What can $1 million dollars of ad spend teach advertisers about Campaign Budget Optimization?
The quick summary: Campaign Budget Optimization is now a best practice that all Facebook advertisers should embrace.
As Facebook reported in February, Campaign Budget Optimization (CBO) will soon become the fixed default for all campaigns in all ad accounts, starting in September 2019. Notifications were sent to ad accounts immediately affected by the change, with the full expansion expected to roll out over a multi-year period.
Facebook is still finalizing timing details, as advertisers provide feedback.
To plainly state the obvious, a migration to CBO is one of the most significant changes to Facebook advertising in years.
Since the feature first became available, we have tested across our client base to understand the true impact. Over the past six months, we’ve spent over a $1 million proving out CBO as a budgetary tactic.
For the study, we focused on eleven D2C clients spending against traditional and CBO Conversion-objective campaigns over a 6-month period (September 2018 – February 2019).
While we’ve tested the CBO tactic since the feature first become available, we originally found Facebook’s CBO 1.0 version limited campaign spend and didn’t yield results greater than our manual optimization.
Several one-off tests months after the initial failures found Facebook had tweaked the tool’s optimization ability and CBO had become a viable tactic. During our 6-month sample time period, the eleven clients were fully invested in testing CBO.
To evaluate our campaign performance, driving revenue and quality traffic were the primary campaign goals.
All revenue-focused metrics (ROAS, CPA, CVR) were positively impacted by CBO, in comparison to ads run with traditional ad set budgets.
- Click-thru rate increased by 13%, but CPC and CPM did negatively inflate.
- The CPC variance was negligible with a $.03 difference between CBO and traditional campaigns.
- A 16% CPM increase should give advertisers pause, especially brands focused on impressions and brand-focused metrics.
The CPM increase is credited to the singular prioritization of ad sets achieving the chosen Conversion Event, rather than advertisers manually optimizing towards a blended campaign goal.
For example, CBO will allocate more budget to ad sets generating revenue if Purchase is the Conversion Event, regardless of any secondary campaign factor like CPM.
What We Learned
After spending a $1 million dollars, here’s what we’ve learned:
- Segment campaigns by audience type
- Set audience budget minimums to maintain control
- Branded metrics are negatively impacted
Segment Campaigns By Audience Type
The main testing takeaway is to segment campaigns by audience type, in order for CBO to work properly.
Prospecting and retargeting audiences should have different objectives. Prospecting audiences focus on driving high-quality traffic, whereas retargeting audiences generate revenue.
To achieve the above goals, the best practice is to:
- Use a View Content or Add to Cart Conversion Event for prospecting
- Use a Purchase Conversion Event for retargeting
While optimization events cannot be combined into a single campaign with CBO, the end goal for each audience type varies. By combining audiences goaled with an upper funnel objective with audiences focused on purchases into a single campaign, CBO will not properly allocate budget and lose efficiency.
If the optimization event is Purchase, CBO heavily prioritizes spend against ROAS-positive retargeting audiences over efficient traffic-driving prospecting audiences. While boosting retargeting spend is a short-term tactic for immediate revenue, investing in high quality prospecting is a winning long-term strategy for brands focused on growth.
Our recommendation is to split prospecting and retargeting audiences into separate campaigns with unique campaign budgets to achieve the best results.
Set Audience Budget Minimums
Budgetary control isn’t completely taken away from advertisers with this change. Budget minimums and maximums can be placed on the ad set level with CBO.
Placing spend minimums isn’t a guarantee the audience will spend the minimum, but it does indicate spend should be spread more evenly across ad sets.
In our tests, we found huge swings in media spend allocation across audiences, which caused issues determining true success or failure.
Five-dollars spent on one ad set vs. $300 spent on another doesn’t provide enough data to make optimization judgments or long-term strategic audience planning.
The recommendation is to place audience minimums to circumvent drastic budget prioritization.
Branding Metrics Decrease
It’s no secret CPMs are rising across every facet of the digital marketing landscape. However, a 16% CPM increase, in addition to the expected social CPM inflation, is concerning.
While a CPM increase may not impact revenue metrics today, decreased impressions can affect brand equity and revenue long-term.
The majority of advertisers focus on the immediate impact on today’s revenue, but tomorrow’s revenue will decrease with such a singular focus. In addition to immediate revenue-generating tactics, the ideal social marketing funnel should include brand awareness tactics like the brand awareness objective, investing in influencer marketing, and strong video creative.
Conversion-focused campaigns may see impressions dip with CBO, so adding true upper funnel tactics into the media mix can help offset the results.
Testing Campaign Budget Optimization
For advertisers who haven’t tested CBO it’s recommended to run a clean test, use split testing, and properly measure the benefit of using this product over more manual efforts. It’s important to understand the impact both on scale and efficiency when looking at this test.
To set up clean CBO tests, we recommend breaking out campaigns by audience type, setting budget minimums as needed, and maintain a full-funnel approach to counteract a CPM increase.
Campaign Budget Optimization is going to become Facebook’s fixed default. All Facebook advertisers across the globe should be testing CBO to determine performance variance for their ad accounts.
As an advertiser, I’m always wary of significant changes to my tried-and-true social marketing methods.
While I want to believe updates are made with the best intentions, I haven’t always found Facebook’s “new thing” to be beneficial for clients, consumers, or my advertising levers.
Hopefully, our testing results give a clear direction to upcoming media plans and calm the fears of fellow Facebook marketing professionals.