At Nord Media, we focus on profitable growth at scale, we help forecast revenue, and contribution margin goals, then build the marketing plans to execute those goals.
This means full clarity on testing, scaling and improving profitability at every aspect of your funnel.
If you want a growth partner aligned with your real business goals, let’s chat
It’s 9am on Black Friday. Sales from your Meta ads look slow for the first hour, so someone on your team swaps in a new ad and raises the budget.
It takes thirty seconds, and those two changes just reset the ad set.
Meta now treats it like it just launched and has to relearn who to show it to, so costs jump around until it collects about 50 sales over 7 days.
Those 7 days land right on the most expensive week of the year.
My working assumption is that the smartest move in late November is to stop making big changes, and to pick the date you stop ahead of time.
Every ecommerce operator I know has a swipe file full of PDP, carts and checkout screenshots they swear they’ll rebuild "someday".
But those designs rarely see the light of day. Why? Because "someday" usually means writing a dev ticket, waiting on a two-week sprint, and praying you don't break your theme code right before Q4.
I’ve been playing around with Rokt Aftersell's new Aftersell Cart, and it completely bypasses that bottleneck.
It actually uses AI to rebuild a cart directly from your screenshot.
The workflow is wild: you take an image of a cart you like, drop it into your LLM with their specific prompt, and import the generated file. Your new cart is live in under five minutes. It requires zero code and doesn't touch your theme files at all.
Once your layout is built, you can turn it into a revenue engine:
Grow the order: Drag and drop 18+ different widgets like one-click upsells, cross-sells, and rewards progress bars.
Personalize the funnel: Set up logic to show specific offers based on customer tags, subtotals, or what’s already in the cart.
If you want to see exactly how this looks in production, the team is doing a live masterclass teardown with Grüns on November 5 at 1 PM ET.
Plus, they are giving away 3 high-converting cart templates to live attendees that you can apply to your store — highly recommend tuning in if you want to optimize your cart before the Q4 rush. RSVP here
Each Big Edit Restarts The Clock
An ad set enters the learning phase when you first launch it, and again after any significant edit.
Meta says an ad set needs about 50 sales (or whatever result you told Meta to aim for) within 7 days of the last significant edit to exit the learning phase.
That word “last” is the catch. A new significant edit starts the 7 days over.
Make three significant edits in one week and the ad set can spend the whole week relearning instead of delivering at its settled cost.
Note that if an ad set can’t reach 50 sales in a week, Meta labels it to Learning Limited.
You don’t need to fix an ad set just because it has that label.
The real risk here is the reset. Each one brings a few days of unstable costs, and late November is when those days cost you the most.
Before your next change, open Ads Manager, look at the Delivery column on the campaigns you plan to scale in November, and note which ad sets still say Learning.
What Actually Resets Learning
Meta publishes its own list of significant edits that reset learning:
1. Any change to targeting
2. Any change to ad creative
3. Changing the optimization event
4. Adding a new ad to the ad set
5. Pausing for 7 days or longer, then restarting
6. Changing bid strategy
Budget changes are the gray area.
A small bump won’t reset learning, but a big jump can, and Meta doesn’t say where the line is.
Its only example: raising a budget from $100 to $101 probably won’t reset it, while raising it from $100 to $1,000 might.
That’s why you’ll see a “20% rule” in a lot of guides.
Meta never published it. Buyers came up with it as a safe habit, and it’s fine to use that way.
Sources also disagree on whether small creative fixes, like fixing a typo, are safe.
Meta’s list says any change to the ad itself resets learning, so that’s the rule I follow.
Lastly, check your tools. If an automated rule or AI agent makes one of these changes, Meta resets learning just as if a person did it.
Know which tools can edit your campaigns before Black Friday.
Your Freeze Date Is…
Pick a date to freeze your ads.
> Make your last big change, like a new ad, a new audience, a new campaign goal or a new bidding setup, by November 17 to 19.
That gives each ad set a full week to settle before the rush. From then on, the ads stay as they are.
> Until then, save up your changes and make them all in one sitting each week.
Four changes spread across a week reset learning four times.
Four changes made together reset it once.
> When you raise a budget, do it in small steps, because one big jump can reset learning on its own.
> Next, check whether each ad set has enough budget to finish learning at all.
It needs about 50 sales in 7 days, so the daily budget has to cover that:
Daily budget needed = (cost per sale × 50) ÷ 7
Example (illustrative):
Cost per sale = $60
$60 × 50 = $3,000/week
$3,000 ÷ 7 = $429/day
If one of your ad sets comes up short, you have two options before November 17.
1. Raise its daily budget to the number from the math
2. Combine it with a similar ad set so their sales count together
During the freeze, you can still rename ads, adjust schedules, and nudge budgets by small amounts.
Write the Freeze Rule Down
Here’s what I’d do this week:
1. Put November 17 on the team calendar next to your email send dates.
2. Pick one day a week for ad changes until then.
3. Run the daily budget math on each ad set you plan to scale in November.
4. Write one rule for the team: after November 17, nobody changes a scaled ad set without a named person signing off.
I’d push hardest on that last one.
The 9am Black Friday change usually comes from someone trying to help.
If sales look slow that morning, wait for the afternoon numbers, write down what you’d change and save it for the week after Cyber Monday.
If you only do one thing, put November 17 on the calendar today.
If you want a growth partner aligned with your real business goals, let’s chat.
I appreciate you reading and as always if there's anything I can do to help you in anyway just respond to this email. I personally respond to every email I get.
- Kody ✌️
