If you only develop variations of high-performing ad creatives and improve them within the same appeal, you only raise the intensity of expression within that appeal, and the creative deteriorates into hooks and aggression. The real efficiency for a brand lies in solving low-performing ROAS, and stable growth that breaks through limitations comes from this process. Meta’s auction determines winners by the total value of bid price, predicted action rate, and ad quality combined. Creative with high spend won many auctions. Yet people didn’t buy right away because we failed to give complete trust and didn’t deliver value that fit them, so we couldn’t persuade them perfectly.
Soulpapamarketing’s branding & marketing story.
Regardless of which ad tool you use, the basic goal of operation is similar. Find creative that performs well and develop it further. Vary the copy of performing creative, change the first screen, and create new versions. Most operate this way, but limitations come quickly with just this approach.
Why do limitations come quickly when you only vary and develop performing creative?
Developing performing creative typically means changing the expression while keeping the same appeal. Even if the copy, first screen, model, and background change, the reason people should buy remains the same.
First, the people who respond don’t change much. A performing appeal has already found people who will respond. If the reason to buy stays the same even when you change the expression, the people who respond are almost the same. As the ad runs more, fewer people among them haven’t purchased yet, and bringing in new customers with only variations becomes increasingly difficult.
Second, the range of variation within the same appeal is narrow. After a few changes, you run out of new expressions to write. After that, there’s no way left but to increase intensity, so the first sentence becomes stronger, the thumbnail becomes more provocative, and you only build curiosity.
Third, the creative ultimately deteriorates into hooks and aggression. It’s hard to notice the deterioration because the initial response actually looks good. In the meantime, what the brand meant to say becomes unclear and more people respond to the stimulus than the product. In the next development, you raise the stimulus once more.
If you only make what’s already performing perform better, you end up saying the same reason to the same people more and more strongly. That’s when you need a different perspective. Look again at the creative that spends a lot but has low efficiency. It’s usually the creative you turn off first, but the answer is right here.
Why does creative with high spend and low ROAS win in auctions?
When you open the creative efficiency report, you see a few creatives with high spend but low ROAS every time. You usually turn off these creatives first. Before turning them off, let’s first check why so much money went in.
This is the actual Ad Manager screen. In the account above, the top-spending creative has a purchase ROAS of 2.53, while the second-place creative with about one-third the spend has 5.02. The account below has a larger ROAS difference. The top-spending creative is 0.42, while the third-place creative with less than one-fourth the spend is 6.98. Both accounts have budgets set at the campaign level, so Meta decided how much money to spend per creative.
Meta opens an auction every time there’s an opportunity to show an ad. According to Meta’s official guidance, the winner is not determined by bid price alone. The ad with the highest total value combining bid price, predicted action rate, and ad quality gets the impression. If high spend went out, then it won many auctions, and it’s still winning in competition now.
Without predicted action and quality rating supporting it, you can’t keep winning auctions. So if spend is large, that means many people showed interest in this appeal.
Common traits of low-efficiency ad creatives: attracted interest but people didn’t buy right away
These creatives’ appeals have something in common. The people who responded didn’t buy right away.
It’s natural that efficiency came out low since they didn’t buy right away. But why did money keep going into this creative? Because people responded to this appeal, and the algorithm saw that response and increased impressions.
If the creative couldn’t attract interest from the start, you need to find the appeal again from the beginning. If the creative attracted interest but didn’t sell, you need to revive the appeal and fill in what’s missing.
When you fail to give complete trust and don’t deliver fitting value, persuasion doesn’t happen
The reason people who showed interest didn’t buy right away is because we failed to give complete trust and didn’t deliver value that fit them, so we couldn’t persuade them perfectly. Persuasion is completed when trust and fitting value are delivered together. And to communicate trust and fitting value, ultimately it matters what actions the brand is consistently taking for this. Without action, there’s no evidence to show in ads and product pages.
| Checklist Item | Checklist Question | Where to Check |
|---|---|---|
| Trust | Did you show evidence to believe the content promised by the appeal? | Location and quantity of evidence such as ingredients, raw materials, reviews, certifications, etc. |
| Fitting Value | Does the utility expected by the interested person match the utility promised by the product page? | Product page screen that appears first when clicking the ad |
| Brand Action | Are there consistent actions the brand is taking for trust and fitting value? | Operating records consistently continued like raw material and manufacturing standards, customer service and after-sales management |
Without trust, people don’t believe even if you talk about value, and if value doesn’t fit, they don’t take it as their story even if they believe. If either one is lacking, they won’t buy.
Why is it difficult for brand advertising to expand through hijacking and benchmarking?
General commerce can expand widely in the direction that produces response. That’s why hijacking competitors’ performing appeals and benchmarking successful ad structures are commonly used.
When playing as a brand, random expansion like this is often impossible. Even if you get response, you can’t use it if it goes against brand philosophy. For a brand to expand and vary, it must align with our consumers while not losing our philosophy. It’s not easy at all to maintain both while expanding.
Why should ad creatives and product pages contain CEP, PP, and USP in images and videos?
Ad creatives and product pages are ultimately content. Both consist of text and images.
Reduce the text explanations and put the customer’s CEP and PP in the images and videos, showing USP that fits. Sense of presence is just background. In one scene, you need the moment when the customer thinks of the product, the inconvenience at that moment, and the product’s strength that solves that inconvenience all together.
CEP (Category Entry Point) is the specific situation the customer faces, the environment where the deficit is revealed. PP (Pain Point) is the deficit revealed in that situation, the pain and problem that needs to be solved. USP (Unique Selling Point) is the unique feature that solves that deficit and problem, and the evidence that persuades all the previous elements.
These three are appeals that manifest and are discovered from within what was mentioned earlier. They are things found by the brand looking into our customers and products.
So when developing through variation, benchmarking, or hijacking, you need to choose the direction as a brand. No matter what approach you use, the CEP and PP to include in images and videos should come from our consumers, and the USP from our product.
Below is the ad library screen of a partner company that has been working with Soulpapamarketing for a long time.
All three ads have CEP as the moment when the pet is active outdoors. The two harness ads answer concerns about whether it will be strong without restricting movement with lightness and safety. The suit ad solves the guardian’s concern about sensitive skin and reluctance to use repellent with an air layer that blocks insects without repellent.
This ad shows a pet coming down rocks. CEP is the moment when the pet moves dramatically on a rough path, and PP is the anxious feeling about whether the harness will hold up. The copy ’15 times stronger, ultra-lightweight safety’ above the image is the USP, and CEP and PP are shown as scene without text.
The order of improving appeals that emerge internally, from within
In the end, brands can’t make hijacking or benchmarking, which are common in commerce, their main direction. Using appeals that emerge internally, and if those appeals’ efficiency is not good, you need to improve from within to break through the limitations mentioned earlier. Even if a creative spends a lot but has low efficiency, if that appeal came from within the brand, you can use it for this task right away. Market interest is already confirmed.
In practice, we proceed in the following order.
- We extract only creatives with high spend and low efficiency. We exclude creatives with both low spend and low efficiency from this list. Because there’s insufficient data to judge if there was interest.
- We write that creative’s appeal in one sentence. We distinguish between expressions like thumbnails or opening sentences and the stories people actually responded to.
- We check whether trust or fitting value was lacking. It’s difficult to judge just by looking at the creative. We watch from clicking the ad through to the product page that appears, and check whether the brand’s actions that back up that trust and value are actually continuing.
- We keep the appeal the same and strengthen what’s lacking. If trust is lacking, we add evidence. If value doesn’t fit, we put the expected utility of those who responded in the first screen of the product page. Before increasing explanation sentences, we put the scene containing CEP and PP and the product’s USP in one shot first. We don’t increase hooking intensity at this stage.
- We group the improved version with the original creative under the same appeal for comparison. You can only know what changed efficiency if the appeal is the same.
Real efficiency as a brand depends on how you solve low-performing ROAS with high spend and grow it. Stable growth that breaks through limitations comes from this exact process.
Frequently Asked Questions
Should I turn off low-efficiency ad creatives right away?
We check the spend level first. Because creatives with high spend and low efficiency kept winning auctions, it’s already clear that many people are interested in this appeal. Even if we turn them off, if we record the appeal in one sentence and write down whether trust or fitting value was lacking, we can carry it directly to the next creative.
Why does Meta ad ROAS stay low even though spend continues to go out?
Meta’s auction determines winners by the total value of bid price, predicted action rate, and ad quality combined. If spend continues to go out, that creative is steadily winning auctions. It’s a signal that people responded but didn’t buy. However, if conversion tracking or campaign optimization event settings are wrong, similar numbers appear, so check the settings before analyzing the appeal.
What happens if you only continue to develop performing ad creatives?
You only change expressions within the same appeal, so the people who respond don’t change much. When you run out of new expressions to write, you raise the stimulus of the opening sentence and thumbnail, and the creative deteriorates into hooks and aggression. If you’ve only raised stimulus in recent developments, check creatives with high spend and low efficiency again.
Is it okay to benchmark competitors’ appeals in brand advertising?
You can reference them. However, it’s difficult to use as the main path for expanding brand advertising. Even if you get response, you can’t use it if it goes against brand philosophy. Competitors’ appeals work because they have evidence and trust backing the competitor’s product, so taking just the sentence doesn’t produce the same response. In competitor ads, only reference the structure, and find the customer’s moment and the product’s strengths to include in the scene from our side.
Related Articles
- Why Meta ad efficiency doesn’t work — 2 essential ROAS settings that 99% don’t do
- Meta ad error 2643186 — Actual solution that worked when ‘Not Showing’ continues
- Low-budget Meta ads, the harm of mindless winning search
→ Brand Directing Approach — Performance and growth marketing agency that builds brands with data
Insights from Soulpapa Marketing — Korea’s digital marketing agency.
Original Korean article: https://soulpapa.co.kr/2026/09/15/ad-creative-high-spend-low-roas/
