Prime Day and the July Slump: Why DTC Underperforms and How to Prepare

Prime Week keeps getting longer, and it keeps getting harder for DTC brands. Amazon posts record order volume — up 99% year over year in 2025 — while most independent stores see steep drops in traffic, conversion, and return on ad spend.

For accounts that had been growing 20% to 30% year over year, July went flat.

It is not only the algorithm and it is not only Prime. It is consumer behavior, and it is predictable enough to plan around.

The slump is real and it compounds

After a strong Fourth of July, most brands hit the usual post-sale slump. Prime Week stacks on top of it.

If you are not running deals on Amazon — or worse, if your ads send people to your own site while Amazon offers the same thing faster and cheaper — cart abandonment rises, ROAS falls, and July flatlines. The increase in cart drop-offs and Meta inefficiency is not theoretical. It is a direct consequence of an alternative purchase path being temporarily better than yours.

The modern buyer journey makes it worse

Consumer behavior is fragmented. People scroll Reels, see a product, screenshot it, then search for it later. Or ask an AI assistant. Or wait for Prime Day. Or forget entirely.

Click-through sales are vanishing, Gen Z especially. Even Meta's conversion-optimized campaigns concentrate delivery on roughly the 20% of users most likely to buy immediately — a tight auction where CPMs climb and performance drops while everyone competes for the same small pool.

The brands handling it well are deliberately stepping outside that pool: add-to-cart objectives, reminder ads, and mid-funnel awareness campaigns that build demand rather than harvesting it. The measurement framing for this is in attribution after last click.

Reminder ads are the most underused tactic on Meta

Reminder ads sit inside Meta's engagement objective and let users opt in to be notified about an upcoming sale. Users get notifications before, during, and after the event without ever joining an SMS or email list.

They are cheap, they are effective, and almost nobody runs them.

In accounts where weekend promos had underperformed, running reminder ads Monday through Friday brought forecasts in — sometimes exceeded — by Saturday. The reason is straightforward: you are not competing for attention during the sale, because you already earned it beforehand.

Meta's own research points the same direction, with users exposed to reminder ads ahead of a drop meaningfully more likely to purchase when it goes live. It is the closest thing to a free lever in the platform, and it builds a retargetable audience without needing a list.

Use the slump as a testing window

July is slow. That makes it the right month to test, provided you do it without damaging what already works.

• Create separate campaigns for sale creative. Do not run promotional tests through your business-as-usual campaigns.
• Keep BAU clean and uninterrupted, so it stays scalable through the back half of the year.
• Treat Memorial Day and the Fourth of July as rehearsals for Q4 structure, not just revenue events.

And test more than creative. Test structure, bid types, event objectives, and new-visitor percentage — the things that determine whether the account can absorb Q4 budget. Budget scheduling tools that let you concentrate spend into timed spikes are worth learning before the moment you need them.

Attribution needs more than one lens

Every year brings another erosion of click-based measurement. Platform changes that strip or limit click identifiers keep arriving, and each one makes single-model attribution more fragile.

The durable answer is a three-pronged measurement model rather than a single source:

1. Click-based
2. View-based
3. Platform-based

Relying only on last click means scaling down your demand-creation channels precisely because they do not convert immediately — which is the most expensive mistake available heading into Q4.

Liquidity is the lever most brands are not pulling

It is easy to talk about testing and scaling. Doing it takes capital.

Smaller brands tend to focus on revenue rather than capital deployment, but performance marketing is capital allocation. You need to trust the algorithm and the operator at the same time, and you need the cash position to act when something starts working.

You do not reach $10M+ in spend by accident. You get there by making good decisions — particularly about who manages your media.

The consumer changed

If you assume Gen Z shops like millennials, the numbers will keep surprising you. Niche cultural interests drive real commercial behavior, entire categories emerge from fandoms that look unfamiliar from the outside, and short-form platforms function as habit loops rather than storefronts.

The shift toward entertainment-first social makes traditional attribution models progressively less useful. Consumers do not click to buy. They consume, get influenced, and decide days or weeks later, often somewhere else entirely.

Which means success depends on understanding niche trends even when they seem foreign, investing in creative that builds emotional stickiness, and planning campaigns for the long game. That argument is developed further in the memory economy.

What to check before Q4

July and August are when momentum gets built. Four indicators tell you whether it is working:

• New customer percentage
• New visitor percentage
• Email opt-ins
• CRM engagement

If those are trending up, you are laying a foundation for Q4. If they are not, the summer is when to fix it — because if you are not planning for Q4 during the slump, you are already behind. What to do once the peak arrives is in Black Friday and peak season.

FAQ

Why does DTC performance drop during Amazon Prime Day?

Because a faster, cheaper purchase path temporarily exists for the same product. Amazon order volume rose 99% year over year in 2025 while DTC brands that had been growing 20% to 30% went flat. If your ads drive to your own site while Amazon offers better price and delivery on the same item, cart abandonment rises and ROAS falls. It compounds the ordinary post-Fourth-of-July slump rather than replacing it.

What are reminder ads on Meta and do they work?

Reminder ads live in Meta's engagement objective and let users opt in to notifications about an upcoming sale, delivered before, during, and after the event without requiring an email or SMS signup. In accounts where weekend promos had underperformed, running them Monday through Friday brought forecasts in by Saturday. They work because attention is secured before the sale rather than bid for during it.

Should you run creative tests during a slow month?

Yes, but in isolated campaigns. Slow periods are the cheapest time to test, and the mistake is running promotional experiments through business-as-usual campaigns and disrupting something that already scales. Keep BAU clean, create separate campaigns for sale creative, and use minor holidays as rehearsals for Q4 structure.

What should replace last-click attribution?

A three-pronged model combining click-based, view-based, and platform-based measurement, rather than any single source of truth. Each successive platform change that limits click identifiers makes single-model attribution more fragile. The specific risk of last-click dependence is that it undervalues demand-creation channels, which leads teams to cut exactly the campaigns that fill the funnel.

What metrics indicate a healthy run-up to Q4?

New customer percentage, new visitor percentage, email opt-ins, and CRM engagement. All four measure whether the audience pool is growing rather than whether current revenue is efficient. If they are flat through July and August, Q4 will be constrained by audience size no matter how much budget is available.