Q4 is won in September
My worst Q4 ever, I did everything right, in November. Found good products, shot good creative, structured the campaigns properly. And got flattened, because I was testing products at $28 CPMs against competitors who'd finished testing in October at $9 CPMs and were scaling proven winners while I burned money learning. That difference isn't skill. It's calendar.
2 to 3x
CPM inflation from early October to Black Friday week
6 wks
Lead the Heartbeat curves give you before each spike
80%
Of my annual profit that comes from Oct to Dec, most years
The whole playbook rests on one idea: separate the testing phase from the scaling phase, and put the expensive one (testing) in the cheap months.
The month-by-month timeline
- 1
August: build the watch list
Pull last year's 24-month curves for gift-adjacent categories and find the Heartbeat shapes, products that spike every Q4 like clockwork. Last year's spike is this year's forecast. I keep 30 to 40 candidates on a Majorka watchlist and let the alerts tell me when curves start bending.
- 2
September: test on cheap traffic
CPMs are at their annual low while everyone else is asleep. Run your normal testing process on the watch list. A product that’s marginally profitable at September CPMs prints money at Q4 volumes, you’re not testing for now, you’re testing for November.
- 3
October: lock and stock
Kill the losers without mercy, pick your 2 to 4 winners, and settle supply: confirm your supplier's real capacity, their holiday cutoff dates, and a backup source for each winner. The Q4 killer isn't ads, it's a supplier who stops answering in late November.
- 4
November: scale, don’t explore
From here you touch nothing unproven. Raise budgets on winners, refresh creative angles on the same products, and watch delivery dates like a hawk. Every dollar spent testing in November is a dollar donated to the auction.
- 5
December: ride it, then pivot to the second spike
Hard stop on standard shipping around Dec 10 to 14 depending on your lanes, after that, sell gift cards or switch angles to 'treat yourself.' January self-improvement products start their curves in the last week of December. The season doesn't end; it rotates.
The three mistakes that repeat every year
- Testing into Black Friday week because 'traffic is high anyway.' Traffic is high and it's the most expensive learning environment of the year. You're paying triple tuition for the same lesson.
- Scaling a winner past supplier capacity. Getting 400 orders you can’t ship is worse than getting 200 you can, refund waves in December kill payment processors’ trust in you, and that damage outlives the season.
- Ignoring the post-Christmas window. Dec 26 to mid-January has crashed CPMs and everyone's holding gift cards and resolutions. Some of my easiest profit ever came from the two weeks everyone else takes off.
Set watchlist alerts now
Whatever month you're reading this, the next seasonal spike is already forming in some category's curve. Add candidates to your watchlist and let velocity alerts do the calendar-watching for you.
“Amateurs feel the Q4 energy in November. Professionals felt it in August, in the data.”
Written by
MaximusFounder of Majorka
Built Majorka after years of running his own stores and being tired of guessing. Every post here comes from real tests, real ad spend, and real mistakes he paid for so you do not have to.
Share article
Put this playbook to work in 30 seconds
Paste any AliExpress link and see the order curve, velocity, and verdict for yourself. Plans from $39/mo.
See it in Majorka4.9/5 from 1,632 sellers
Orders · 24 wks
Scale23,491
