BFCM: The Backwards Plan

What the ten weeks buy

Black Friday Cyber Monday is the one week where an email-led store earns a disproportionate share of its year, and the week itself is unimprovable: inboxes are full, ad prices triple, and nothing you decide after Wednesday makes the weekend better. Every lever that matters sits weeks earlier. The list you will mail is decided by the cleanup you ran in September. The margin you keep is decided by the floor you wrote in early October. The five sends that carry the weekend are good or generic depending on whether they were written in a calm week or a loud one.

The one rule: the week is won in September. This board exists to move every decision into a dated slot long before the noise, so that by Black Friday morning there is nothing left to decide. If you are writing copy on the Friday, the board failed. If the week feels boring, it worked.

How to think about it

Backwards. The anchor is your year's BFCM Friday, and every phase is dated from it, not from whenever you happen to start.

• T-10 weeks is list work, and it is unskippable. Deliverability is earned on the sends before the season, not during it. Sunset the dead weight, run one honest re-engagement pass, and accept the smaller number: a list that opens is worth more in November than a list that is large.
• T-8 weeks is the offer, and the offer is three decisions, not one. What the deal is, what the floor is (the number below which nothing goes without editing the card that states it), and what you will NOT do. The cut list is the highest-leverage card on the board: every BFCM disaster story is a decision made in the loud week that a calm week would have refused. A store that discounts nothing can still run this board; the offer phase then answers "nothing site-wide" and the week sells on something better than a percentage.
• T-6 weeks is the early-access segment. The head start is the one BFCM mechanic that adds rather than subtracts: it rewards the customers who already love you, it pulls revenue into a quieter window, and it gives the weekend sends a warm list to land on. Define VIP from behaviour, not flattery.
• T-4 weeks is production, all of it. Five touches, written, staged, and individually approved a month early. The gift guide is deliberately the odd one out: content on the Saturday when every other brand is shouting percentages is how a store sounds like itself in the one week everyone sounds the same.
• T-2 weeks is QA and the holdout. The holdout is what makes December's numbers honest: without it, the week's revenue claims every order the store would have taken anyway.
• The week runs; you watch. Sends fire on their scheduled dates through the approval gate. The only live decisions are the ones the cut list already anticipated.
• The debrief is the point. Lift over holdout, margin after discounts, and a keep-or-kill card per choice. Next August the board is reopened, not reinvented.

Numbers to hold it against

These are planning numbers, not promises; what moves them is list quality, category, and how honest the deadline is.

• Start the cleanup 8 to 10 weeks out. That is the practitioner norm, and it exists because re-engagement plus sunset plus warmup takes about six weeks to show up in inbox placement.
• BFCM and the holiday season carry 30 to 50 percent of annual revenue for roughly a quarter of stores. That share is the argument for a floor: a discount decision made casually here moves the year's margin, not the week's.
• Early access typically books a fifth to a third of BFCM email revenue before the public doors open, at a lighter discount. If the VIP segment is under a tenth of the engaged list, the definition is too tight to matter; over a third and it is not a reward, it is the list.
• Five touches is the shape, not a law. Fewer and the weekend leans on one send's subject line; more and unsubscribes start paying for the extra revenue. Watch unsubscribes per send, not per week, and let the cut list veto the sixth send somebody proposes on Saturday.
• Hold out 5 to 10 percent of the engaged list. Report lift, and expect it to be smaller than the raw total: it always is, and knowing by how much is what the debrief is for.
• A deadline that is real converts; one that quietly extends trains the list to wait. The Monday close email works exactly once per year per store, and only if it is true.

What has to be true before a card asks for approval

The census. Real counts from the platform, source and date named: engaged in the last 90 days, lapsed, never-opened. "The list is about 40k" is not a census.

The sunset rule. A definition, the number it removes, and the send that gave those addresses a last chance. Approving it should feel slightly uncomfortable; that is what makes it real.

The offer and the floor. One sentence each, plus the margin-scenarios card the floor was read off. The floor names its holder. Exceptions edit the card first.

The cut list. Each line has a reason and an owner. A cut list of platitudes ("stay on brand") protects nothing. A cut list that names the sixth send, the extra code, and the discount on the new collection protects the week.

The five touches. Each one staged, on-voice, one call to action, no placeholder surviving into approval, and a send date on the card. The gift guide sells nothing above the fold.

The holdout. A percentage, a definition, and the exclusion applied in the email platform, not just promised on the board.

The debrief. Lift over holdout, margin after incentives, and keep-or-kill cards specific enough that next year's board can be built from them without archaeology.

The seed cards are the first instance, not the ceiling

The seeds show the shape: a census, a sunset rule, the margin scenarios, the offer, the cut list, a VIP definition, an invite, five touches, one creative, a QA checklist, a holdout, and the debrief pair. A live run is denser in two places. Production usually grows past five emails, because segments diverge: the VIP close and the public close are not the same email with a different code. And the debrief should end with more keep-or-kill cards than feels polite, because they are the only part of the board that appreciates: the next run starts by reading them.

The layout is one arrangement of many

The columns run by countdown slot, because that is the order the thinking must happen in, and the dates are the point. Once the week has run, some stores re-arrange by outcome: a column for what carried the weekend, one for what underperformed, one for next year. Both arrangements survive the one rule; the countdown is the one that keeps September honest.

When this is the wrong playbook

If the list is under a few thousand engaged subscribers, the ten weeks are better spent on the welcome flow and acquisition: BFCM amplifies a list, it does not create one. If margins cannot survive any site-wide discount, keep the board and change the answer: run the gift guide and the early access, and let the offer phase say "nothing site-wide" out loud. And if it is already November, do not run this board at double speed. Run the free BFCM sprint board, do the week as well as the remaining days allow, and put one card at the top of it that says: next year, we start in September.