Hey,

Short one today, and it's a follow-up.

Three weeks ago I told you Q4 gets decided in the summer, and I gave you the creative math. Today's the honest check-in: did you start?

Because the deadline got two weeks closer.

Black Friday is about 15 weeks out. Whatever ads you'll be running that week need to be proven winners by late October, because peak season is the worst possible time to be guessing.

Proving a winner takes 2 to 3 weeks of testing and iteration. Which means your real deadline for shooting new concepts is September, and your testing engine needs to be running at full speed now.

Here's the math again, sharper this time.

Across our accounts, roughly one concept in five becomes a winner worth scaling. The unit is the concept, an actually different angle or format.

A recolored thumbnail on the same script is still the same concept.

So the worksheet is three lines:

1. How many proven concepts do you want running in November? For most brands at your stage, five is the floor.

2. Multiply by five. That's roughly 25 concepts to test.

3. Divide by the weeks until late October, when testing has to be done. From this week, that's a bit over two new concepts every week, no skips.

Run your own numbers.

If you want five winners and you're currently testing one new concept a month, the math says you'll arrive at Black Friday with one proven ad and a lot of hope.

That's how brands end up paying peak CPMs to run experiments in November, which is the most expensive classroom in ecommerce.

The pushback I always get: "we can't produce two concepts a week."

You probably can, once repackaging counts properly. A winning concept re-cut into a new format is a new test at a fraction of a fresh shoot, and a proven hook re-recorded by a different creator is another.

The quota measures testing volume, and testing volume is allowed to be scrappy. The one thing it can't be is postponed, because the shoot has to happen in September either way.

Two quality notes so the quota doesn't turn into noise.

A concept only counts if it could fail on its own. New hook, new angle, new format, or new persona. Five variations of your winner still count as one concept.

And hold the bar at "good enough to stop a scroll." Anything past that is money spent on your own taste.

Paid creative lives a week in a fast feed, and I've watched a raw phone video outperform a $3,000 production too many times to count. I wrote about where the quality bar actually belongs.

The brands that look lucky in Q4 are the ones doing boring reps in August. Two concepts a week from now until October beats any amount of November brilliance.

Put the quota on this week's creative meeting agenda, give every concept an owner and a ship date, and the countdown takes care of itself.

Talk soon,
Andrej

PS: Run the three-line worksheet and reply "quota" with your weekly number. I'll tell you if it's realistic for your setup, and what I'd cut if it's not.