Testing & Scaling – Questions about structure
Hello,
I hope you’re doing well.
I have a few questions regarding the logic between testing and scaling campaigns within ASC+.Current setup
- Testing campaign (ABO)
One creative per ad set
No audience targeting or signals
Exclusion of existing customers (via customer file + pixel), to ensure we’re testing on a cold audience
Tests run over several days with a budget equivalent to 3/4x the target CPA
When a creative performs well, I move it into scaling.
- Scaling campaign (ASC+)
All winning creatives from different products are grouped into a single ad set
Exclusion ofAdd to CartandInitiate Checkoutaudiences
- Retargeting campaign (CBO)
Multiple ad sets:
Add to Cartvisitors 30 daysInitiate Checkoutvisitors 30 days
Abandoned payment 30 days
- Shopping campaign
Includes all products from Shopify
Performs very well alongside the other campaigns
ASC+ structure: strategic question
The reason I run my ABO tests without audience signals is because it doesn’t feel relevant to include all signals from different products into a single ASC+ ad set.I’m in the supplement space, and each product addresses very different needs and symptoms which means each has its own ideal audience.
So here’s my question:
Is it still recommended to group all creatives into a single ASC+ ad set using broad targeting?
Now that ASC+ allows multiple ad sets, wouldn’t it make more sense to build separate ad sets by product line, each with relevant audience signals?
Or would that break the core principle of how ASC+ is designed to work?Catalog ads in ASC+
One last point:
I currently have catalog ads activated within my ASC+ campaign.At this stage, I’m wondering if that’s really the right choice:
Doesn’t it risk biasing delivery in favor of catalog items instead of my tested creatives?
Could this undermine my creative-based scaling approach?
Especially since I already have a dedicated Shopping campaign running in parallel that’s performing very well.
The short answer is:
Continue with the single broad ad set in your ASC+ scaling campaign, but use Advantage+ audience signals within it to suggest your different product audiences.
For testing, your current ABO structure is great, but isolate your testing budget from your scaling budget.
For the catalog ads, it’s generally best practice to keep them active in the ASC+ as they provide the AI with valuable product-level purchase intent data, but if you suspect they’re cannibalizing the performance of your tested creatives, you can test a structure where you exclude your catalog from the ASC+ and let your separate, high-performing Shopping campaign handle all catalog-based dynamic ads.
The long answer is:
Your current campaign structure is quite sound, leveraging ABO for controlled testing and ASC+ for automated scaling, which is a common and effective hybrid strategy.
Regarding your strategic question on ASC+ ad set structure, while Meta now allows multiple ad sets in Advantage+ Sales campaigns, the core principle of ASC+ remains maximum consolidation and broad targeting for the best AI optimization.
Creating separate ad sets by product line with distinct audience signals would break the intended function of the ASC+ algorithm, which thrives on a large, undivided audience pool to efficiently find the best customers across all your creatives.
Your instinct that different products appeal to different audiences is correct, but the solution in ASC+ is to let the AI figure that out based on creative performance, not by segmenting the ad sets.
Instead, you should utilize the Advantage+ audience signals feature within your single ASC+ ad set to upload your custom audiences (like past buyers of a specific product line or lookalikes based on those buyers) as suggestions for the algorithm, rather than hard constraints.
The AI will prioritize these signals without limiting its ability to prospect broadly and allocate budget dynamically across all your creatives to the highest-converting audience, which addresses your concern about the different product audiences.
For your catalog ads in ASC+, the presence of dynamic catalog ads alongside your static/video tested creatives is an intentional feature of ASC+, designed to combine brand and direct-response performance.
The system’s objective is to deliver the ad most likely to convert a user, and a dynamically personalized product ad is often a strong performer, especially for people who have browsed your site.
This is not necessarily “biasing delivery” in a negative way, but rather the algorithm performing its job to maximize conversions.
Since you already have a very high-performing dedicated Shopping campaign running in parallel, you have a safe alternative.
If you are noticing that your manually tested creatives are struggling to gain spend, you can experiment by excluding your product catalog from the ASC+ campaign, or you can use the Catalog Ad option to only allow the AI to use your catalog for retargeting, which will reserve the ASC+ budget almost entirely for prospecting with your proven image/video creatives.
A critical component to ensure maximum performance from both your ABO and ASC+ campaigns is maintaining the highest possible quality of first-party data flowing into Meta, which is where server-side tracking becomes essential.
To future-proof your tracking and maximize your ASC+ performance, an excellent and cost-effective solution is to set up a server-side tracking stack using Facebook Conversions API (CAPI), your Shopify data layer, Google Tag Manager (GTM) in a server-side environment, and a dedicated server-side tagging service like Stape or a basic Google Cloud Platform setup.
This approach solves the data accuracy issue caused by iOS updates, ad blockers, and browser restrictions, which significantly undermine the effectiveness of the old Facebook Pixel.
When implemented, the Shopify API data feeds into a server-side GTM container, which then sends a direct, secure, and complete stream of events like ViewContent, AddToCartInitiateCheckout, and Purchase
By using server-side tracking, you bypass browser limitations and data loss, allowing you to achieve a much higher Event Match Quality score by consistently sending rich customer information (like email hashes and phone numbers) back to Meta.
This more accurate and complete conversion data is the fuel that Meta’s AI needs to make smarter, more profitable optimization decisions in your broad ASC+ campaign, effectively enabling the algorithm to overcome the problem of diverse product audiences by giving it cleaner purchase signals to model from.
Stape offers a cheap, simplified hosting solution for your server-side GTM container, making this advanced setup far more accessible and cheaper than managing a full Google Cloud or AWS instance, which provides a major performance edge for a minimal investment.
About The Author
