Common Knowledge | Wondering something?
Facebook: The Breakdown Effect
Below is a one-sheeter from Facebook explaining this phenomenon.
Introduction
This page explains the “breakdown effect” and how you can properly judge the performance of your campaigns when using automation between ad sets, placements and ads.
A common point of confusion when evaluating Facebook ads reporting between ad sets, placements, and ads is that our system appears to shift impressions into underperforming ad sets, placements or ads. In reality, the system is designed to maximize the number of results for your campaign dependent on what ad set optimization you choose. This understandable misinterpretation is called, “The Breakdown Effect.”
To understand why this is common in ads reporting, we need to dig into how the delivery system works.
Discount Pacing
The delivery system uses both bidding and pacing to determine how to deliver your ads. Pacing is the mechanism that allows the budget to last the entirety of the schedule. Discount pacing lowers your bid when appropriate to help get the lowest-cost results available, balanced with ensuring we spend your full budget by the end of the campaign.
Over the duration of the campaign, there are only so many optimization events you can get for your budget. When your budget is smaller, it’s easier for our system to spend all of it on the least-expensive ones using discounted bids. When your budget gets bigger, we search for more optimization events to spend it on to continue delivering results. Once we’ve gotten all the least-expensive ones, we have to go for more expensive ones. This means we may not be able to discount your bid as much (or at all). Your average cost per optimization event typically goes up at this point, but we’re still optimizing for the lowest costs possible.
Next, we need to understand how Facebook uses machine learning automation to help drive strong results for you based on your inputs (creative, budget, audience, objective, etc).
Automation
Facebook’s delivery system uses machine learning to automate the delivery and management of impressions between ads, ad sets, and placements to maximize your results. This automation drives more results compared to manual campaign management and cuts down on user time spent managing Facebook campaigns.
The Breakdown Effect Explained
The breakdown effect occurs at the intersection of our discount pacing approach and our system’s automation. It’s easiest to understand this effect by using a fictitious example. Keep in mind, this example is simplified to illustrate the effect.
Let’s say you choose to run a campaign utilizing the conversions objective. You choose two placements to deliver your one creative asset to – Facebook Stories and Instagram Stories. The total budget is $500.00 USD for a single ad set using ad set level budget.
When the campaign begins, our system starts to deliver ads to both placements to see which will drive the most efficient results for your target audience — this is called the learning phase.
Facebook Stories starts out driving cheaper acquisitions, but then our system identified an inflection point at which the cost per acquisition (CPA) on Facebook Stories begins to exceed cost per acquisition on Instagram Stories.
The cost per acquisition on Facebook Stories is $0.35 on the first day, compared to $0.72 on Instagram Stories. However, as the campaign continued Instagram Stories received more budget, even though it still has a higher cost per acquisition. At the end of the campaign, Instagram Stories delivered significantly more budget compared to Facebook Stories, even though Facebook Stories originally had a lower cost per acquisition.
Final results at the end of the campaign:
| Placement | CPA | Spend |
|---|---|---|
| Instagram Stories | $1.46 | $450 |
| Facebook Stories | $1.10 | $50 |
| Day | Instagram CPA | Facebook CPA |
|---|---|---|
| 1 | $0.72 | $0.35 |
| 2 | $0.78 | $0.50 |
| 3 | $0.88 | $0.75 |
| 4 | $1.02 | $1.10 |
| 5 | $1.20 | $1.55 |
| 6 | $1.42 | $2.10 |
| 7 | $1.68 | $2.75 |
| 8 | $1.98 | $3.50 |
| 9 | $2.32 | $4.35 |
| 10 | $2.70 | $5.30 |
Prior to the inflection point (where Facebook costs were still lower than Instagram), the system (for example) may spend $50.00 on Facebook Stories and $50.00 on Instagram Stories to test. Since the system can detect that Facebook Stories CPA was rising faster than Instagram Stories, it shifts the remaining budget of $400.00 to Instagram Stories to have a cheaper CPA over the duration of the campaign. In the table, you’ll see that prior to Day 4, Facebook Stories achieves a lower CPA; however, the costs would have grown faster than Instagram Stories. As you see above, Facebook Stories reached up to $5.30 by Day 10 while Instagram Stories was delivering at half the cost.
Why Did Our System Do This?
At the start of the campaign, the system began delivering ads to both platforms to explore where the lowest cost opportunities were. Early on, the lowest cost results were garnered because the system front loads low-cost results through our discount pacing approach. In the example, the system recognized that although Facebook Stories was driving the most efficient results initially, it predicted the cost was going to increase throughout the duration of the campaign. Based on the anticipated rising costs, the system was able to pivot and shift the budget to Instagram Stories in order to drive a more efficient average CPA for the duration of the campaign. As a result, the system made the right decision and ultimately drove more conversions.
This is where the reporting may not match your expectations. If you had only judged the decision by looking at the CPA in Ads Manager, it would show less budget went to the lower average CPA placement, Facebook Stories. This may be confusing because it appears the decision was incorrect; however, the system pivoted the budget in real time to Instagram Stories which drove more results.
How to Evaluate Results
When using multiple ad sets, placements and ads, the best way to measure the effectiveness of your campaign is by evaluating your results at the aggregate level. For example, when using campaign budget optimization, the system will optimize to deliver the highest number of results between the ad sets – meaning that to properly evaluate success, you must evaluate at the campaign level.
How to evaluate results with each automated solution in ads manager:
- When using campaign budget optimization, always evaluate your results at the campaign-level.
- When using automatic placements (without campaign budget optimization), evaluate your results at the ad set level.
- When running multiple ads in 1 ad set, evaluate your results at the ad set level.
Conclusion
The Breakdown Effect has led to some confusion when interpreting ads manager campaign results. It’s a trade-off designed to utilize automation and discount bidding. It may seem misguided without context, but ultimately it helps drive significantly more value for folks using our advertising system. The important thing to remember is that it will usually work to your favor when you create flexible campaigns through lowest cost bidding (discount bidding), automatic placements, and campaign budget optimization features that work together to maximize performance.