Shopify BFCM 2026: The September Checklist for a Store That Holds Up

Shopify reported $14.6 billion in sales across its merchants over Black Friday to Cyber Monday 2025, up 27% on the year before. The figure is the platform's; the share of it a store took depended on what the store had done by the end of September, because the eight weeks after that are for running the campaign, not building it.

Shopify publishes a checklist every year, and most merchants read it in November. This is the same ground in the order the work has to happen, with the parts that fail most often first. It is written for a merchant whose store already sells, on a marketplace agency's reading of what breaks on the weekend.

The BFCM visitor arrives from an email or a social ad on a phone, often on a slow connection, with twenty other tabs open. A product page that takes four seconds to become usable loses a share of them before it has said anything. Shopify's own theme and checkout are fast; what slows a store is what was added to it, and the site speed article sets out which apps, scripts and images do the damage and how to find them.

The September job is to measure the home page, the top five product pages and a collection page on a mid range phone, remove what is not earning its place, and measure again. A store that has not been measured this year has apps installed for campaigns that ended in 2024.

Buy something. On a phone, with a discount code, with the express payment button, with a shipping address that has an apartment number, and with a payment that fails. Every one of those paths is used a thousand times on the weekend and most merchants have used none of them since the store launched. Shop Pay, Apple Pay and PayPal in the express row are the difference between a checkout that takes ten seconds and one that takes two minutes on a phone, and the two minute checkout is where the abandoned cart figures come from.

The discount is the other test. Automatic discounts do not need a code and do not get mistyped; a code in an email is a code that will be typed wrong by a share of the people who try. Set the BFCM discount to apply automatically where the offer allows it, and test that it stacks or refuses to stack the way it was meant to.

The BFCM offer decides the demand and the stock decides whether the demand becomes orders. A store that emails 40,000 people a 30% offer on its best product and has 300 units is a store that oversells by Saturday morning and spends December refunding. Inventory in Shopify has to be true on the morning of the sale, across every channel that sells from the same units, and the article on supplier feeds covers keeping it that way when a supplier's feed is the source.

Fulfilment is the same question at the packing bench: who packs the Saturday orders, on what carrier, with what pickup. Shopify Fulfillment Network, a third party warehouse or Amazon's multi channel fulfilment take the bench out of the equation for the weekend, and the multi channel fulfilment article sets out when the last of those is worth its fee.

The offer, the email sequence, the SMS, the ad creative and the on site banners are one campaign with one calendar, and they are built in October and scheduled, not written on the Thursday. Shopify's Campaign Autopilot from the Spring 2026 Editions plans and runs the sequence across email and the store from a single brief, for merchants who want the machine to do the scheduling; the brief still has to say what the offer is and who it is for.

The traffic the campaign brings lands somewhere, and the somewhere is usually a collection page. Why Shopify collection pages do not rank is about search, and the same fixes, a real description, a real title and products in an order somebody chose, are what make the BFCM landing page convert the traffic that was paid for.

The buyer on the weekend wants to know two things before paying: when it arrives and what it costs to send. Both belong on the product page, not discovered at checkout, and both have to be true for the volume the weekend brings and the carrier surcharges the season carries. The last shipping date for Christmas delivery goes on the site in early December and comes off the day it passes.

Many stores read this checklist in November rather than September. The weeks left are too few to rebuild anything, and enough to make sure nothing breaks. The list shrinks to what protects the weekend.

Test the checkout today. Every path a buyer will take: phone, discount code, express payment, a failed card. It is the one item that can cost the whole weekend, and testing it takes an hour.

Choose one offer and keep it simple. A single clear discount, set up and tested, is better than a clever structure built in a hurry. Decide which discount type the store will use and check that it applies to the right products and stacks the way it should.

Count the stock honestly. Set inventory to what is really on the shelf, and decide what happens when a product sells out: hide it, show it as sold out, or take pre-orders with a stated date.

Write the emails now and schedule them. Three or four messages across the weekend, written and scheduled this week, not on the day.

Put the shipping promise on the page. The order by date for delivery before the holidays, and the cost of shipping, visible before checkout.

Prepare the support replies. Saved answers for where is my order, can I change my order and when will it ship. Most weekend messages are those three.

Then stop changing things. The freeze below matters more for a late store than an early one.

September. The building weeks: speed measured and fixed, checkout tested as a buyer on every path, the offer decided and costed, and the stock ordered against the supplier's real lead time rather than a hopeful one.

October. The campaign built once: the emails written and scheduled, the advertising set up and tested at small budgets, the discounts configured and tried, and the support replies for the questions that come every year written in advance.

Early November. The freeze. No new apps, no theme changes, no new payment settings and no migrations from about two weeks before the weekend. Every change made in the last fortnight is a change nobody has watched under load.

The weekend itself. Running, not building. One person watches the store's speed and checkout, one watches stock and pauses advertising on anything about to sell out, and one answers customers. Each knows in advance what they are allowed to change and who decides the rest.

The week after. Write down what broke and what sold, while it is remembered, because that note is the first page of next September's checklist.

Most BFCM failures are caused by a change made in the last week: a new app, a theme edit, a discount restructured on the Wednesday. A freeze, decided in advance and kept, removes that whole category of risk.

Set the date. A day, usually a week or more before Black Friday, after which the theme, the apps and the checkout settings do not change. Write it in the calendar and tell everyone who can edit the store, agencies included.

What the freeze covers. The theme and its code, installing or updating apps, checkout and payment settings, shipping rates and the structure of the offer. These are the changes that break things nobody notices until orders stop.

What it does not. Product stock counts, scheduled emails, banner copy and advertising budgets can still change, because they have to, and they rarely break the store.

A rule for emergencies. If something must change during the freeze, one person decides, the change is made with a way to undo it, and it is tested as a buyer straight away.

Keep a copy of the theme. Duplicate the live theme before the freeze. If something goes wrong on the weekend, publishing the copy restores the store in minutes.

Lift it after the rush. The freeze ends after Cyber Monday's orders are out, not on the Tuesday morning. The week after the weekend is still busy, and the same caution applies.

Conversion rate by the hour, against the same hours last year or the previous weekend. A sudden drop with traffic steady is almost always something broken: a discount not applying, a payment method failing, a page that stopped loading on phones.

Checkouts started against checkouts completed. If people reach checkout and stop there, the problem is inside it: shipping costs appearing late, an express button missing, a code that will not apply. That is fixed in minutes once it is seen, and costs the whole weekend if it is not.

Hours of stock left on the best sellers. Not units, hours, at the rate they are selling now. The advertising on a product with six hours of stock left is paused before it sells out, not after, so the money goes to products that can still be bought.

The weekend ends on Monday night. The work it created runs for another week or two, and how that week goes decides the reviews, the returns and whether the new customers buy again.

Ship the backlog oldest first. Orders from the weekend arrive faster than most stores can pack them. Ship strictly in the order received, and tell customers whose orders will take longer before they ask.

Answer the where is my order wave. Most messages that week are about delivery. A short email with tracking, sent as each order ships, prevents most of them.

Expect the returns later. Returns from the weekend arrive through December and into January, especially for gifts. Plan the space and the time to process them, and keep the return policy for gifts clear.

Ask for reviews at the right time. A review request sent after the product has arrived and been used, not the day it ships, turns a busy weekend into lasting social proof.

Welcome the new customers properly. Many weekend buyers are new. A useful first email about the product, rather than another discount, is the start of the repeat purchase that makes the weekend's acquisition pay.

Reconcile the stock. Count what is left against what the store thinks it has, before the next promotion relies on the numbers.

Write the post-mortem. What broke, what sold out, which emails worked, what to start earlier next year. Written in the first week of December, it is next September's checklist.

The support desk is the part of the checklist nobody wants. Where is my order, can I change the address, the code did not work, I want to return it. Templates for each, written in October, and a person who answers them on the Saturday. A store that answers in an hour keeps the customer; a store that answers on Tuesday keeps the refund request. For the brands we run, this checklist is what Shopify management means in September, with the marketplace stores run from the same stock and the same calendar.