What is Facebook’s Simplified Campaign Structure? When getting started with advertising on Facebook, your campaign structure needs three parts to run, and knowing how they work together will help your ads run the way you want, and reach the right audience. These components are:
In February of last year, Facebook released its “Best Practices” update for structuring campaigns. This “many campaigns and many ad sets” account structure worked previously when AI was not managing account optimization. All those separate ad sets let advertisers test dozens of different combinations of audience settings, placement settings, and more. It was a manual way to find the most profitable niches in the advertising landscape. That structure worked well if all you have humans managing your campaigns.
As Facebook advertising has evolved, the algorithms have taken more control, automating more of intraday campaign management. It’s become clear that the AI will get optimal performance from less “campaigns and ad sets” in the account setup.
As such, advertisers will benefit by restructuring their campaigns based on Facebook’s new Simplified Campaign Structure, which is a template for how to scale your Facebook advertising. Here we share more on how the algorithm enhancements work.
With Facebook advertising in 2019, the algorithm can manage your intraday account changes better than you can. The algorithm is better at tracking, interpreting, and acting on the dozen or more “signals” generated by your ads hundreds of times a day.
Advertisers should let Facebook’s machine algorithm take over account optimizations. This gives you less control, but if you’ve chosen your campaign goals well, the account’s performance will improve.
Here’s why that means letting go of all those ad sets: Because the algorithm can crunch data so effectively, it doesn’t need all the “on/off” switches that human managers use different ad sets for. It actually operates better with a simplified account structure.
The simplified structure also allows your campaign to get out of the “learning phase” faster and stay in the optimized phase longer.
Facebook’s learning phase is a type of slowed-down campaign status while Facebook is gathering all the necessary data to deliver your campaign as stable as possible. It gets triggered whenever there’s a “significant change” made to a campaign, or a campaign has just launched. Most advertisers want to minimize how long their campaigns are in this status because campaign performance typically drops about 30% during the learning phase.
The learning phase is just another example of how Facebook ad management has changed now that the algorithms manage the platform. After the Best Practices update, fewer changes and giving more control to the algorithm resulted in better performance.
While performance slows down during the learning phase, important things are happening. The algorithm is calibrating the ad delivery system for the “new” campaign, new ad set or bid and budget changes so it will achieve optimal performance. Nearly a dozen different signals are being measured, and the system is also taking into account the ad’s history, the campaign’s history, the account’s history, and the user’s history… ultimately, so the campaign can perform even better.
After about 50 conversions per week per ad set, Facebook’s advertising algorithm will have enough data to shift a campaign out of the learning phase. At this time, your ad set can experience fewer performance fluctuations. You can make an informed decision about your ad set. If you’re satisfied with your results, you can let it keep running or increase its budget. If you’re unsatisfied, you can edit the ad set to try to improve its performance or pause it.
This is an important aspect of the learning phase to understand: Ad Sets aren’t put into the learning phase for any particular length of time. They’re put into the learning phase until they’ve accrued enough data that the system will release them into full delivery.
Because how the system works, if an advertiser specifies too many criteria for a campaign, it’s possible to inadvertently require the system to keep the campaign in the learning phase for longer than might have been necessary.
In other words, the more limitations you put on a campaign’s performance, the longer the learning phase may be.
Of course, almost every campaign has performance requirements. Maybe it has to work within a tight budget, or needs to generate conversions at a particular price. Just go easy on how many constraints you set. The more flexible (or “liquid” as Facebook would say) you can be with campaign settings, the more room you allow for the algorithm to work.
Fortunately, there are several ways to get the results you want, to still minimize the learning phase, and to get great results, and they’re all considered Facebook best practices.
Increase the size of the audience you’re targeting in order for Facebook to gather enough data more quickly. For instance:
If you aren’t happy with your campaigns’ performance, optimize your campaigns for an event further along in your sales funnel. For example, instead of optimizing for clicks or web page views, optimize for an event that only happens when someone initiates checkout. That’s the type of event that truly drives revenue.
If you optimize a campaign for an event further along in the funnel, you’ll probably be optimizing for a conversion event that happens less often than, say, the first click. That’s okay, but it will slow down how quickly a campaign can accrue signals, and so it will keep your campaigns in the learning phase longer. In our experience that extra wait is worth the upfront cost.
Implement the following:
Choosing a bid strategy is really an exercise in balancing between more conversions or more control. The algorithm is pretty good at its job, so we do recommend letting it do the heavy lifting of account management, especially if you’re willing to spend some money getting your campaign up to speed.
There are three bid strategies to pick from:
The Facebook algorithm is taking over and it’s important to understand how the algorithm works and to give it the data, the budgets, and the creative assets needed for optimal performance. The Facebook algorithm will take away the budget lever from humans when Campaign Budget Optimization becomes mandatory in September 2019.
Keep in mind that machines can’t do creative development yet, and that’s now the biggest competitive advantage advertisers have. User acquisition managers should invest in creative development and creative testing. They should keep their skills current and stay agile. We expect many more changes from Facebook and Google in the coming months and years.