How to prepare for Black Friday without losing money

Black Friday is a logistics problem wearing a marketing costume. The offer takes an afternoon to decide; the stock, the shipping cutoff, the support inbox and the question of whether the discount leaves you anything take six weeks. Most small stores spend six weeks on the offer and an afternoon on everything else, which is the whole reason the day goes badly.

So this is how to prepare for Black Friday backwards, from the date. Pick the date, count back, do what is due.

The discount arithmetic, before anything else

Decide the discount with one calculation, because it determines everything downstream: how much stock you need, how much cash you commit, and whether a record day leaves you poorer.

Take an illustrative item at $40 that costs you $16, so gross margin is $24. Now discount it:

At 40% off you need three times the units for the same gross profit — and you pay three times the packaging, picking, card fees and returns. The real break-even is worse than the right-hand column, not better.

Two conclusions fall out. A 50% storewide discount on a 60%-margin product is a volume bet most small stores cannot win. And a smaller discount on a narrow selection — one hero product, a bundle, free shipping over a threshold — usually beats a storewide number, because the headline draws people and the basket pays you.

If your margins are thinner than the example, redo it with your own figures before you promise anything. Pricing arithmetic is the same calculation in the other direction.

Six weeks out — stock and the one decision you cannot reverse

Choose what is on offer and order the stock. Suppliers get slower in November, not faster. Whatever you order in the last fortnight arrives in December.

Decide the quantity by last year, or by honesty. If you have last year's numbers, use them. If this is your first season, size the order to what you could sell at normal pace over the following two months, so unsold stock becomes January inventory rather than a loss.

What goes wrong: ordering for the day you hope for. Discounted stock that does not sell is money you paid full price for, sitting in a spare room.

Four weeks out — the systems that break under load

Load does not break a store gracefully. It finds the place where two systems disagree and widens it.

• Overselling. The same last unit sold three times in a minute. Check that stock decrements at the point of sale, decide now what you do when it happens, and write the apology email while you are calm.
• Discount codes. Test every code. Test whether two can be combined. Test whether a code works on a product that is already reduced, and whether that combination can take an item below cost.
• Checkout. Put a real payment through and refund it, the same test as on launch day. A payment path that quietly broke in September will not announce itself on the busiest day of the year.
• Email bursts. A large campaign sent in one go from an address that normally sends twenty messages a day is how mail starts landing in spam. Warm up with something useful beforehand.
• Shipping cutoffs. Get the carrier's last dispatch dates in writing, subtract your own packing time, and publish the resulting date. Then stop promising delivery and start promising dispatch — the carrier owns the rest and will not be answering the phone.

Three weeks out — plan the support volume, not just the orders

Support volume rises faster than order volume, for three reasons that compound. Buyers are disproportionately first-timers who have never used your store. Many orders are gifts, so the deadline is real and the questions are urgent. And discount mechanics generate questions of their own — does the code apply, why did it not apply, can I use it on yesterday's order.

Practical preparation:

1. Write five canned answers now: where is my order, does the code work on X, what is the shipping cutoff, can I change the address, how do returns work in a sale.
2. Publish the cutoff date and the returns window on the storefront, not only in the confirmation email.
3. Decide the refund rule for the person who bought Thursday at full price. "Within seven days of the sale starting, we refund the difference" costs little and prevents a public argument.
4. Block out the hours. The inbox is the job on sale day; the parcels are the job on the two days after.

This is where an AI support agent earns its place — mAutomate's AI Customer Support answers on your website, and the Pro plan at $149 extends that to an AI call center on the site, with a real phone number on Scale at $349. It helps because most sale-week questions are the same five questions. It does not replace you on the hard ones: when an order is wrong or a customer is upset, some questions should reach a human.

Two weeks out — build the audience you will sell to

The people who buy on your sale day are mostly people who already knew you existed. Two weeks out is when you tell the list the date, not when you try to acquire strangers at auction prices.

• Email the list with the date and one sentence on what will be discounted. Do not reveal the whole offer.
• Schedule the social posts now, while you still have time to look at them. mAutomate's AI Marketing Suite drafts and schedules campaigns across channels, and nothing publishes until you approve it — the setting you want in a week when a wrong price in a post becomes a hundred emails.
• Set up the abandoned-cart message if you do not have one. Sale-week carts are abandoned while the buyer comparison-shops the same product elsewhere.
• Ad costs rise in late November. If you have never run ads, this is the most expensive possible week to learn.

One week out — the dress rehearsal

Do all of this on a Monday, not on Thursday night:

1. Create the discount, apply it to a test order, and confirm the final charge is the number you intended.
2. Buy something on a phone, on cellular data, and complete the purchase.
3. Confirm the confirmation email arrives and does not land in spam.
4. Pack one real order and time yourself. Multiply by the number of orders you expect. If that exceeds the hours in the day, recruit help now rather than at midnight.
5. Count the packaging. Boxes, tape, labels and printer ink run out mid-batch, and every shop is shut.
6. Take a backup or export of your product and customer data.

The day, and the two days after

Launch in the morning so you have daylight for the first failures. Watch stock on the first hour's bestsellers and pull anything that oversells before it compounds. Reply to support faster than usual: on a sale day a slow reply becomes a refund request.

Then treat the two days after as work days. A sale whose orders arrive late converts a good day into a month of complaints and chargebacks.

What preparation does not fix

Two things, honestly.

A sale cannot create demand that was not there. If a normal week brings 30 visitors, Black Friday brings slightly more than 30 visitors, discounted. The day amplifies an audience; it does not supply one. If you have no list yet, the list is the project — before launch and after, since the first ten customers come from people you contact directly, not from a discount.

And a discount trains buyers to wait. Run 40% off every quarter and your regulars stop buying in between: you have not grown the business, you have moved the same revenue into lower-margin months. One deep sale a year is a strategy. Four is a new price list.

The timeline in one table

mAutomate plan prices are on the pricing page, and every plan includes a 14-day free trial with no card required.