Black Friday for very small shops

Everyone in your feed is planning 40% off and you're quietly wondering what happens if you just don't.

Mostly, nothing bad. For a shop doing under a hundred orders a month, Black Friday is usually a bad trade: you sell roughly the volume you were going to sell anyway, at a worse price, to a mix of people who'd have bought in December and people who will never buy again. If you make 40% gross margin and you knock 20% off, you have to sell twice as many units to end up with the same money in your account. Not 20% more. Twice.

That number is the whole article. Work it out for your own margin before you decide anything, because the answer for a 70%-margin digital product and a 25%-margin resold item are completely different businesses.

The multiplier, in one table

The arithmetic is simple and nobody does it. To hold the same gross profit after a discount, you need to multiply your unit sales by margin ÷ (margin − discount).

Your gross margin	10% off	20% off	30% off	40% off
25%	1.7×	5×	Impossible	Impossible
40%	1.3×	2×	4×	Impossible
55%	1.2×	1.6×	2.2×	3.7×
70%	1.2×	1.4×	1.8×	2.3×

Read the 25% row again. If you buy something for $30 and sell it at $40, a 20% discount means you need five times the orders to break even against your normal week. Five. And "impossible" is literal: at 30% off on a 25% margin you're selling below cost, so every extra order makes the hole deeper.

This is the sentence that should stop you: a discount you can't fund is a loss you scale. Volume doesn't rescue a bad unit economic, it multiplies it. The shops that get destroyed in the last week of November are not the ones that sold nothing. They're the ones that sold a lot.

Two things make it worse than the table suggests.

Ad costs peak in the same week. Everybody with a budget is bidding for the same attention, so whatever you normally pay to reach a thousand people, you pay more in that window. If your discount is funded by ads, you're paying twice for the same order.

Returns run higher on discounted orders. People who buy on impulse at 40% off change their minds more than people who bought at full price after thinking about it. A return isn't just a refund, it's outbound postage, return postage, and an item you may not be able to sell again.

Work out what you'd actually clear

Do this on one product, on paper, before you commit to anything.

Take your best seller. Write down:

1. Retail price.
2. What the goods cost you: materials, or wholesale cost, plus packaging.
3. What it costs to get it to a buyer: postage, and the packing time if you're honest about it.
4. What the payment takes. Card processing is a percentage plus a fixed fee, and the fixed fee is what mauls small baskets.

Subtract 2, 3 and 4 from 1. That's your contribution per order. Now apply the discount to line 1 only, because none of the other costs go down when you cut your price. That's the part sellers get wrong: they model a 20% discount as if it takes 20% off everything, and it takes 20% off the only line that was paying for the other three.

If you want the fee side of that done properly, how payment fees eat a small order has the arithmetic on small baskets, including why a fixed per-transaction fee hurts a $22 order far more than a $60 one.

What to sell instead of price

There are four things a very small shop can offer in the last week of November that don't touch the unit margin.

Capacity. If you make things by hand, the scarce thing isn't price, it's your Saturday. Selling twelve slots at full price with a named dispatch date is a better offer than 25% off, and it's a much better offer for you.

A bundle at a real price. Two items that normally cost $32 and $18 sold together at $44 is a 12% discount on the pair, and it raises the value of the order rather than lowering it. You've also moved a second item without a second postage cost. Bundles protect the anchor price, which matters in January when you go back to full price.

A free add-on with a known cost. A sample, a card, a small thing you already make that costs you $1.40. It reads like a gift and it costs a fraction of a percentage discount. The trick is that the buyer values it at its retail price and you pay its cost price.

Early access rather than a lower price. Existing buyers get first pick of a limited run on Wednesday, everyone else on Friday. Nothing is discounted. Your best customers get something they actually want, which is not being beaten to the thing they wanted.

If you're going to test any discount at all, the mechanics of doing it without wrecking your baseline are in how to run a discount without losing money.

The cash trap

The version of this that ends shops isn't the discount. It's the cash cycle around it.

The pattern runs like this. In September you buy stock on the strength of a good August. In October you buy more, because Black Friday. In November you sell it at 30% off and it feels like the best week you've ever had. In December the invoices for the September stock come due, the courier bills you for the return leg on the refused parcels, and the money from your marketplace hasn't cleared yet.

You had a record month and no money. That's not a rare story, it's the standard one.

If you're funding Black Friday stock out of cash you need for something else, the discount isn't the risk. The purchase order is.

If you do run one, run it properly

Deciding to participate is fine. Doing it vaguely is not. Four rules.

Pick one product, not the shop. A shop-wide percentage is a decision you've made about products you haven't looked at, including the ones with your worst margins. One product, one price, one reason.

Cap the units and publish the cap. "30 at this price" does three useful things: it caps your downside, it creates a real reason to order now, and it gives you an honest end to the promotion that isn't a countdown timer you'll extend on Sunday.

Set an end time and keep it. The seller who extends the sale "by popular demand" has taught every buyer to wait next time. You only get to train them once.

Decide the cut-off for Christmas delivery before you launch. A November discount that generates orders you can't ship until the third week of December is a customer service problem wearing a sales costume. The dates and the backwards planning are in Christmas orders without the chaos.

Worked example: Dana, candles, Portland

Dana pours soy candles in a garage in Portland. One product does most of her revenue: a 9oz jar at $32.

Her real costs, per unit:

Line	Cost
Wax, wick, fragrance oil	$4.80
Jar and lid	$2.90
Label, box, tissue	$1.60
Postage, US domestic	$6.40
Card processing, roughly	$1.23
Total	$16.93

So on a $32 sale she clears $15.07, which is a 47% contribution margin. She sells about 55 a month.

Now the Black Friday plan she was going to run: 30% off. That's $22.40 retail. Costs don't move except processing, which drops a few cents. Contribution falls to roughly $5.75.

To make the same money as a normal month she'd need to sell 144 candles. She can pour about 30 a day and she has a job. It was never possible, and she'd have found that out in the middle of it.

What she did instead:

• A two-candle gift set at $56 (a 12.5% discount on the pair) in a single box, one postage. Contribution on the set: about $25, which is better per order than a single at full price.
• 40 sets, capped and stated. They sold out in four days.
• A free hand-poured tea light with every order over $50, costing her $0.90.
• No shop-wide sale. Everything else stayed at full price all week.

She sold 40 sets and 21 singles. Gross profit for the week beat her previous best November. She also still had a price to go back to on the first of December, which is the part that doesn't show up until later.

The other thing she did, which was free: she counted those Black Friday buyers again in March. Eleven of the 40 set buyers had ordered again. That's a number worth having, and if you never measure it you'll keep running promotions on faith. How to get repeat buyers is where the follow-up sequence lives.

The week after is the expensive one

The promotion is the fun part. The bill arrives in the following ten days, in four forms, and none of them are on the spreadsheet.

Packing. Forty orders is not four orders ten times over. You run out of boxes, then out of tape, then out of the specific label size, and the post office queue in the last week of November is not the queue you're used to. Buy packaging for double what you plan to sell, in October, when it's in stock.

Messages. Every promotion generates a wave of questions that have nothing to do with the promotion. Where is it, can I change the colour, does it come gift wrapped, can you hold it until December. Budget an hour a day for a week.

Returns. These land two to three weeks later, which is the same fortnight you're trying to ship Christmas orders. A discounted item coming back in mid-December is worth less than it was in November because you've now missed the window to resell it at all.

Your baseline. December after a heavy November is usually softer than December after a quiet one, because you pulled the demand forward. Compare November and December together, never November alone. A shop that measures the promotion week in isolation will run it again next year on bad evidence.

The case for doing nothing

You do not need to run Black Friday. Nobody has ever unfollowed a shop for not having a sale.

The specific situations where sitting it out is the right call:

• You sell handmade and you're already at capacity. More demand at a lower price is a worse business, not a bigger one.
• Your margin is under 35%. The table above says the volume you'd need doesn't exist.
• You've been trading under six months. You don't yet know your baseline, so you won't be able to tell whether the promotion did anything.
• Your buyers are gift buyers who'll buy in December anyway. Discounting in November just moves December's revenue backwards and makes it smaller.

There's also a positioning argument. A shop that has never discounted has a price that means something. A shop that discounts every November has taught its buyers that November is when they buy, and the other eleven months quietly get worse.

The honest limitation

If you want to run a discount code on Sailo, coupons are a Business-plan feature, which is $49 a month or $468 a year. On the free plan and on Pro at $19 a month, there's no coupon field. You'd be editing the price on the product itself and putting the reason in the description, then editing it back on the Monday.

That's fine for the one-product, one-price, capped promotion described above. It's genuinely annoying if you wanted a code to give to a specific group. Worth knowing before you plan around a feature you don't have.

And on card payments specifically, Sailo takes 1–3% of the goods after discount, excluding delivery and tax, on top of whatever Stripe charges. On Dana's $56 set that's 56 cents to $1.68, depending on her plan. At her volume the plan she's on costs more than the rate does.

Do this in October

Open a spreadsheet and put your best seller's real cost per unit in it, all four lines, including your time. Work out the multiplier for a 10%, 20% and 30% discount at that margin. If any of them says "impossible", you've just saved yourself a November.

Then design the bundle instead, cap it at a number you can physically make, and write the cut-off date into the product description before you post about it anywhere.

If November's promotion is really about shifting stock you regret buying, that's a different problem with a different answer, and it's in end of year stock clearing.