Prime Day 2026 runs June 23 to 26.
Amazon moved the whole event up to June this year after nearly a decade of July dates.
My first job inside marketing was actually Amazon-specific. So I have a lot of Prime Day experience.
My biggest learning? the brands that win Prime Day built their demand months ago, and you can’t change that now. The job this weekend isn’t to build a huge Prime Day. It’s to not mess it up and capture what you built in the past.
Start With What You’re NOT Doing
The fastest way to blow the next four days is trying something new you’ve never tested on huge spend levels.
No new creative concepts. You don’t have time to test, and running untested concepts on the highest-traffic days of the summer is a good way to end up with an unprofitable ROAS.
No optimization strategies you’ve never run. Prime Day is not the moment to try click campaigns, new bidding setups, or anything you haven’t already validated at normal spend.
Friday: Three Decisions (About 2 Hours)
Decision 1: Which channel gets the better deal?
If your site evergreen promo beats your Amazon deal, you split conversions and confuse buyers. If Amazon undercuts your site, your DTC offer looks weak all week.
Or, third, run a “prime time” sale on your site matching the Amazon offer.
Most brands let both channels run whatever was already planned and find out Tuesday that they’re competing with themselves.
Pick a winner on purpose. If your site unit economics are better, give a better site offer (just make it a GWP or offer bar so it's not scrapeable, as long as it's not identical SKUs, Amazon can’t match it back). If you are trying to build Amazon reviews, give a better Amazon offer.
Decision 2: Spend to your inventory, not your ambition.
It’s too late to inbound more FBA units. Count what’s actually sellable and pace your spend against that number.
If units are tight, spend steadily across four days instead of blowing the budget on day one and going dark.
Decision 3: Raise your budgets now.
Prime Day traffic will drain a normal daily budget by mid-morning. Then your ads sit dark during the highest-intent hours.
Raise Amazon campaign caps today. If you’re pushing Meta traffic to Amazon, decide that number today too. That way you have a plan and don’t panic-adjust.
The Weekend: Creative and Email (About 3 Hours)
Badge your winners.
Take your top 3 to 5 ads and put a Prime Day banner on them. That’s it. Your winners already work, they just need a deal flag.
Then knock out a handful of quick Prime Day statics. Product shot, deal callout, urgency line. Make the offer and the product clear.
Don’t edit your live winners, because editing resets learning on the ads currently working. These should be net new ads set up, then paused after Prime Day.
Segment your list
An email to everyone can send your best customers to Amazon, where you lose the subscription, the data, and the margin.
Example segmentation:
Active subscribers & recent buyers never see a Prime Day email. They already pay you regularly. Leave them alone.
Lapsed customers (no recent purchase, not subscribed) get sent to Amazon. They weren’t coming back to your site anyway. A Prime Day purchase from a dead segment is found money.
Then schedule everything now: a launch-day send, a mid-event nudge, a last-chance email Friday. Write the post-event “missed the deal” email this weekend too, scheduled for Saturday the 27th.
Fix your links.
One thing on every ad and email pointing to Amazon.
First, attribution links, so you collect Amazon’s 10% Brand Referral Bonus. It’s free margin so claim it.
Start the push Saturday, not Tuesday.
Amazon Attribution counts conversions for 14 days after the click. Traffic you send this weekend still gets credited when those people convert at deal prices on Tuesday.
Weekend clicks are also cheaper than event-day clicks, because every brand in your category piles into the auction Tuesday morning. So if you’re pushing Meta traffic to Amazon, start Saturday. Not at full budget levels, but a small “get ready” test is worthwhile.
Break it into its own campaign so you can measure it post-window and decide if it's worth increasing or cutting next year.
Monday: Keep the gas on.
The instinct says Prime Day steals from your site, so pull back DTC spend and let Amazon have the week. I recommend the opposite.
Buying intent rises everywhere during Prime Day. People are in shopping mode, and site conversion often lifts during the event even when you’re not pushing traffic there. Pausing your regular Meta campaigns means pulling out of the market during one of the biggest shopping windows of the summer, and resetting learning on the way out.
Leave your DTC campaigns running.
Then verify everything actually renders.
Monday is for checking. The coupon clips. The strikethrough price shows. Subscribe & Save settings are what you think they are. Redirects resolve to the right listings. Scheduled emails point at the right segments.
And check what the final stacked price actually is. A Prime Day coupon that stacks with your Subscribe & Save discount can put you at 40%+ off without anyone deciding that. Check for yourself.
And set your day-of cadence.
Pick three check-in windows for Tuesday: a morning budget check, a midday pacing look, an evening scale-or-kill call. Decide the schedule Monday so Tuesday is execution instead of refreshing dashboards all day.
The Whole Thing on One Page
Friday: pick which channel wins (or run a parallel offer on purpose), pace spend to inventory, raise budgets.
Weekend: badge your top ads, quick statics launched as new ads, segment the list (protect subscribers and recent buyers, push 90+ day lapsed to Amazon), schedule every send including post-event, attribution links and redirects on everything, and start the Meta-to-Amazon push Saturday to load the 14-day attribution window.
Monday: leave DTC running, verify everything renders including the final stacked price, set the day-of cadence.
Tuesday: 12:05 a.m. check, then execute.
The brands that win Prime Day built their demand in March and April. This is just the final checklist.
Did I miss anything? Hit reply and let me know!