Black Friday falls on 27 November this year and Cyber Monday on the 30th. From today that is five weeks and a couple of days.
Every year I get the same two calls. The first comes in late October from someone who wants to rebuild their account before the sale. The second comes around 24 November from someone whose cost per purchase has tripled overnight and wants to know what to change. The first call is useful. The second one almost never is, because by then the only honest answer is to stop changing things and ride it out.
What follows is the order I would work in if I were preparing an account today.
What actually goes wrong in Q4
It is not that traffic dries up. It is that traffic gets expensive.
Between the first week of November and Cyber Monday, auction prices in most consumer categories rise somewhere between thirty and seventy percent against the October baseline. Every advertiser with a seasonal budget is bidding for the same attention at the same time, and the platforms are simply clearing the market.
That price rise does not break a healthy account. What it does is remove the margin that was quietly absorbing every weakness the account already had. A tracking gap that cost you a little efficiency in September costs you real money in November. A checkout that leaks a tenth of its sessions leaks the same proportion of much more expensive sessions. A retargeting pool that is too small to matter is still too small, except now you are paying peak prices for cold traffic because there is nothing warm to talk to.
So Q4 preparation is not really about Q4. It is about removing the weaknesses you can still afford to fix.
Weeks one and two: make the numbers true
Start with measurement, because everything after this depends on it and because it is the only part that gets harder to fix as you get closer.
Check that every purchase event fires once and only once. Duplicate events are the most common cause of a Black Friday report that looks spectacular and a bank balance that does not move. Compare the last fourteen days of platform-reported purchases against what your store actually recorded. If the two are more than about fifteen percent apart, find out why before you spend another pound.
Check your server-side setup properly. On Meta, open Events Manager and look at the event match quality score for your purchase event. Anything below about seven out of ten means you are sending weak customer data, and weak matching gets punished hardest exactly when competition is highest. Make sure email, phone, name and location are all being passed where you have them.
Decide your attribution window now and write it down. If you switch from seven-day click to one-day click in the middle of the sale because the numbers look bad on the Saturday, you will have no way to compare anything to anything. Pick the window, record what a normal week looks like inside it, and leave it alone.
Finally, work out the number you are actually optimising towards. Not last year's return on ad spend. The break-even figure for this year, with this year's cost of goods, this year's shipping rates and whatever discount you are about to offer. A store running a thirty percent sitewide discount needs a materially higher return than one running ten percent off a single range, and plenty of people do not do that arithmetic until January.
I wrote about the general version of this problem in why three platforms report three different numbers, and the reconciliation method there is the one to use here.
Week three: decide the offer before you make the creative
The offer does more work than the creative in this period. A good ad for a vague promotion loses to a plain ad for a clear one, every time.
Three questions settle it. How deep is the discount, what is it on, and when does it run?
Sitewide percentages are simple to communicate and brutal on margin. Discounting a hero range or a bundle protects the margin but needs more explaining. Free shipping over a threshold is often the cheapest offer you can make, because it removes the single biggest reason people abandon a basket without touching product pricing at all.
Pick one, then build everything around it. The offer should be legible in the first second of the video, in the first line of the headline, on the product page, in the cart, and in the email. If a visitor has to work out what the deal is, you have already paid for them twice.
Timing matters more than most people expect. The window that has grown fastest in recent years is the one before the weekend, not the weekend itself. Opening on about the 20th gets you cheaper auction prices and a head start on inventory, and it gives your retargeting something to retarget before everyone else floods the feed.
Week four: build the pool you will sell to
This is the part that gets skipped and it is the part that pays.
The audience you convert on Black Friday is mostly assembled in the weeks before it. Spend some of your budget now on reaching cold audiences with content that does not ask for a sale: product demonstrations, comparisons, the thing you are actually good at. Those viewers become a retargeting pool that is cheap to reach when the expensive window opens.
Grow the email and SMS list in the same period, with a sign-up offer tied to early access rather than a generic discount. Early access is worth more than money off to the people most likely to buy, and it costs you nothing.
While you are at it, check that the retargeting audiences you are relying on are actually populated. A three-day window on a site with modest traffic will be too small to deliver. Widen to fourteen or thirty days, segment by behaviour rather than by recency alone, and keep the people who added to cart separate from the people who read a blog post, because they need different messages.
Week five: freeze
From roughly the 13th of November, stop restructuring.
No new campaign architecture. No migration to a new bidding strategy. No switching the optimisation event. No installing a new tracking tool. Every one of those resets the learning the account has accumulated, and the learning phase in a high-cost auction is the most expensive place to be.
Budget increases are the exception, and they have a rhythm. Raise an existing campaign by around twenty percent at a time and let it settle for a day, or duplicate the structure into a separate higher-budget campaign and let the original keep running. Tripling a budget overnight on the 26th is the single most reliable way to turn a profitable account into an unprofitable one in twenty-four hours.
If the account genuinely needs restructuring, do it now, this week, and accept the dip. There is still time for it to repay. In three weeks there will not be. The account architecture I would restructure toward is in how to structure a Meta Ads account that actually scales.
What to leave alone entirely
The checkout. Do not redesign, re-platform or add a new payment provider in November. If something there is broken, fix that one thing and test it on a phone yourself, twice.
The landing page, on the day. Build and test the sale page in week three or four, not the night before. If you want it to beat your homepage, it needs traffic through it before the stakes are high — the reasoning is in why paid traffic should rarely land on your homepage.
The optimisation event. If you are optimising for purchases and volume supports it, stay there. Moving to add-to-cart for the sale because it feels safer gives the algorithm a cheaper signal to chase and it will happily chase it.
Your nerve, on the Saturday. Mid-sale, daily numbers swing violently and mean very little. Judge on a rolling three-day view, not on what you see at eleven in the morning.
The week after is where the profit is
Most stores treat 1 December as the end. It is the most valuable moment of the whole period.
You have just bought a large number of first-time customers at peak prices. Whether that was a good trade depends entirely on whether they buy again, and that is decided in the following fortnight by your post-purchase flow, not by your ads. If those emails are not built, the discount you just gave away was pure cost. That is the case I make in full in why empty email flows mean paying for the same customer twice.
Plan the December follow-up now, while you have the attention to do it properly.
The short version
Fix measurement in the next fortnight. Settle the offer and build creative around it. Spend November filling the retargeting pool and the email list. Freeze the structure from the middle of the month and scale budgets in steps. Leave the checkout alone. Then spend the first half of December turning new buyers into repeat ones.
Nothing on that list is clever. It is just all of it done in the right order, early enough to matter.
If you want a second pair of eyes on an account before the window opens, the ad account audit is a fixed-price review of structure, tracking and creative, and the free growth audit gives you the three biggest problems we can see in writing. Either one is more useful now than it will be on 25 November. Our Meta Ads and Google Ads work runs to the same plan.