Meta ASC Campaign Cannibalized My BAU Ads Completely
If you run Meta ads, you may have seen your Advantage+ Shopping Campaign (ASC) take over all your business-as-usual (BAU) campaigns. This is a common issue. The system learns fast, but it can also cause complete attribution cannibalization. In this article, I explain why it happens and how to fix it.
What Is Meta ASC and BAU Advertising?
Meta's ASC is an automated campaign type. It uses broad targeting and dynamic creative. BAU campaigns are your standard manual campaigns. They have more control. Both can overlap if you do not set clear boundaries. ASC relies on machine learning to find buyers, which makes it very powerful but also very hungry for impressions.
BAU campaigns usually target specific audiences, placements, and products. They often have a clear structure. When ASC enters the same account, it can start competing with those manual campaigns. The result is that one campaign wins and the other seems to lose. Most of the time, ASC wins.
Why ASC Cannibalizes BAU Ads Completely
ASC uses a different bidding system. It often gets better delivery because it can optimize across many placements and audiences. Meta prioritizes ASC due to its automated nature. This pushes BAU out of the auction. The attributed conversions flow to ASC, making BAU look dead even when it is still doing useful work.
Audience Overlap
When ASC and BAU target similar people, they compete. Meta does not always de-duplicate the overlap. The same user sees both campaigns. ASC usually wins because it has more flexibility. This starves BAU of impressions and creates a false impression of failure.
Attribution Bias
Attribution models often give all credit to the last click. ASC often touches the same user later in their journey. So ASC gets the conversion. Your BAU campaigns may have assisted, but the report shows zero. This is why understanding attribution modeling is critical before making changes.
Budget Allocation
ASC has a dedicated budget. If ASC budget is too high, it drains the account. BAU campaigns get leftover. This causes complete cannibalization of reporting, even if incrementality is not real. The platform will always spend where it expects the cheapest conversions.
How to Diagnose ASC Cannibalization
Before making changes, verify the problem. Use Meta's reporting tools. Look at attribution windows, overlap, and placement data. Do not rely only on the default view.
- Compare ASC attributed conversions vs BAU conversions over the same period.
- Check frequency and audience overlap reports between ASC and BAU campaigns.
- Run a geo holdout test: turn off ASC in one region and compare total conversions.
- Look at placement data to see if ASC is taking over the same placements as BAU.
How to Fix Meta ASC Cannibalization
You can stop ASC from eating BAU. The goal is to separate audiences and attribution. Use one or more of the following strategies. Do not apply all at once. Test each change and measure the impact.
Separate Audiences
Exclude your BAU remarketing and prospect audiences from ASC. This forces ASC to find new people. It also prevents direct overlap. You can use customer lists and custom audience exclusions inside ASC settings.
Use Exclusions and Caps
Set bid caps or cost controls in ASC. This slows down its aggressive delivery. You can also exclude existing customers if your BAU focuses on retention. This keeps ASC from stealing cheap repeat buyers.
Reduce ASC Budget
Start with a small ASC budget. Increase it only when you see incremental lift. Do not let ASC have more than 30% of total budget initially. This prevents ASC from dominating the auction and leaving no room for BAU.
Use Different Attribution Windows
Compare ASC with a longer attribution window. ASC often wins on 1-day click but loses on 7-day click. This helps you see true contribution. If ASC only looks good on short windows, it is likely stealing credit from BAU.
Key Takeaways
- ASC cannibalizes BAU when audiences overlap and budgets are unbalanced.
- Diagnosis requires looking at incrementality, not just platform attribution.
- Fix overlap with exclusions, separate audiences, and bid caps.
- Reduce ASC budget until it shows true incremental lift.
- Use holdout tests to validate real performance before scaling ASC.
Comparison Table: ASC vs BAU Campaigns
| Feature | ASC Campaign | BAU Campaign |
|---|---|---|
| Control level | Low, automated | High, manual |
| Targeting | Broad, machine-driven | Manual, specific |
| Budget flexibility | High but may dominate | Set per campaign |
| Attribution credit | Often overstates | Understates due to overlap |
| Best use | Prospecting new customers | Retargeting and controlled segments |
Frequently Asked Questions
Why does ASC get all my conversions?
ASC often touches users last because it has broad reach and aggressive optimization. The attribution model credits the last click, so ASC looks like the hero. This hides BAU assistance and makes BAU appear ineffective.
Should I turn off ASC?
Not always. ASC can drive incremental reach if you separate it from BAU. Turn it off only if holdout tests show no incremental lift. Otherwise, optimize it with exclusions and budget caps.
How do I know if cannibalization is real or just reporting?
Run a geo holdout or audience holdout. Compare conversions in regions with ASC versus without. If total conversions stay the same, ASC is just stealing credit. If total conversions increase with ASC, it is adding incremental value.
Can I stop ASC from targeting BAU audiences?
Yes. Use audience exclusions and customer list exclusions. Also set placement controls if available. This forces ASC to focus on new audiences instead of overlapping with your manual campaigns.
How much budget should ASC get?
Start with 10–20% of total budget. Increase only when you see incremental lift. Avoid giving ASC more than 50% until proven. This keeps BAU campaigns alive and lets you measure true performance.
Final Thoughts
Meta ASC is powerful but can cannibalize BAU ads completely. The fix is not to panic. Separate audiences, reduce budgets, and measure incrementality. This way you get the best of both worlds: automation for scale and manual control for predictable results.