What does Meta's Push delivery to this ad setting do, and when is it worth forcing spend onto one creative?

TofuAnswer by the Tofu team · checked

Push delivery to this ad sets aside a percentage of your campaign or ad set budget for one ad, for a set number of days, to help it get spend. Advertisers first reported it in March 2026. Meta hasn't documented it or said which accounts have it; the feature details come from Jon Loomer's tests. Use it when one new ad must get seen, like a webinar ad, in an ad set that already works. To pick a winner fairly, use a creative test or an A/B test.

Ask Tofu next

Pick a question and Tofu Ads Agent answers it in chat.

As of November 2026.

Push delivery vs creative test vs A/B test vs a separate ad set

Meta shows each person the ad in your ad set that it predicts will get "the lowest cost per optimization event," so "each of your ads won't necessarily be delivered the same number of times." Its troubleshooting page adds: "it's normal for some ads to have more impressions than others." All four options below work around that. Only the last three are in Meta's help center. The push delivery column comes from Jon Loomer, who writes about Meta ads and tested the setting in June 2026.

Push delivery to this adCreative testA/B testSeparate ad set with its own budget
What it doesReserves a share of the campaign or ad set budget for one ad, new or already runningCopies one ad into 2 to 7 test ads inside your live campaign and spends a budget share on themSplits your audience so nobody sees both versions, then compares resultsGives the ad a budget no other ad in the old ad set can take
Where you set itOn the ad, in its setup: "show more settings," then below the scheduling setting (per Loomer)At the ad level, in the Creative test section, Set up testA/B Test in the Ads Manager toolbarWhen you create the ad set
BudgetA percentage you type in. No published minimum or maximumA share you choose. Meta suggests no more than 20% and aims to spend it each dayMeta recommends the same budget for each versionThe ad set's own budget
How longTooltip suggests up to 7 days "or until your ad set ends." Loomer found 1 to 30 days allowedA duration you setMeta recommends at least 7 days. The maximum is 30Until you turn it off
Gives you a fair winner?No. It buys delivery, not a comparisonPartly. Meta emails "high performers if any were found," with no confidence levelYes. It's the cleanest comparisonNo. It starts its own learning phase, and it isn't a side-by-side test
NeedsAt least 2 ads in the ad set (per Loomer)The Highest volume bid strategyBudget for every versionIts own budget, ideally outside a shared campaign budget
Meta documentationNoneYesYesYes
Main limitCreative tests are greyed out while a push runs, and Ads Manager warns that several pushes may not each get their minimum (per Loomer)Highest volume only, and the test adds new ads to your campaignSplits a small budget furtherMore to manage, and it can compete with your other ad sets

When Push delivery to this ad is the right pick

When you need one ad seen in an ad set that already gets leads, and you don't want to rebuild anything. For example:

  • A time-bound ad. A webinar, an event or a launch. Loomer suggests adding the new ad to the existing ad set and pushing a share until the promotion ends.
  • A new ad stuck near zero spend. In March 2026, Loomer passed on advertisers' reports that the Delivery column flagged some published ads as getting "limited or no ad spend" and offered the push from there.
  • A short read on a low-impression ad. Loomer's advice is to push mainly when the ad set is underperforming overall, for a few days.

It is not the right pick to find out which ad is best. The pushed ad gets delivery Meta didn't choose to give it, so its results aren't a fair match against the ads Meta picked. Loomer also warns against pushing every ad that isn't delivering: making a habit of it, he writes, "feels like a recipe for driving up costs."

When a creative test is the right pick

When you want new versions of one ad tested inside the live campaign. Meta built it so winning ads keep running in the same campaign, with "no need to merge them into another campaign where the learnings would reset." It needs Highest volume bidding. Loomer found Set up test greyed out while a push is running. The full setup is in how to test new B2B ads fairly on Meta.

When an A/B test is the right pick

When the decision matters, like a new offer or a new message for your whole funnel. Meta shows "each version to a segment of your audience and ensure[s] nobody sees both." Meta recommends tests of at least 7 days and a budget "that will produce enough results to confidently determine a winning strategy." On a small B2B budget, test two versions, not five.

When a separate ad set is the right pick

When an ad needs guaranteed budget for weeks, not days, or targets a different audience. Give it its own budget. Inside a shared Advantage+ campaign budget, Meta can still move money away, and Meta says "A minimum spend limit is a target range, not a guarantee" (see ad set spending limits). Expect a new learning phase.

How much budget can a push reserve, and for how long?

Meta hasn't published limits. Here is what Loomer reports:

  • Budget share. You type a percentage of the campaign or ad set budget. Loomer found no stated minimum or maximum and tested 10%. Treat any fixed range you see quoted elsewhere as unconfirmed.
  • Days. The tooltip suggests up to 7 days. Loomer found the field accepts 1 to 30.
  • When it ends. The ad goes back to normal delivery. Loomer believes you can turn a push off early.
  • Floor or ceiling? Loomer couldn't confirm whether Meta can spend more than the pushed share on an ad that does well.

Do the math before you push. Our arithmetic, not Meta's: at $150 a day, a 15% push for 7 days is about $158 on that ad. At $90 a lead, that's under 2 leads. That's enough to get the ad seen, not enough to judge it on leads. Judge it on what the spend can show, like click rate and cost per landing page view, or keep the push longer.

Why does Meta warn that minimum spend may not be met?

Because every push takes from the same budget. When Loomer added a second push, Ads Manager said "Another ad is using this setting" and warned that "minimum spend per ad may not be met." Meta hasn't explained the warning further.

The warning doesn't block you. In his June 11 test, Loomer pushed 10% to each of five ads, half the budget, and over the first $75.02 Meta split spend as set. He didn't test bigger budgets or longer pushes. Pushes that add up to most of the budget leave little for the ads Meta prefers, so push one ad at a time.

What Meta hasn't said

  • Whether a push resets learning. Meta lists adding a new ad as a significant edit, and big budget changes can be one too. It doesn't say whether a push counts.
  • Which objectives and bid strategies support it. Loomer couldn't confirm it works the same for every performance goal.
  • How it shows in reporting. Loomer's posts don't describe a label or column for pushed ads.
Sources (12)

Related answers

More on