Prime Day 2026 wrapped on June 23 to 26, and the advertising data now coming out of that event tells a different story than the “outspend the competition” playbook most sellers default to. Across the US, brands actually pulled back on ad spend and still converted better than the year before. That’s worth sitting with before Q4 planning starts in earnest.
The Numbers That Moved
According to a CommerceIQ analysis, US ad spend fell 8.8% year over year during the event, while conversion rate climbed 17.1%, reaching a blended 20% conversion rate. A few other figures stand out:
- Product page traffic dropped 10.3% year over year, so fewer shoppers landed on listings overall.
- ROAS held almost flat at 4.78x, versus 4.73x the year before.
- CPC dipped 0.7% year over year on average, but still sat 37.3% above the pre-event baseline, so clicks stayed expensive relative to a normal week even as the year-over-year trend cooled.
- Sponsored Products still carried 88% of event ad spend, though that share slipped 5.6%. Sponsored Display spend fell 70%, and Sponsored Brand spend fell 25%.
Important note: category performance varied sharply underneath those averages. Beauty saw CPC jump 97% year over year with ROAS barely moving (down just 0.5%), while Tools & Home Improvement saw ROAS fall 38.2% against baseline. Furniture was the only category that held positive gross margins through the event.
DSP Took a Bigger Slice of the Budget
Separately, Tinuiti’s Prime Day recap found Amazon DSP spend up 59% year over year, growing to roughly 35% of total Prime Week ad investment. Day 1 spend concentration also eased, dropping to 28% of the event total versus 31% in 2025, with more budget flowing into days three and four. Advertisers weren’t just spending less on search ads, they were shifting more of what they did spend into upper-funnel DSP placements to stay visible across the full event window rather than front-loading everything into the opening day.
Why This Happened
Fewer shoppers hit product pages, but the ones who did were further along and converted at a higher rate, helped by discounts averaging 18.5%, about 16 points deeper than baseline. That combination changes the math on aggressive bidding. Chasing every click made less sense than it did in past years, because a smaller, more qualified pool of shoppers was already primed to buy. Sellers who kept bidding as if traffic volume were flat likely overpaid for clicks from browsers who were never going to convert, while sellers who tightened targeting caught the shoppers who were ready.
What Sellers Should Take Into Q4 Planning
1. Rebuild bids around efficiency, not volume
With traffic down and conversion up, the campaigns worth funding are the ones already converting, not the ones with the most impressions. Trim spend on keywords that are pulling clicks without sales, and reallocate toward proven performers rather than raising budgets across the board.
2. Stop front-loading Day 1
Pacing data shows spend spreading more evenly across multi-day events now. Build BFCM and holiday campaigns to run a day or two past the official event window instead of concentrating budget on the opening day.
3. Test a DSP allocation, even at modest scale
DSP’s share of spend jumping to over a third of Prime Week investment signals where larger advertisers see the best return on visibility right now. A small retargeting or audience-building test ahead of BFCM is worth running even for mid-size accounts that haven’t touched DSP before.
4. Check your category’s CPC-to-ROAS relationship before scaling anything
If your category looks more like Tools & Home Improvement than Beauty, protect margin with tighter TACOS-based bid caps rather than chasing share on rising clicks.
The clearest lesson from Prime Day 2026 is that budget size mattered less than where that budget landed. That’s a cheap adjustment to make now, while the data is still fresh and there’s a full quarter to apply it before the next big event. This is exactly the kind of shift Mazescale builds into client accounts every quarter.
