Every brand we take over in August has the same two dates on the wall. The FBA inbound deadline and the deal submission deadline. Both are real, both are published, both are on a Seller Central page with Amazon’s name on it, and every operator we speak to can recite them.
Nobody has a creative deadline, because Amazon has never published one. There is no notification, no banner, no cutoff after which you can’t change an image. You can swap your hero on December 18th if you want to. The platform will let you.
That’s precisely why it’s the one that gets missed. The Q4 creative deadline is real, it’s arithmetic rather than policy, and for most brands it lands in the second week of September. Which is four weeks from today.
Three Q4 deadlines, and only two of them announce themselves
Amazon opened holiday deal submissions on July 8 this year. Prime Big Deal Days submissions close September 8, with Black Friday Week and Cyber Monday submissions closing October 20. On the inventory side, FBA shipments for PBDD need to arrive by September 16 if you’re on Amazon-optimized shipment splits (September 9 on minimal splits), with the Black Friday cutoff at October 28.
Amazon has said PBDD returns at the same time as last year, which was October 7-8. Treat that as the working assumption and confirm the dated version in your own Seller Central notifications rather than off any blog, including this one.
Those two deadlines behave like deadlines. They have a date, a consequence, and an email. Miss the inbound window and your units sit in a receiving queue while the event runs. Miss the deal window and you’re not in the event.
The creative deadline has none of that. Miss it and nothing happens. Your images just go into the biggest traffic window of the year without ever having been measured. That’s not a failure that generates an alert. It’s a failure that generates a quiet feeling in January that peak was “rough this year.”
The arithmetic that sets the date
Creative results don’t arrive when you ship them. They arrive when the metric they move has accumulated enough data to be legible, and those windows are not negotiable:
- CTR reads in 2-3 weeks. It’s a top-of-funnel signal with high volume, so it stabilizes fastest.
- CVR reads in 30-60 days. Fewer events, more noise, longer window.
- Return rate reads in 60-90 days. Because the buyer has to receive the thing, live with it, and decide.
Now work backwards from an early-October event.
A creative change that goes live September 12 gives you roughly three weeks of CTR data before PBDD. That’s enough to know whether the thumbnail is winning or losing before the year’s most expensive traffic hits it. It is not enough to know what happened to conversion.
A change that goes live September 25 gives you ten days. Ten days is not a read, it’s a glance. You’ll have a number, it will move, and you will have no idea whether the movement is the change or the fact that the whole category’s traffic mix started shifting toward holiday intent.
A change that goes live October 3 is untested creative running on your highest-value traffic of the year. Sometimes that’s fine. It’s still a decision that nobody in the business consciously made.
So: if you want a CTR read before peak, your creative goes live by roughly September 12. If you want a CVR read, it needed to be live around now. That’s the deadline. It’s not a policy, it’s a measurement window, which is exactly why nobody circulates it.
The dependency almost everyone forgets: the approval queue
Going live isn’t the same as being finished, and A+ content is where this bites.
Practitioner consensus puts standard A+ review at 7-10 business days, with peak-season queues extending it — reports of 14 business days in Q4 are common, and a rejection restarts the clock rather than pausing it. Amazon doesn’t publish a guaranteed SLA, so treat these as working ranges from the field rather than a commitment, and build buffer accordingly.
Two consequences that reorder your production schedule:
A+ submitted the same week you want it live is a coin flip. If your September 12 target includes A+ changes, those files needed to be in the queue in late August, and they needed to be clean, because one rejection on a resubmit puts you into October.
Batch submissions during peak are a bad idea for a second reason. If you’re pushing A+ across a catalog on a shared template and something in that template trips a compliance category, you don’t get one rejection, you get a batch of them, in the queue that’s already running long. Run the compliance read on the template before it multiplies, not after.
Images are faster — they process in hours, not days — but “faster” isn’t “instant,” and a hero that renders on your monitor and hasn’t reprocessed on Amazon’s side yet is not live.
What still ships after the line
A creative freeze isn’t a prohibition on touching anything. It’s a distinction between changes that change what’s being measured and changes that fix something broken, and the second list is longer than most brands think.
Ship these in October without hesitation:
- A hero that fails the 85% frame-fill standard and can be re-cropped from the file you already own. You’re adding pixels to a thumbnail, not changing an argument.
- Any image under 1,600px on the longest side that isn’t engaging zoom properly. That’s a file problem, not a creative decision.
- Fine print on an image that is factually wrong — a count, a size, a certification that changed. Wrong facts on a listing during peak are a return machine.
- An empty Item Highlights field on a revenue driver. That’s a blank line on your search card, and search cards changed shape last week.
- A missing or wrong compatibility statement. Every day it stays wrong, it’s generating returns you’ll be paying for in January.
Do not ship these after the line:
- A rebuilt hero on a top-20 ASIN. You will never know whether it worked, because you’ll be measuring it against traffic that has nothing in common with September.
- A resequenced image stack. Same problem, larger blast radius.
- An A+ rebuild that has to clear the queue.
- Any A/B test. More on this below, because it’s the one that costs the most.
The sorting rule: if the change fixes something that is objectively wrong, ship it. If the change is a bet you’d want to grade, it’s too late to place it.
Do not run creative tests into peak
We say this every year and every year somebody runs a Manage Your Experiments test through Black Friday, gets a decisive-looking winner, and rolls it across the catalog in January.
Peak traffic is not your traffic. The mix shifts toward deal-seekers and toward people buying for someone else, both of which behave differently at the thumbnail and at the buy box than your baseline shopper. A creative variant that wins on that traffic has won a contest that isn’t running for the other ten months of the year. Roll it out in January and you’ve made a permanent decision on a temporary population.
There’s a second contamination problem specific to this year. Something like 984 million product titles have been updated since Amazon’s title changes landed, and Item Highlights started rendering as its own line on desktop search cards this month. Every competitor in your comparison set re-rendered inside the same window you’d be testing across. Any CTR read that spans late July to now is a number produced by the entire shelf moving at once.
The practical implication: pull a fresh baseline this week — organic rank on your top ten keywords, 30-day CTR and CVR by ASIN, screenshots of your search cards and your competitors’ — and treat it as day zero. You cannot grade a Q4 creative decision against a memory, and you can’t grade it against a pre-July baseline either.
The calendar
This week (Aug 11-15). Pull the baseline. Screenshot search results on a phone for your top five revenue ASINs. Pull CTR and CVR by ASIN for the last 30 days. Note which listings have empty Item Highlights. This is the file everything else gets measured against.
Now through Aug 29. Do the real creative work. This is your last comfortable window. Prioritize by contribution, not revenue, and promote any SKU with a compatibility, sizing, or count dimension a tier — a wrong expectation during peak costs you the contribution, the fees both directions, frequently the unit, and a return rate reading that follows you into next year.
By Aug 29. A+ submissions in the queue. Clean, compliance-read at the template level, with buffer for one rejection.
By Sept 12. Everything live. Images processed, A+ approved and rendering, Item Highlights populated on the top 20. After this date, changes are fixes only.
Sept 12 through the end of peak. Freeze on creative bets. Weekly ten-minute diff on your top 20 to catch changes you didn’t make — auto-published listing edits run on a 14-day clock that is shorter than most brands’ review cadence, and peak is exactly when nobody’s looking.
January. Unfreeze. Grade what shipped in September against the August baseline, not against November. Restart testing on a normal traffic mix.
FAQ
Is there an actual Amazon deadline for changing images before Q4?
No. Amazon publishes inventory and deal deadlines, not creative ones. The date in this post comes from measurement windows — how long CTR and CVR take to become readable — not from policy. That’s why it gets missed.
What if my creative is genuinely bad and I only just noticed?
Then fix it, and accept that you’re shipping an unmeasured change into peak. A hero that’s clearly failing is worse than an untested replacement. Just be honest in the file about what you did and when, so January’s post-mortem has something to work with.
How long does A+ content take to approve?
Field reports cluster around 7-10 business days, extending in Q4, with rejections restarting rather than pausing the clock. Amazon doesn’t publish a guaranteed turnaround, so build three weeks of buffer before any date you actually care about.
Can I run a test during Prime Big Deal Days if I use Manage Your Experiments?
You can. You shouldn’t. The tooling is fine; the traffic isn’t representative. A winner declared on deal-event traffic is a statement about deal-event shoppers, and you’ll be rolling it out to a completely different population.
What’s the single highest-value creative change to make in the next four weeks?
For most catalogs: the frame that prevents your most common return. Pull FBA Customer Returns for your top revenue SKU, read the buyer comments on the not-as-described and not-as-expected codes, and find the pattern that appears three or more times. Then check whether any image in your stack was supposed to prevent it. Usually there isn’t one, and that’s your shot list.
—
If you’re looking for a team that manages every lever — creative, advertising, and operations — Velocity Sellers works with brands doing $100K+/month on Amazon. Contact us for a free account audit.
