Why One Ad Gets All the Spend in Advantage+ and What to Change
Meta's delivery system backs the ad it predicts will win. When to leave that alone and how to give new creative a fair read, with settings by weekly conversion volume.
On this page
When a Meta Advantage+ campaign budget puts nearly all the spend on one ad and new creative is not delivering, the system is behaving as documented. Meta says as much on its troubleshooting page: "If your ad is not receiving enough results but other ads using the same budget are, that's okay." Our position is that this concentration is what you should want from an optimizer, and that the frustration comes from expecting it to double as an experiment. A fair read for new creative has to be set up on purpose, and the right setup depends on weekly conversion volume. Meta's pages are quoted as we read them on October 7, 2026.
Why does delivery pile onto one ad?
Advantage+ delivery concentrates spend on one ad and ignores new creative because every impression follows a prediction, and the fix begins with accepting that predictions tend to favor whatever already has history. Meta describes the auction winner as the ad with the highest total value, a combination of three major factors: the bid, estimated action rates and ad quality. By our reading, ads inside one ad set go to auction under the same budget and bid settings, so the last two factors separate them, and an ad with months of purchases behind it gives the system far more to estimate from than one uploaded yesterday. The troubleshooting page finishes the thought: "The ad delivery system predicts which ads will generate the best results. Sometimes this means that other ads in the same ad set or campaign won't deliver."
Budget pooling follows the same logic one level up. According to Meta, Advantage+ campaign budget "may not spend your budget equally for each ad set."
When one ad taking the budget is the right result
A Facebook Ads Advantage shopping campaign that concentrates budget on one ad is often doing what you pay it to do. The developer documentation from Meta describes Advantage+ sales campaigns as the updated, streamlined version of Advantage+ shopping campaigns, so we treat the two names as one campaign type.
Leave the concentration alone when three things hold: the campaign total is at or above target, frequency on the lead ad is steady, and no pending decision needs data on a second ad. The campaign budget page tells advertisers to analyze results at the campaign level rather than the ad set level, and we apply the same discipline to ads. Judge the total against break-even, which our guide to what ROAS is works through, with the ROAS calculator as a further resource. A platform-reported figure counts attributed sales, a different question from the sales the ads caused, and our piece on incrementality covers that gap.
Concentration becomes a liability when the lead ad tires. The troubleshooting page names the signals: high frequency, or a Delivery status that reads "Creative fatigue." A campaign with nothing tested to replace that ad is exposed the week the status appears, which is the argument for testing early.
What happens when you add new creative to an existing ad set?
When you add new creative to an existing ad set in Meta ads, delivery and spend change in two ways, and as of October 2026 neither one amounts to creative testing in an Advantage+ campaign. First, Meta lists "Adding a new ad to your ad set" among the significant edits that send an ad set back into the learning phase, a period that "can lead to less stable performance and a higher cost per result." Second, the new ad has to win impressions from the incumbent on predicted results. The troubleshooting page says the learning phase usually ends after "about 50 optimization events in the week after the ad set's last significant edit." Dropping five fresh ads into the winning ad set every Monday restarts that clock each week and still promises none of them a read.
Can you force budget onto the other ads?
To force budget to other ads in an Advantage shopping campaign, the documented routes are a creative test, a spend limit on a second ad set, a separate test campaign and an A/B test. None of them is a budget set on a single ad, and we found no ad-level spend control in the Meta help pages we read.
The creative test built into Ads Manager
To force Facebook to test new creative in an Advantage+ campaign, start with the creative test, which Meta says lets you "compare up to 7 creative variants with delivery provided to new test ads." The settings, from that page:
- Bid strategy: Highest volume. Creative testing is supported with that bid strategy only. Switching to it counts as a significant edit, per the significant edits page.
- Where: at the ad level, in the Creative test section, click Set up test.
- Copies: 2 to 7 copies of the ad, each edited after you confirm the setup.
- Test budget: Meta suggests "no more than 20% of your existing budget" and aims to spend that amount per day on test ads.
- Duration and metric: set the test length and the comparison metric, then confirm and publish.
Two details matter once the test ends. The test ads keep running on the original budget, and the page states that the test "does not make any automatic changes based on the results," so pausing the losers is your job. The results identify the top performing test ads against your comparison metric, with no confidence level included, which leaves the sample-size judgment with you. We found no statement on that page about whether creating test ads counts as a significant edit.
A second ad set with a spend limit
Spend limits belong to ad sets. Meta offers minimum and maximum ad set spend limits when Advantage+ campaign budget is selected and Advantage+ on shows on the Budget card, set as a percentage or a dollar amount under Budget & schedule, in Ad set spending limits. For new creative that means a second ad set in the same campaign, holding only the new ads, with a minimum spend limit. The page attaches cautions: Meta calls a minimum a "target range" and issues no refunds when delivery misses it, recommends as few spend limits as possible, and advises against a minimum and a maximum on one ad set at the same time.
Adding that ad set does not push the others back into learning, according to the significant edits page. When the budget requirement is specific, Meta's own suggestion is an ad set budget, and by our reading the cleanest form of that is a separate test campaign with its own money. Keep the original campaign on campaign budget: Meta says a campaign with multiple ad sets that is not using campaign budget turns Advantage+ off, while optional spending limits keep it on.
An A/B test
When the answer has to survive a budget meeting, run an A/B test. Meta shows each version to a segment of the audience and ensures nobody sees both, recommends the same budget for both versions, and does not recommend testing informally, "such as by turning ad sets or campaigns on and off manually." The troubleshooting page points to the same tool: "If you want your ads to deliver equally, try creating an A/B test." Duplicating a campaign for the test duplicates its ad budget too, so we keep this route for decisions that move a large share of spend.
Which route fits your budget and weekly conversions?
The thresholds in this table are our working rules, built on two numbers Meta publishes: about 50 optimization events in a week for the learning phase to end, and no more than 20% of budget on test ads. Divide 50 by 0.20 and you get 250, the weekly volume at which a test ad set holding a fifth of the budget could reach the learning figure by itself, assuming it converts at the campaign's average cost per result.
| Weekly results in the campaign | Route | Settings |
|---|---|---|
| Under 50 | Keep one ad set and use the creative test | Highest volume, 2 copies, test budget at 20% or less, a longer test window |
| 50 to 250 | Creative test in the existing campaign | Highest volume, 2 or 3 copies, test budget at 20% or less, 14 days as our default |
| Over 250 | A standing test ad set, or a separate test campaign | New ads only, a minimum spend limit set as % or $, no maximum on the same ad set |
| Any volume, high-stakes decision | A/B test | Duplicate, change only the creative, same budget for both versions |
A hypothetical example, with assumed numbers. A store spends $300 a day at $40 per purchase, which is 7.5 purchases a day and 52.5 a week. A creative test at 20% is $60 a day, or $840 over 14 days. If the test ads convert at the campaign's $40, a generous assumption for unproven creative, that buys 21 purchases. Split evenly, which Meta's page does not promise, that is 7 per ad with three copies and 3 per ad with seven. Three purchases will not separate one ad from another, so this store tests two or three ads at a time. At the same $40, the 250-a-week line equals $10,000 a week, or about $1,429 a day. That $40 is a platform figure, and our guide to CAC covers the fully loaded one.
What to do when a test ad wins
Our view is that the winner should then be allowed to take the budget. Pause the losers, keep the winner running where it is, and expect delivery to pool again. If the result justifies more budget, raise it in measured steps: the significant edits page treats a budget change from $100 to $101 as unlikely to restart learning and $100 to $1000 as one that may. Our guide to scaling paid ads without killing ROAS covers the pacing, and our paid media team is there if you want a second opinion on the structure. One ad carrying the campaign is a comfortable place to be once its replacement has had a fair read.
Sources
- Meta, Troubleshoot ad performance: You're not receiving enough results
- Meta, About ad auctions
- Meta, About Advantage+ campaign budget
- Meta, Advantage+ Campaign Experience for Sales, App, and Leads, Marketing API documentation
- Meta, Significant edits and learning phase
- Meta, Set up a creative test in Meta Ads Manager
- Meta, About ad set spend limits with Advantage+ campaign budget
- Meta, About A/B testing
- Meta, Choose your budget strategy for Advantage+ campaigns