Meta has been pushing Advantage Plus Shopping hard for two years. Every rep pitch, every interface nudge, every account review ends with some version of "you should really be running more Advantage Plus." The pitch is convenient: hand Meta the keys, the algorithm will outperform your manual structure, you spend less time in ads manager, everyone wins. Sometimes it works. Often it does not. This is the decision log I apply before turning ASC on for a DTC brand.
Product feed is clean, titles and images audited
ASC works from the catalog. Garbage feed, garbage ads.
What ASC actually does
Advantage Plus Shopping Campaigns (ASC) is Meta's automation stack wrapped around a single campaign type. You provide a catalog, a budget, and some creative. Meta handles audience selection (including whether to target existing customers or prospects), placement allocation across Facebook, Instagram, Audience Network, and Reels, and creative rotation from the pool you uploaded. You lose granular targeting control. You gain, in theory, access to Meta's signal-rich optimization at a scale a single operator cannot match.
The pitch depends on Meta having high-quality signal to work with. If the signal is good, ASC compounds. If the signal is weak, ASC burns spend faster than a manual campaign because the automation cannot course-correct as well as a human operator watching daily.
When ASC is worth trusting
Three conditions need to be true simultaneously.
Condition one: your catalog is clean. Product titles, product descriptions, images, variant structure, and GTINs are all populated and accurate. Meta reads the catalog directly, so bad feed data becomes bad ads. If your Shopify product titles are like "SKU 4821 Red L" instead of "Women's Performance Running Shirt, Red, Large," ASC will underperform because Meta cannot read what it is advertising.
Condition two: you have enough weekly conversion volume. ASC works best at 100+ purchases per week minimum. Below that, Meta does not have enough events to optimize against and the automation oscillates. A DTC brand at $3M annual revenue running $150K/month across channels with roughly 30 percent of spend on ASC should be producing at least 100 ASC purchases a week. If it is producing less, ASC is too small relative to the brand and you should run manual campaigns until scale justifies automation.
Condition three: your server-side CAPI is clean and match quality is above 8. ASC is signal-hungry. If you are feeding it browser-only pixel data with match quality at 5, the automation is working from a compromised map. The DTC Stack Audit covers the CAPI and match quality questions, because these are prerequisites for ASC working at all, not separate concerns.
When all three are true, ASC often outperforms my manual campaign structure by 10-20 percent on blended MER. When any of the three is false, ASC underperforms.
When ASC is the wrong tool
Three patterns where I explicitly avoid ASC.
You are testing new creative concepts. ASC rotates creative based on early performance, which means the creative that wins the first 48 hours gets most of the spend. For creative testing, you need equal budget per concept so you can read comparative performance. Use a manual ABO campaign, not ASC.
You have meaningful gross margin differences across SKUs. ASC optimizes for conversions, not for margin-weighted revenue. If your $50 hero product has 60 percent gross margin and your $25 upsell has 20 percent gross margin, ASC will happily compound the upsell because it converts more easily. Your CPA looks great. Your gross profit drops. Manual campaigns let you set margin-aware audience and creative splits.
You need audience isolation for a new launch. If you are launching a product to a cold prospecting audience only, ASC will mix in existing customers and re-engaged audience segments in a way that makes the launch data unreadable. Use a manual prospecting campaign with audience exclusions. Measure clean.
The blended pattern that works
At DTC brands in the $3-8M range, I usually end up with a dual structure. ASC handles a portion of the prospecting pool where the conditions above are met. Manual CBO and ABO handle everything else: creative testing, geo-split campaigns, new product launches, retargeting. The CBO vs ABO decision log covers the manual side.
The ASC portion is typically 30-50 percent of paid social spend. Never 100 percent. The brands I see running ASC at 100 percent are almost always leaving margin on the table because they have abdicated the entire operating system to Meta's algorithm.
What to measure once it is live
Do not measure ASC performance inside ads manager alone. ASC's reported ROAS is particularly unreliable because the audience mixing makes attribution fuzzy. Three measurement layers that matter.
First, the blended MER across the account including ASC. If ASC is pulling MER up versus the three months before it launched, it is working. If MER is flat or down, the ASC ROAS claim inside the platform is suspect.
Second, gross margin per dollar of spend, not conversions per dollar. If ASC is shifting your product mix toward lower-margin SKUs, CPA goes down, gross profit goes down, ASC looks like a win in ads manager and a loss on the P&L.
Third, new customer rate. ASC's audience mixing often inflates retargeting-to-prospecting ratios, which inflates reported ROAS while starving the top of funnel. Look at new-customer purchases as a percentage of total ASC purchases weekly. If it is climbing relative to pre-ASC, you are buying new revenue. If it is dropping, you are recycling existing customers.
“ASC is not a campaign. It is an opinion about who should make the optimization decisions. If your opinion of Meta's algorithm is "they have more signal than I do," ASC makes sense. If your opinion is "they know less about my product than I do," it does not.
”
Rollback conditions
Set hard rollback rules before you launch ASC. Mine are:
- MER drops below a defined floor for two consecutive weeks. Pause ASC, run manual for one month, re-evaluate.
- New-customer rate drops below 40 percent of ASC purchases for two consecutive weeks. Pause ASC.
- Gross margin per dollar of spend drops more than 15 percent versus the pre-ASC baseline. Pause ASC.
Write these into your playbook before launch. Without rollback rules, ASC campaigns quietly drift toward what looks good inside ads manager and away from what is actually working on the P&L.
Does ASC work for small DTC brands under $1M annual revenue?
Usually not. Below $1M annual, you probably do not hit the 100-purchases-per-week threshold for ASC to have enough signal, and the overhead of setting up a clean feed may not be justified. Stick with manual CBO at smaller scales.
Should I cap ASC at a percentage of total spend?
Yes. I usually cap at 50 percent of Meta spend for brands in the $3-8M range. This keeps the manual campaigns as a baseline to compare against and prevents full abdication to the algorithm.
How do I tell Meta to exclude existing customers from ASC?
ASC includes an "existing customer budget cap" setting that lets you define a percentage of budget Meta can spend on existing customers. Set this based on your business model. For most DTC brands, 15 percent is a reasonable starting point.
Related reading
This decision log lives inside the paid social for DTC operators hub. The CBO vs ABO decision log covers the manual side. If ASC is underperforming because your data layer is weak, the underlying cause is usually the stuck Meta learning phase problem plus CAPI quality gaps. The DTC Stack Audit runs the prerequisite checks.