BOD model
At the end of 2019, Facebook launched a new budget optimization model called CBO(Campaign Budget Optimization). This model allows you to set a budget for the campaign and let the platform distribute the budget among the ad sets, prioritizing the best performing ones.
However, it is necessary to pay attention to a few points so that the algorithm can work optimally.
That's why we at Pareto have put together a list of best practices for the BOD, with some issues that you should pay attention to so that this budget optimization model works in the best possible way.
Start Using the Lowest Cost Bidding Strategy
This type of strategy will make Facebook look for the lowest cost conversions within each ad group.
By using it, you won't limit the algorithm's performance right from the start and consequently it will achieve the highest number of conversions by spending as little as possible within that audience.
With the accumulation of data, it is possible to evaluate changing the bidding strategy to the one that best suits your objectives with a given campaign. We'll see more about bidding strategies in this article.
Use Segmentation with Audiences of Similar Size
As we mentioned earlier, the user will define the budget at campaign level, and let Facebook distribute it among the ad sets. That's why it's first necessary for the audiences to be of a similar size.
This is because, if one audience is much larger than the other, the algorithm will naturally invest more in the adset with the largest audience. Since the chance of finding the users with the cheapest conversion within that pool is greater.
And this will limit the learning of the smallest audience, and consequently we won't be able to know their real potential.
Use Segmentation with Audiences with Similar Characteristics
Another extremely important point when organizing audiences using the BOD is the characteristics of each of them, both in terms of size and qualification In terms of qualification, since a bottom-of-the-funnel audience is more qualified than a middle-of-the-funnel audience, if you place them in the same campaign, the algorithm will naturally prioritize the hottest audience. For example: if a cart audience is together with a "home" audience, the algorithm will naturally optimize the former, since it is closer to purchase (considering this objective in the campaign).
Leave Lookalike Audiences in a Separate Campaign
In some cases, lookalike audiences perform better than interest audiences, which is why it's important for them to be in a separate separate campaign since their characteristics are not necessarily similar to those of the audience of interest.
Another reason for this separation is that it makes it possible to find the best lookalike for your account. Test audiences with different levels of similarity to find the one with the best performance for your objective.
Important: don't forget to check the size! Make sure they don't differ too much so that the algorithm can work as well as possible.
Always Analyze Results at Campaign Level
In the CBO, you need to analyze the result at campaign level. This is because the algorithm will optimize for the best overall result. If the results are analyzed at the adset level, we can jump to conclusions. The graph below explains this:
Source: https://www.facebook.com/business/m/one-sheeters/breakdown-effect
In this case we have a campaign that works with two ad groups. One for Facebook and one for Instagram. Notice that at the start of the campaign Facebook appears to be more attractive in terms of CPA. However, in the long term, Instagram has the lowest CPA.
This graph shows us that the algorithm will always look for the best CPA at campaign level, and not necessarily at adset level. If the user analyzes by adset, they could pause Facebook in the first few days, and in the end they wouldn't have the best possible CPA at campaign level.
This practice does not exempt us from monitoring the audiences present in the Ad Sets in order to propose improvements once the campaign has accumulated enough data.
Conclusion
This article has shown us that using good CBO practices can be an excellent alternative for optimizing Facebook Ads campaigns. With it, you don't have to worry about the budget at the ad set level, which means that optimizations can be focused on other points.
Do you want to know how to improve your Facebook campaigns even more and no longer have to worry about budget issues? Get in touch with our team and we'll help you out.
