candle, so even a 4% wax discount moves COGS by 1.6%. That is why the structure matters more than the headline percentage: a commitment that goes wrong for one month can cost more than a year of the difference between 2% and 4%. The difference itself, at 100 kg a month, is ₹12,177.60 a year.

Worked: 100 kg a month under each structure

The worked buyer (labelled; use your own) uses 100 kg of CSI 464 a month, 543 candles of 200 g, and today buys 25 kg a week at list. Illustrative terms: 2% off any single order of 100 kg or more; 4% for a 12-month take-or-pay commitment of 100 kg a month released weekly; 3% for a commitment with a ±20% flex band (80–120 kg); 4% with price reversion (no penalty, but list price for any month under 100 kg). Cost of money 1% a month (worked).

The per-order break costs something the invoice does not show: to hit the 100 kg threshold you buy a month at a time instead of weekly, so your average stock rises by 37.5 kg and ₹19,027.50 of cash sits on the shelf. At a worked 1% a month that is ₹190.28, which is why a 2% break nets ₹824.52. If the break needed a quarter's stock in one order, the cash side would grow and the sums would belong in the three-months-of-wax post.

Now let the forecast be wrong. On judgement odds (your own will differ), usage averages 70 kg with 20% probability, 100 kg with 60%, 125 kg with 20%:

The ranking flips once the forecast can miss. Take-or-pay wins at 100 kg and 125 kg and loses ₹13,192.40 in a 70 kg month, enough to make it an expected loss of ₹913.32 a month. The flex band loses less in the bad month (₹3,856.24) and stays positive on average (₹522.62), though on these odds it trails the plain per-order break. The reversion clause has the best expected value here because its downside is losing the discount, not paying for wax: if your supplier will offer that structure, it is the one to ask for.

When the forecast is wrong

The rule that settles take-or-pay is one line: break-even usage = committed volume × (1 − discount). Commit 100 kg at 4% and you need to use at least 96 kg in a month for the commitment to cost no more than buying what you used at list. Commit 100 kg with a band floor of 80 kg at 3%, and the floor's break-even is 77.6 kg. Below those lines, every kilo you did not need was paid for anyway.

The cash basis in the tables is deliberate. In a 70 kg month the take-or-pay buyer has paid ₹14,613.12 for 30 kg of wax they did not need. If the range is steady and that wax will be used next month, the real economic cost is only carrying it, about ₹146.13 a month at the worked 1%. If the range changes, a grade is dropped or summer storage spoils the bags, the cash basis is the honest one. You do not know in advance which month you will have.

Two cousins of this question have their own posts. If the supplier's better price comes with a much larger minimum order rather than a monthly commitment, the cash-versus-saving maths is the MOQ and working capital post. If the commitment would make you dependent on one supplier for a critical wax, read the two-approved-suppliers post first; and if the supplier you would commit to is cheap but often out of stock, price that with the stock-out cost post.

Scheduled releases, flex bands and price holds

Scheduled releases

You agree a quantity for a period (say a quarter) at the volume price, and take delivery in dated releases: weekly, fortnightly or per production run. Your shelf holds what a weekly buyer holds, so the carrying cost of the per-order break disappears. Ask for release dates to be movable by a week or two with notice; production slips more often than it runs early.

Flex bands

Instead of a single committed figure, a range: 100 kg ± 20% means anything from 80 kg to 120 kg a month keeps the agreed price. The worked illustration gave up one point of discount (3% instead of 4%) for the band, and in exchange lost ₹3,856.24 instead of ₹13,192.40 in a bad month. Set the bottom of the band at or below your low case.

Price-hold windows

A price hold keeps today's list price for a set period, often in exchange for a forecast rather than a commitment. It saves nothing on day one. It earns its keep when list prices rise: an illustrative 5% rise on 100 kg a month is ₹2,537.00 a month, ₹15,222.00 over a six-month hold. It is also what lets you print a retail price list for the season without guessing. If you are deciding between holding more stock and holding a price, note that the hold ties up no cash.

What never to commit to without tests

Fragrance needs the same discipline with one extra twist: oils are bought by the kilo across many scents, so a commitment usually means a total fragrance volume with the split by oil left open. Commit to scents with a track record, not to a launch you have not sold yet; how much oil a production run really needs (trials, wastage, safety) is in the 1,000 candles at 10% fragrance purchase post, and what to ask about batch consistency before a long-term arrangement is in the batch consistency and availability post.

With CSI: what is listed, what is a conversation

Beyond the list, we discuss volume, scheduled dispatch and price holds on WhatsApp, case by case. We publish no committed-volume discount and will not invent one here; CSI dispatches from Pune and does not publish lead times or freight, so release dates are confirmed per order. And we will not ask you to commit to a volume of a grade you have not tested.

The waxes and oil in the worked case

Three waxes that show the three kinds of listed ladder (flat, stepped, and bigger-is-dearer) and one oil with a per-order step. Chosen on page statements, price and stock on 4 October 2026; CSI publishes no throw or burn data, so nothing is ranked on performance. Counts use 184 g of wax and 16 g of oil per candle.

Frequently asked questions