How to sell candles online without losing a summer of stock

You've poured maybe forty candles, four people have asked what you'd charge, and you've been quietly pricing tins at 2am. That's the point where it stops being a nice smell in the kitchen and starts being a question about money.

The short version. An 8 oz soy tin costs you somewhere around $4.50 to $5.20 in materials once you count the box, and it sells for $24 to $28. You cannot pour it and post it the same week, because soy needs one to two weeks of cure before it smells like anything from across a room. You'll spend your first month burning candles rather than selling them, because the wick is the product and you get it wrong three times before you get it right. And from roughly May to September, in most of the United States, you are shipping a solid that would rather be a liquid.

Everything below is the detail underneath those four sentences.

What a candle actually costs to make

Weigh your own, once, properly. These are the numbers a small maker in the US tends to land on in 2026, buying in hobby quantities rather than pallets.

Line	8 oz tin	Where the number comes from
Soy wax	$0.95	About 6 oz of wax once you allow headspace, at roughly $2.60 a pound in a 45 lb case
Fragrance oil	$0.95	8% load is about 0.48 oz, at around $2 an ounce
Wick, sustainer, sticker	$0.14	Cheaper in 500s, and you will buy three sizes
Tin and lid	$1.35	At 100 units. Under 50 units it's closer to $1.90
Label and warning label	$0.30	Two labels, not one
Mailer box and insert	$1.25	A candle mailer with a die-cut insert
Waste and test pours	$0.35	About 7%, and that's optimistic in month one
Total	$5.29

At $26 retail that's a 79% gross margin, which is why people say candles are a good margin business. They're right about the wax and wrong about the afternoon.

The afternoon is the thing. A pour session where you melt, add fragrance, pour, top up the sink holes, cool, label, and clean the pitcher is around two and a half hours for a batch of twenty. That's about seven minutes a candle, and if you pay yourself $20 an hour, your real unit cost is closer to $7.60. The full method for this, including the five costs makers forget, is in how to price what you make. Do it before you order 200 tins, not after.

One buying note that saves real money. Fragrance oil is the line that moves most, and it moves by a factor of three depending on whether you buy 4 oz bottles or 16 oz. A scent you've committed to should be bought at 16 oz. A scent you're testing should not.

Cure time means you can't pour to order

This is the constraint that reshapes the whole business, and almost nobody mentions it in the starting-out posts.

Soy wax needs to sit. The wax crystallises and the fragrance binds into it over days, and a candle burned 24 hours after pouring smells noticeably weaker than the same candle burned two weeks later. Most soy makers cure for one to two weeks. Some blends genuinely want the full fourteen days. Paraffin is faster, coconut blends sit somewhere in between, and your wax supplier's data sheet will give you a number to start from.

So "made to order" isn't available to you. If someone orders on Tuesday and you pour on Wednesday, you're shipping a candle that hasn't finished becoming a candle, and the buyer's honest review will be "smells lovely unlit, can't smell it burning". That review is correct and it's your fault.

What you do instead is pour to stock on a schedule and sell what's cured. Practically:

• Pour on the same day each week. Sunday is popular because Monday is a shipping day.
• Label the batch with the pour date on the bottom sticker, not just the box.
• Nothing goes on the shop until it's fourteen days old.
• Keep a two-week buffer of your top two scents and let the rest run out rather than panic-pouring.

That means your shop is a stock list, not a menu, and your product page needs a quantity that goes to zero. A candle maker who lets people order what isn't cured has invented a two-week delivery time and hasn't told anyone. Setting honest windows is covered properly in delivery times you can actually promise, and candles are the case where it matters most.

The wick decides everything and you'll get it wrong three times

The wick is the product. The wax is a delivery mechanism for the wick.

A wick that's too small tunnels. The first burn leaves a ring of unmelted wax up the sides, the candle drills a hole down the middle, and the buyer gets maybe 40% of the wax they paid for. A wick that's too big mushrooms, smokes, blackens the tin, and in a glass vessel it can get the glass hot enough to be a genuine problem.

Testing is boring and there's no shortcut. You burn a test candle for four hours, note the melt pool depth and whether it reached the edge, let it cool completely, and burn it again the next day. You keep a notebook with the wick code, the fragrance, the vessel and the room temperature. Four hours is the standard test burn because that's roughly what a normal person does on a normal evening.

Every fragrance changes the wick size. A scent swap in the same tin is a new burn test, not a new label.

That's the sentence that costs people money. Fragrance oils carry different amounts of vanillin and different viscosities, and a wick that's perfect in your unscented test will tunnel in your bakery scent. Ten scents in two vessel sizes is twenty burn tests, not two. Budget for it as stock, because that's what it is, and knock about $60 of wax and oil off your first month.

Two other things that only show up in testing:

Frosting and wet spots. Soy frosts. It's the wax crystallising and it's cosmetic. In a tin nobody sees it. In clear glass, the buyer sees white patches and a gap where the wax pulled away from the side, and they think it's a defect. This is the main practical reason new makers start in tins and amber jars rather than clear glass. If you do use glass, warm the vessel before pouring and cool the candles slowly, away from any draught, and accept you will still get some.

Sink holes. Wax shrinks as it sets and pulls a crater around the wick. You fix it with a heat gun across the top or a small second pour. Do it before the label goes on, not after.

Your fragrance supplier will publish a maximum usage rate for candle applications. Use it. Loading more oil than the wax can hold doesn't make a stronger candle, it makes an oily surface, a wick that drowns, and a flame that behaves badly.

Summer is a shipping problem, not a packing problem

Soy container wax softens well before it looks melted. A parcel spends its day in a metal van and its afternoon on a porch, and the inside of a closed vehicle in Phoenix in July runs far hotter than the air outside it.

What arrives is not a puddle. It's worse than that in a way, because it looks like carelessness. The smooth top is now a lumpy re-set surface. The label has slid. Fragrance has beaded on the surface and left an oily ring inside the lid. Sometimes the wick is off-centre because it floated. The candle burns perfectly fine. It looks like a second.

Nobody will pay you for that. Heat damage is not something couriers compensate for, and I'd check what yours says before assuming otherwise, because most of them exclude it plainly.

What actually works, in rough order of usefulness:

1. Ship Monday and Tuesday only in hot months. A parcel that enters the network on Thursday sits in a depot over the weekend, and the weekend is when nothing moves and everything bakes.
2. Say it on the product page. "Candles ship Monday and Tuesday. In hot weather they can arrive with a soft or uneven top, which sets again and doesn't affect the burn." A line like that turns a complaint into an expectation.
3. Sell wax melts through the summer. They cost less to make, weigh less to post, survive heat better in a clamshell, and they're a genuinely good product rather than a consolation prize.
4. Lean on local pickup and hand delivery. A candle handed over at a market in August is a candle that arrives intact.
5. Tell people the fix. A candle with a lumpy top self-levels the first time it burns to the edge. Say that in the note in the box and half your messages stop.

Insulated liners and cold packs exist and they cost more than the candle. For a $26 tin, they don't pay. This is one of the places where the honest answer is that you plan around the weather rather than engineer through it.

The rest of the packing problem is ordinary and solved. Rigid mailer, die-cut insert so the tin can't move, tape over the lid seam so it can't pop open, and the whole thing snug enough that it doesn't rattle when you shake it. Packaging that survives the courier covers the mechanics; the heat is the part specific to you.

Worked example: Maya, 8 oz soy tins, Phoenix

Maya pours 8 oz soy tins in black matte with a kraft label. Six scents. She sells at $26, or three for $70.

Her materials come to $5.29 a tin. She pours forty on a Sunday afternoon, which takes her about four hours including cleanup, so call it six minutes each. At $20 an hour that's another $2, giving a real cost of $7.29. At $26 that's a 72% gross margin before shipping and fees.

She sells around 45 candles a month. Roughly 25 of those go out at a farmers market on the first Saturday, which costs her $40 for the table and where she takes cash and card on a reader. The other 20 come from Instagram, and they split about evenly between local pickup at her front door and posted orders inside Arizona and to California.

The posted ones are where her margin actually lives or dies. A single tin in a mailer costs her $2.15 in packaging and $8.40 to post, and she charges $8 flat for shipping. So she's 55c down on shipping and fine with it, because a three-candle order costs $11.20 to post and she charges $8 there too, which hurts. Her fix was a $70 bundle where the shipping is priced in, which moved her average order from $26 to $48 and made the postage line stop mattering.

From June to September she stops posting candles altogether. She swaps the shop over to wax melts at $9 for a six-cube clamshell, keeps candle sales going for local pickup only, and pours ahead for October. Her September revenue is about 40% of her November revenue and that's fine, because she planned it.

The number that surprised her: her first-month wick testing burned through $71 of wax, oil and tins, and produced nothing she could sell. She'd budgeted $0 for it.

What to list and what to keep off the shop

A candle shop with 22 listings converts worse than one with six. You want the buyer to pick a scent, not to research.

• Three to six scents, maximum, at launch. One fresh, one warm, one sweet, one that's clearly the odd one and might become your signature.
• Name them so a person can smell them. "Tobacco, fig, cedar" beats "Wanderlust". Put the top note first.
• One vessel size until you're sure. Two sizes doubles your burn tests, your stock holding and your photography.
• A bundle. Three for $70 does more for your average order than any discount code.
• A sample or melt. A $9 entry product turns a browser into someone with your name in their order history.

Photograph the candle lit and unlit, and photograph the melt pool. Buyers who've been burned by tunnelling candles look for that shot specifically. One window and a plain surface is enough kit, and how to photograph what you sell is an afternoon well spent before your first listing.

On labelling: most markets expect a warning label with burn instructions and a way to identify the batch, and what exactly it must say varies by country and sometimes by state. Look up what your own market requires rather than copying another maker's label, because copying a label from a seller in a different country is how you end up with wording that doesn't apply to you.

Taking the order and getting paid

For a candle maker doing 45 units a month, the payment question is smaller than it looks.

Most of your orders will arrive from a photo. Someone sees the tin, messages you, and you send a link with the scent, the price and the pickup or postage option on it. On bank transfer, cash on pickup, WhatsApp, Instagram DM or Telegram, Sailo never touches the money and takes nothing at all. On card it's 1–3% of the goods, taken after any discount and excluding delivery and tax, so about 13c on a $26 tin. That 13c is not what decides anything. The $49 a month Business plan that card payments require is.

Run it honestly. At 20 online orders a month averaging $32, that's $640 through the shop, and $49 is 3.1% of it. At the market table you're already taking cash and a reader, so those don't count. Whether that's worth it depends on how many of your buyers refuse to do a bank transfer, and the only way to find out is to ask the next five. Do you need card payments to sell online runs that calculation at other price points.

The honest limitation, and it's a real one for a candle maker with six scents: the free plan stops at 10 products. Six scents plus melts plus a bundle plus a seasonal set is already eleven listings, and if you add a second vessel size you're at seventeen. You'll bump the ceiling faster than you expect. The other one is that Sailo can't tell you a bank transfer arrived. Only your bank can, which means ten minutes on a Sunday matching names to amounts, and if you get sloppy about it you'll post a candle to someone who never paid.

Month four, when it goes wrong

Someone says the candle "doesn't smell". Ask two questions before apologising: how long did you burn it, and how big is the room. A candle burned for twenty minutes in a large kitchen won't throw. If it burned two hours and reached the edge and still nothing, that's a real fault and you replace it. Track which scent it was, because it's usually the same one.

A tin arrives with a lumpy top in August. You already warned them on the product page. Send the self-levelling explanation and a $5 code, not a refund. Nine times in ten that's the end of it.

Your best scent gets discontinued. Fragrance suppliers reformulate and drop oils with no notice. When a scent becomes a bestseller, buy 32 oz and ask the supplier whether it's a stable line. Losing your signature scent with 40 tins of demand and no oil is a genuinely bad month.

Two people want the last three of a scent. Whoever paid first has it. Say that in the description before it happens.

Christmas. Your December order volume peaks in the last week of November, and cure time means those candles were poured in mid-November. Work backwards from the date people actually buy gifts, not from the date they give them, and stop taking posted orders about eight days before the courier's own cutoff rather than on it.

A wholesale enquiry. A shop asks for 60 tins at 50% off retail. At $13 a tin against a $7.29 real cost, that's $5.71 each and around fourteen hours of pouring, cleaning and labelling for $343. Do that maths before you say yes, and remember they'll want them all at once, cured.

What to do next

Buy one bag of wax, one vessel, and three wick sizes in the range your wax supplier recommends for that diameter. Pour three test candles this weekend, identical except for the wick, and burn them for four hours each with a notebook next to you. That single afternoon tells you more than a month of reading, and until you've done it you don't have a product, you have an intention.

Then work out your real cost per unit with your hours in it, pick three scents, and pour a batch of twenty on a Sunday. Mark the date on the bottom. Photograph one of them by a window on the day it finishes curing, put it on a link with a stock count and an honest shipping note, and post it once.

If it's June when you read this, pour melts instead and save the candles for October.