Facebook is an unstoppable ad-serving machine.
Every day, marketers and business owners use the self-serve ad platform to get their business in front of new customers and drive sales.
But for most people the inner workings of Facebook remains a mystery.
Sometimes things work great.
Other times it doesn’t go as you expected.
Either way you’re not totally sure what’s going on.
It can be frustrating for sure.
Today I want to explain one of the big concepts that helps understand the Facebook Algorithm.
And that is the ‘Breakdown Effect’.
Have you ever ran an ad campaign where you were getting a low cost per result at the start but by the end of the campaign your costs had risen?
Maybe even pretty dramatically with a big budget?
That’s happened to all of us.
If you dig into the data it might seem like Facebook was shifting the budget to more expensive placements and ads.
But believe it or not, the ‘Breakdown Effect’ is actually the result of the Facebook algorithm getting you the best bang for your buck.
Let’s take a deep dive into the Facebook ad allocation algorithm and see exactly how this works.
Table of Contents
How Facebook Paces Your Ad Budget
When you set up a new Facebook campaign and assign a budget, Facebook’s automated delivery system has two big goals in mind:
- Getting the lowest cost results available.
- Ensuring your budget lasts for the full length of the campaign.
To do this, the algorithm paces your ad spend through a system aptly called ‘Discount Pacing’.
It starts by looking for the lowest possible bids that will bring you results.
If you only have a small budget, you might only get these low-cost results.
For bigger budgets though, these low-cost results can dry up relatively fast.
This means that as the campaign progresses, Facebook will transition your ad spend into the next best fit in terms of ad bids.
With this system, what you usually see is the bids by the end of your campaign are quite a bit higher than those at the end.
The reason why is because the lower-priced options have already been exhausted.
Explaining the Breakdown Effect
It is this process of pacing an automatic allocation of ad spend that leads to what we know as the ‘Breakdown Effect’.
There are three distinct steps in the automated process that go on without any input from you after you set the campaign.
Learning Phase
During the learning phase, the algorithm takes all the options you have selected for placement (like newsfeed ads, placement in stories etc.) and runs quick tests to determine which one will convert better.
This way, it is established which placements are going to be the best ‘bang for buck’.
Ad Allocation
Next, ads are allocated based on your preferences and what was revealed during the learning phase.
This begins with the lowest Cost Per Acquisition (CPA) placements.
There is, of course, a limit on how many of these lower-cost options are available when using a large ad budget.
The algorithm can make projections about how much the CPA on a certain placement is likely to increase based on the current spend and shift to other options before your ad spend becomes inefficient.
To make sure your campaign runs all the way until the end and utilizes the entire budget, there becomes a point where the ad spend must be shifted to other options.
Inflection Point
This point where the algorithm detects that the CPA for the first selected option is increasing beyond its feasibility is called the inflection point.
It is here where the algorithm makes the shift to other, sometimes more expensive CPA placements.
A real time adjustment like this saves money in the long term, but if you don’t understand the Breakdown Effect it might look like Facebook is shifting your budget to the more expensive placement.
The Breakdown Effect is real, but all you need to do is measure the success of your campaigns correctly to see that it is actually helping you achieve a better CPA overall.
It Comes Down to How You Measure Success
You can see the benefit of the automatic placements in your campaign when you look at the results through the correct lens.
Here are a few tips:
- Generally, when you are using CBO you should look at your results at the campaign level. This will show you the overall CPA rather than being bogged down in individual ad set or ad level.
- When using automatic placements without campaign budget optimization, you should be looking at your results at the ad set level as a whole, not the breaking it down by the placement on the different platforms (Facebook, Instagram, etc).
- If you have multiple ads in one set, you should also evaluate results at the ad set level (not the ad level, since the ads work as a whole in the ad set).
Doing this will give you the real picture of what running a campaign with automatic placements has accomplished considering the Breakdown Effect.
Maximizing Results Through Automation
A very common reaction many Facebook marketers have is to assume they can do better than Facebook’s algorithm.
We try to build more complex campaigns to outsmart the machine.
In reality, it is unlikely or even impossible to match the instant, real-time adjustments of the automated ad allocation system.
The area where you CAN have an impact is in the way you set up and optimize your ad campaigns.
Choosing the right campaign structure and having the best ad creatives is where we have found our time is most valuable.
Optimizing Your Campaigns
The days of cheap clicks on Facebook are gone.
It is better to let the algorithm do its thing.
But there are a few ways you can optimize your campaign set up to help:
- Test lots of creatives to make sure that yours are compelling and engaging with your target audience. This is your biggest lever to create massive success
- Test new audiences and targeting
- Test a full funnel strategy with top of the funnel, middle of the funnel, and bottom of the funnel retargeting ads.
Running Facebook ads at scale can be hard work, but with the right set up you can work with the Facebook algorithm to get better results.
Want Major Impact Media to launch & scale your Facebook Ad campaigns? Click Here To Get Started
0 Comments