Question from Reddit user:
I’ve always thought you were supposed to select the event you want even if it won’t get a lot of conversions. But could it work much better to select an event that leads up to that event if it will mean much more data? Ex: Instead of choosing purchase for your event, selecting “view content” and have it be the button before people get to the final page where they either buy or don’t.
Answer from Nabil:
The short answer is:
Yes, it can absolutely work much better to select a higher-funnel event like Add To Cart or Initiate Checkout instead of Purchase if your purchase volume is too low for the ad set to exit the “Learning Phase,” which typically requires around 50 optimization events per week.
Optimizing for a higher-frequency event gives the Facebook algorithm enough data to learn what a good customer looks like, which can lead to more purchases overall, even though you are not directly optimizing for the final sale.
The only reason this strategy is necessary is because of data limitations, so you should simultaneously work to solve the root cause by implementing the Facebook Conversions API for more reliable and higher-volume Purchase data.
The long answer is:
Your idea is not only valid but is a common and necessary strategy used by performance marketers, often referred to as “optimization event shifting.” The primary goal of selecting an optimization event for a Facebook conversion campaign is to feed the algorithm the most relevant, high-quality data signal in a sufficient volume.
When you select Purchase and only get ten sales per week, the ad set is starved for data and remains “Learning Limited.” The algorithm struggles to find the patterns that lead to a sale, resulting in inconsistent performance and higher costs.
By shifting your optimization event to a higher-funnel event that occurs much more frequently, like Add To Cart or a strategically placed View Content on a high-intent page, you provide the algorithm with the required 50+ signals per week.
This allows the ad set to exit the learning phase, stabilize, and become more efficient at finding people who perform that chosen action.
While the algorithm is optimizing for Add To Cart, the users it finds are still highly qualified and more likely to progress to the final Purchase than those found by optimizing for a low-intent event like Landing Page Views.
However, be cautious with an event like View Content on a button click, as that may be too far up the funnel and could lead to a large volume of low-quality events that don’t translate to sales.
A good rule of thumb is to choose the highest-frequency event closest to the final purchase.
To address the fundamental data problem for long-term scale and to ensure you can eventually optimize for Purchase, you should implement the Facebook Conversions API using a server-side solution, such as Google Tag Manager running on a service like Stape or Google Cloud Platform.
This setup sends event data directly from your server to Facebook, which is far more reliable than the browser-based Pixel.
This server-side tracking captures conversions that the Pixel misses due to ad blockers, poor connectivity, and browser restrictions, thereby increasing the number and reliability of your Purchase events.
With a more complete and accurate Purchase signal, the algorithm gets better data quality and a higher volume of the event you truly want, significantly increasing the chances of successfully optimizing for Purchase without having to rely on the event shifting strategy.
About The Author
