Creative volume is the new targeting.
Privacy changes shrank audience targeting precision, and Meta rebuilt delivery around the ads you give it. The new edge is a steady supply of different concepts, sized to your spend.
What changed in this revision: Added how Meta’s Andromeda retrieval engine changes the role of creative, a worked table for sizing monthly concept supply to spend, a decision rule for when to fix supply before raising budget, and a question section. Removed unsourced performance multiples.
For about a decade, the brands winning at paid social were winning on targeting. Better audience definition, sharper interest segmentation, smarter lookalike modeling. The pixel collected enough behavioral data that the algorithms could find buyers reliably, even with mediocre creative.
iOS 14 weakened that era. Audience signals shrank, pixel data became partial, and lookalikes became less precise. The brands that had been competing on targeting precision watched that advantage fade, while the brands that had been competing on creative were far less exposed.
The new edge is creative. But not the kind most brands produce, and not in the quantity most brands assume.
What changed in 2026: the system reads the audience from the ad
Meta has been explicit about the direction. Its engineering team describes Andromeda as the retrieval engine that selects which ads are candidates for each person, rebuilt because the number of eligible ads keeps growing as Advantage+ automates audiences, placements, and creative. The published goal is more capacity for personalization and, over time, a more diverse set of ad candidates.
Read that as an operator and the implication is plain. When the system chooses who sees an ad by matching people to the ad itself, the creative carries the targeting signal. A hook aimed at a first-time buyer, a demo aimed at a skeptic, and a price-led ad aimed at a comparison shopper are three audiences, whether or not you ever define them in an ad set.
That changes what "more creative" should mean. Twenty near-identical edits give the retrieval system twenty ways to reach the same person. Five genuinely different concepts give it five different people to find. Volume only works when it is volume of difference.
What kills accounts now
In our judgment, the most common reason a Meta account plateaus is not the bid strategy or the structure. It is creative supply falling behind spend. The account keeps asking the same few concepts to carry a growing budget, and they tire.
Creative fatigue shows up in a recognizable order. Frequency climbs among the people most likely to respond, click-through rate softens, then cost per purchase rises while the ad still looks fine in the headline numbers. How fast that happens depends on audience size, spend, and the concept, so any fixed "half-life" is a guess. The pattern is reliable even when the timing is not.
Brands that produce one polished hero spot per quarter are asking a single concept to carry a quarter of spend. It will not, and the account drifts down while the team waits for the next shoot.
What creative volume actually looks like
Volume does not mean twenty variations of the same hero spot. That is almost as limiting as one spot. Volume in the useful sense means structured variety across genuinely different concepts, with each concept then expressed in the formats the placements need.
- Distinct hooks per concept, varying the first two to five seconds
- Format variants for Reels, feed, Stories, and in-stream placements
- Length variants, from short cut-downs to longer explainers
- Style variants: creator-style, founder-led, lifestyle, product demo
- Different reasons to buy: problem-led, proof-led, price-led, story-led
The first and last items create new concepts. The middle three adapt a concept to placements. Count them separately. A month with forty assets built from two concepts is a month with two concepts.
How many concepts does your spend need?
The useful question is not "how much creative should we make" but "how much creative does this budget consume." Two numbers from your own history answer it: how much spend a winning concept absorbs before its cost per purchase drifts up, and what share of the concepts you test become winners.
| Monthly Meta spend | Winners needed (spend ÷ $15,000) | Concepts to test (winners ÷ 20%) |
|---|---|---|
| $20,000 | 1.3 | About 7 |
| $60,000 | 4 | 20 |
| $150,000 | 10 | 50 |
| $150,000, if a winner absorbs $30,000 | 5 | 25 |
Two things fall out of the arithmetic. First, creative supply scales with spend, so a brand that doubles its budget without doubling its concept pipeline has quietly halved the life of every ad. Second, the two levers are the concept’s staying power and your hit rate. Stronger concepts and better research reduce the volume you need. Neither removes the need.
To get your own numbers, pull ad-level spend for the last 90 days. For each concept that won, add up the spend it carried until its cost per purchase rose meaningfully above the account average. The median of those totals is your absorption figure. Divide the winners by the concepts launched in the same window for your hit rate.
Decision rule: if the concepts you launch per month are below the "concepts to test" figure for your spend, fix supply before raising budget. More budget on a thin pipeline buys fatigue faster, not growth.
Why polished creative often loses
Counterintuitively, expensive polished creative often underperforms native, creator-style content on social platforms. The reason is platform fit. Reels and TikTok reward content that looks like it belongs in the feed. Commercial-style creative interrupts the feed; creator-style content blends into it.
People scroll fast and have learned to skip anything that looks like an ad in the opening seconds. A polished spot often trips that reflex before its message lands. That is an observation about format, not a rule about quality: a well-made product demo can beat a lazy creator clip, and the only way to know in your account is to test both.
This is why so many brands have moved to creator partnerships and partnership ads. The content already looks native, the creator already has credibility, and production cost per concept falls, which is what makes the supply numbers above achievable.
How to actually build a creative engine
Most brands struggle with volume because their production process is built for polished one-offs. Long briefing cycles, agency rounds, legal review, multiple approval gates. By the time a single ad ships it has cost two weeks of effort. Producing twenty concepts a month with that process is not possible.
The brands that have solved this rebuilt production around supply. Lightweight briefs that name one concept and one reason to buy. Creator partnerships for production. Batch shoots that cover several concepts in a session. Approval steps designed for fast iteration, with brand and legal rules written down once rather than argued per asset.
AI tooling helps at the iteration layer: copy variants, thumbnail concepts, voiceover options. It does not replace the judgment of which concept is worth making, and in our view that judgment is where most of the value sits. For the operating detail, the creative production system walks through briefs, cadence, and roles.
How to tell whether volume is working
Volume is a means, not a score. Judge it at three levels. At the ad level, more concepts should mean more of spend sitting on ads that are not yet tired. At the account level, cost per purchase should hold steadier as budget rises. At the business level, the only number that settles it is whether new customers are arriving at a cost the margin can carry.
That last check lives outside Ads Manager. Read spend against total revenue and new customers, as covered in reporting beyond Ads Manager, and check that the cost of a new customer pays back in a reasonable time with the CAC payback calculator. Keep the testing itself disciplined, with a creative testing system that separates a concept win from noise.
The new playbook
The brands that win paid social over the next few years will be the ones running creative engines, not creative one-offs. Structured testing, creator partnerships, lightweight production, and a willingness to ship concepts that turn out to be wrong.
That is uncomfortable for teams who came up when targeting was the lever. But targeting precision is not coming back, and the platforms have rebuilt delivery on the assumption that it will not. Creative supply is also one of the few growth constraints fully inside your control, which is why it sits near the center of how to scale an ecommerce business.
If your account is still built around one polished hero spot per quarter, the question is not whether it will tire. It is how much budget it will waste while you wait for the next one.
How many new creatives should a Meta account launch each month?
It depends on spend, not on a universal number. Divide monthly spend by the spend a winning concept absorbs before it tires to get the winners you need, then divide by your hit rate to get the concepts to test. A small account may need a handful; a large one may need dozens.
Is a new edit of an existing ad a new concept?
Usually not. A new crop, length, or caption adapts an existing concept to a placement. A new concept changes the hook, the reason to buy, or who the ad is speaking to. Count concepts and adaptations separately, because only concepts give the delivery system new people to find.
Does Andromeda mean audience targeting no longer matters?
It matters less as the main lever. Exclusions, geography, and the conversion event you optimize for still shape delivery. But with broad audiences and Advantage+ automation, the creative increasingly decides who an ad reaches, so creative variety does work that audience segmentation used to do.
How do we know a concept has fatigued rather than had a bad week?
Look for the sequence rather than one number: frequency rising among engaged audiences, click-through rate softening, then cost per purchase drifting up over more than a few days. A single bad day with steady frequency is more likely noise, a tracking gap, or a site problem.
Should small brands chase creative volume too?
Yes, scaled to their spend. A small account needs fewer concepts, but it also has less data per concept, so each one should be more clearly different. Five distinct ideas teach you more than fifteen versions of one.
Written by Sophie Mills, performance marketing strategist. If this resonated and you want to apply it to your own account, you can book a strategy call or run a free audit.
How we research, source figures, and handle corrections: editorial policy.