If you run Meta ads at any real volume, you know the problem.
You've got a winning ad. It's been spending for six weeks, ROAS is holding, and the algorithm loves it. You upload three new creatives — new hooks, fresh visuals, different angle — and drop them into the same ad set. You check back in 48 hours.
The old ad is at $800 in spend. Each new one sits at $11.
Meta's algorithm has made its decision. It already knows what works in this ad set, and it's not interested in your new ideas until they prove themselves — which they can't, because they're not getting any spend.
This is the creative rotation trap. And as of July 10, 2026, Meta gave you a direct lever to fix it.
It's called Push Delivery to This Ad. Here's what it does, how to set it up, and how to use it without breaking what's already working.
The Problem: New Creative Goes Dark While the Old Winner Keeps Spending
The reason this happens isn't Meta being arbitrary. It's the Andromeda algorithm doing exactly what it's designed to do.
Andromeda optimizes delivery at the individual ad level, drawing on a massive pool of signals — your ad's historical performance, the creative's visual and copy attributes, the audiences it's been shown to, and the conversion outcomes that followed. An ad that's been live for 45 days with consistent ROAS has an enormous data advantage over an ad that was uploaded this morning.
When you add a new creative to an existing high-performing ad set, the algorithm is comparing a proven asset with 45 days of signal against a blank slate. The proven asset wins on every delivery decision. Your new creative sees trickle-level impressions that give it almost no chance to prove itself.
The old workaround was to duplicate the ad set — create a fresh copy with only the new creative, let it enter the learning phase separately, then graduate the winners back to your main ad set. It worked, but it was messy. You'd have multiple ad set variants competing for the same audience, overlapping delivery, and learning phase disruption every time you rotated.
Push Delivery is the first real, built-in alternative Meta has given advertisers.
What "Push Delivery to This Ad" Actually Does
Push Delivery to This Ad lets you allocate a specific percentage of your ad set's daily budget directly to one ad for a defined number of days.
The percentage you choose goes exclusively to that ad during the push window. The rest of the budget continues to be distributed by the algorithm across your other ads. The pushed ad gets guaranteed impressions and spend — enough to generate real signal, not just statistical noise.
When the push period ends, the ad re-enters normal algorithmic distribution. Critically, any performance data accumulated during the push carries forward. The ad doesn't lose its learning history and restart the learning phase. It enters algorithmic competition with real data behind it, which is the point.
This is not a "force this ad to win" button. Push Delivery gives a new creative the spend it needs to be evaluated fairly — it doesn't guarantee it performs. If the pushed creative generates poor ROAS during the push window, that signal will count against it when the push ends.
Why This Is Different From a Separate Testing Campaign
Foxwell Digital makes this distinction clearly, and it's worth repeating: Push Delivery is a delivery assurance tool, not a testing tool.
Meta's creative testing campaigns (via the dedicated A/B testing or separate testing ad sets) are built to statistically compare creative variants against each other and identify a winner. They're designed for directional insight — which hook resonates, which visual format converts, which offer framing works.
Push Delivery is built for a different job: you've already identified what you want to run, and you need the algorithm to actually look at it.
The two tools operate at different points in your creative process:
- Testing campaign → Use when you're running a controlled comparison and need a stat-sig answer
- Push Delivery → Use when you have a new creative that you're confident in and you want it integrated into an existing performing ad set without the dead-zone spend problem
If you're not sure whether the creative is worth pushing, that's a sign you should be running it in a test first.
How to Find and Enable Push Delivery (Step-by-Step)
Push Delivery rolled out at 100% of advertisers on July 10, 2026, so it should be available in your account now. Here's how to access it.
- Open your ad set in Ads Manager. You need at least two ads active or in draft — the feature doesn't appear with a single ad.
- Click into an individual ad (not the ad set level — the setting lives at the ad level).
- Navigate to Ad Setup. Scroll past the creative section.
- Look for the push delivery toggle. It sits under the advanced settings, near multi-advertiser ads. It's off by default.
- Set your percentage and duration. Enter the percentage of the ad set's daily budget to allocate to this ad, and the number of days to run the push.
- Save and publish. The push starts immediately on the next delivery cycle.
If you don't see it yet, check from the ad creation flow in an existing ad set rather than from the ads table. Some accounts surface it more reliably there.
The Settings Playbook: What Percentage and How Many Days
This is what none of the mechanics guides cover, and it's where most brands will get this wrong.
On duration:
The default starting point is 7 days. That's appropriate for most ecommerce brands running $200–$500/day ad set budgets, because it gives the creative enough time to accumulate meaningful signal across the purchase consideration window — especially for fashion and health products where a customer might need 3–5 touchpoints before converting.
If your product has a longer purchase cycle (higher-ticket health programs, custom swimwear) or if you're running to cold audiences with broader funnels, consider 10–14 days. Past 14 days, you're likely not getting new signal — you're just paying for impressions.
Don't go to 30 days unless you're testing a full seasonal campaign creative that needs runway across a multi-week window (Black Friday, holiday, major drops). Even then, 21 days is usually enough.
On percentage:
Start at 20–30% of your ad set's daily budget. Here's the practical math:
- $300/day ad set → Push 20–25% → $60–75/day to the new creative for 7 days = $420–525 total push spend
- $500/day ad set → Push 20% → $100/day for 7 days = $700 total push spend
- $1,000+/day ad set → Push 15% → $150+/day for 7 days = $1,050+ total push spend
The goal is to give the pushed ad enough daily spend to generate purchase events. For most ecommerce brands, you need at least $50–75/day in spend for a creative to generate enough conversion data to mean anything. Don't push at a percentage so low that the new creative is still effectively invisible.
Don't go above 40%. If you're pushing more than 40% of your budget to one unproven creative, you're gambling your ad set's performance on something with no data. Let the algorithm keep doing its job with the proven ads while the new one earns its spot.
Push only one ad at a time per ad set. Meta has flagged that pushing multiple ads simultaneously degrades overall delivery. If you have two new creatives you want to test, stagger them.
When to Use Push Delivery — and When NOT To
Use it when:
- You have a new creative you're confident in (based on research, winning hooks, or proven angles from other contexts) and you need it to get real spend fast
- You're refreshing creative in a high-performing ad set and want to maintain performance continuity rather than duplicating the ad set
- You're launching a time-sensitive seasonal creative that needs to accumulate signal before the window closes (pre-Q4, product drop, sale event)
- The incumbent ad in your ad set is showing signs of creative fatigue (declining CTR, rising frequency) and you need to rotate in a replacement
Don't use it when:
- You're not sure if the creative is ready — Push Delivery is not a validation tool, it's a delivery assurance tool. Run a test first.
- You want to force a winner — if the algorithm consistently underdelivers to your new creative even with a push, that's signal. Don't override it by extending the push window.
- You're running ASC (Advantage+ Shopping Campaigns) — Push Delivery operates at the ad set level within manual campaigns. ASC controls creative distribution differently and has its own creative ranking logic.
- You have only one active ad in the ad set — you need at least two for the feature to appear.
Fashion/Swimwear and Health Brand Use Cases
For the ecommerce brands Dash Activate works with, Push Delivery is most useful in two specific scenarios:
Fashion and Swimwear: Seasonal creative rotations
Fashion and swimwear ads have a seasonal expiration date. A winning summer campaign creative from June starts to feel stale in late July. But you can't afford to pull it until something else is proven — you're in peak season.
Push Delivery is exactly right here. Keep the June winner running. Push 20% budget to your August creative for 7–10 days. If it can match or outperform the incumbent, you have clean data to justify rotating. If it underperforms, you've spent a controlled amount finding that out — and the June creative is still running the account while you learned.
This is especially useful for Q4 prep. September is when you should be pushing your Black Friday and holiday creative into performing ad sets and letting it accumulate signal, so by October you're competing with proven assets — not blank slates.
Health and Natural Products: Compliance-updated creative
Health and supplement brands frequently need to update creative for compliance reasons — a claim gets flagged, a visual needs updating, or policy changes require a revised hook. When this happens, you don't want to shut down the existing ad and restart cold.
Push Delivery lets you introduce the compliance-updated version alongside the original, give it 7 days of spend to establish signal, and then transition delivery once the new version has a track record. You maintain account stability while making the necessary update.
For health brands running policy-sensitive claims, this also gives you a way to test revised messaging angles (softer language, reframed benefits) without burning a full ad set on an unproven variant.
Where Push Delivery Fits in Your Creative Testing System
Most ecommerce brands running Meta ads at $300+/day have some version of a creative testing pipeline: a structure for introducing new concepts, identifying winners, and graduating them into scaling ad sets.
Push Delivery doesn't replace that system. It fills a specific gap in it.
Here's how it fits:
Stage 1 — Concept validation: Run new creative concepts in a dedicated testing ad set or A/B test to identify directional winners. This stage is about figuring out what resonates, not delivery continuity.
Stage 2 — Push Delivery: Once a concept proves itself in testing (above your ROAS threshold, reasonable CTR, enough conversion signal), promote it to your performing ad set and use Push Delivery to integrate it cleanly. This is where the delivery guarantee matters — you've already de-risked the creative, now you're ensuring it gets a fair seat at the table.
Stage 3 — Algorithmic rotation: After the push window, the new creative competes on equal footing. Monitor ROAS, frequency, and click behavior over the following 7–10 days to confirm it's holding. If it is, you've successfully rotated creative without the dead-zone spend problem.
This is the system we use with clients running at scale. Concept validation in isolation, delivery assurance in integration, algorithmic confirmation in rotation.
The brands that struggle with creative rotation are usually skipping Stage 2 — dumping new creative directly into performing ad sets without any mechanism to get it spend, then concluding the creative "didn't work" when it never had the chance.
Push Delivery solves that specific problem. Use it at the right stage.
The Bottom Line
Push Delivery to This Ad is the most useful manual override Meta has given advertisers in years. For ecommerce brands running at volume, it's a direct solution to the creative rotation problem that most teams were working around with ad set duplication and structural hacks.
The mechanics are simple. The discipline is knowing when to use it, what settings to dial in, and where it fits in your broader creative system — not just clicking the button because it's there.
Start at 20–30% of budget, 7 days, one ad at a time. Validate the creative first. Use the push to integrate, not to discover.
If you want a second set of eyes on your current creative testing system — what's rotating, what's stuck, and where you're leaving efficiency on the table — book a strategy call. No pitch, just a real look at your account.
