The date on your plan is 27 November. The date that actually matters is around 5 October.
That is roughly the last point at which you can launch warm-up campaigns and have them out of the learning phase, with three or four weeks of real creative data, before auction prices go up. Published guidance puts it about there, and it is a week away.
Everything after that is more expensive, including your mistakes.
Why early October and not early November
The cost curve does the arguing here.
Meta CPMs in Q4 run something like 35% to 45% above the annual average, with November on its own around 41% higher. On the peak days, Black Friday and Cyber Monday, they can reach two to three times the annual baseline. Ecommerce specifically saw roughly a 65% jump from October to November last year, then about a 42% correction into December.
Google is not gentler. Search auction pressure peaks when every retailer and every finance advertiser is bidding on the same intent, and reported Google Ads CPMs during last year’s Black Friday and Cyber Monday period were above $20 and up about 10% year on year.
So a broken pixel discovered on 20 November costs you several times what the same broken pixel would have cost on 1 October. Same error, different price, because you are now paying peak rates to learn it.
Q4 is not the quarter for finding things out. It is the quarter for executing things you already tested.
The pre-flight list, this week
None of this is glamorous and all of it is cheaper now than later.
Verify the tracking end to end, as a customer. Buy something. Watch the conversion arrive in every platform, with the right value attached. Then reconcile what the platforms report against each other so you know your baseline gap before the volume arrives, because a gap you discover in December is indistinguishable from a tracking failure.
Check the product feed properly. Disapprovals, missing GTINs, prices that disagree with your site, out of stock items still serving. A feed error in Q4 is a budget that spends on items you cannot ship.
Test the pages that will take the traffic, on a phone, on mobile data. Not your laptop. The landing pages and the checkout, at the weight they will actually be once the promotional banners and the countdown timers are installed.
Build the email flows now. Abandoned checkout, browse abandonment and post purchase are doing a large share of the work in November and they cannot be built in the week they are needed.
Count your creative. Work out how many distinct assets you will need across the whole period rather than discovering on 20 November that you have three and they are fatigued. Then produce them in October when your designer has time and your costs are normal.
The freeze, and what breaks it
This is the part that separates a calm November from an expensive one.
Meaningful edits reset the learning phase. Budget changes beyond roughly 20%, bid strategy swaps, adding new creative into a live ad set, targeting changes. Which means the instinct to fix something on Black Friday morning is the single most expensive instinct in the quarter.
So pick a date, write it down, and tell everybody. From that day, no structural changes. If you need new creative during the peak, it goes into new ad sets rather than into the ones already performing.
Two weeks before the peak is a reasonable freeze date. Have the offer campaigns live one to two weeks ahead so they exit learning before prices rise, then leave them alone and let the bidding do what you are paying it to do.
What you are still allowed to do: pause obvious failures, raise budgets in increments small enough not to trigger a reset, and answer the phone.
Pace against the days that convert
A flat daily budget across November is an averaging error dressed up as discipline.
The period has a shape, and your spending should match it rather than the calendar. There is a pre-peak window where costs are still reasonable and intent is building. There is the Black Friday to Cyber Monday concentration. There is a genuine lull in the first half of December. Then a second surge running into the last guaranteed shipping date, which for many retailers converts better than Black Friday because the buyer has run out of time and stopped comparing. And then the post-Christmas window, which is gift cards and self-gifting and is the cheapest traffic of the whole quarter.
Decide the split across those five windows now, in a spreadsheet, while you are calm. Whatever you allocate on the day will be driven by whatever the dashboard did yesterday, which is the worst available input.
And if your total budget is modest, concentrate rather than spread. The floor logic gets harsher in Q4, not easier, because you are competing for the same impressions as people spending fifty times more.
Decide January in October
The decisions nobody makes in advance, and they are all easier to make now.
What happens to the discount on 2 January. If you do not decide, it stays, and you have taught your customers to wait for it. Write the end date into the campaign now.
What you do with the customers you acquired. Somebody who bought once at 30% off is not a customer yet. The second purchase decides that, and the flow that produces it should be built in October rather than improvised in February.
Your January budget. Media gets significantly cheaper after the December correction, and January is usually the best value of the year for acquisition. Most retailers cut spend in January out of exhaustion, precisely when the cost per acquisition is at its most favorable.
The retrospective date. Put a meeting in the calendar for the second week of January, now, while you still remember what you were worried about. In February nobody can reconstruct what actually happened.
Before lunch
An hour, three things.
Put the freeze date in the shared calendar and tell whoever might edit the account.
Then place a test order on your own site and confirm the conversion arrives everywhere with the right value.
Then open the product feed and clear every disapproval, because that is the failure that quietly caps everything above it.
Then, this week rather than this morning, get the warm-up campaigns live. The window on that closes in a few days, and it does not reopen.
My notes on performance marketing cover the rest of the mechanics. If you’d like a second pair of eyes on your own Q4 plan, drop me a line on email, WhatsApp or LinkedIn and we can have a quick chat. I’m contracted full time so this isn’t a pitch.