Question from Reddit user:
I have an account with a relatively long purchase conversion funnel, and we get many more leads than purchases.
The true goal is purchases. Should I launch my FB campaigns to optimize to purchases and have very little data to work with? Or should I set them up to optimize to leads and then just see how many purchases they also get?
In my POV it comes down to whether Facebook considers lead events as food for the algorithm while optimizing to purchases.
In other words, can the campaign optimize to purchases but also automatically reference how many leads a particular ad, etc is driving to know when it’s on the right track?
Google conversions are so much more intuitive to work with.
I wish I could just set up these campaigns to optimize to both actions like I can in Google.
Answer from Nabil:
The short answer is:
You should launch your campaigns optimizing for the most frequent, high-intent action right before the final purchase that still allows Facebook’s algorithm to exit the “Learning Phase” – this is likely to be an event like Add To Cart or Initiate Checkout, not the lower-volume Purchase event or the less intent-driven Lead event.
The algorithm does not automatically use the Lead data as a positive signal when you are optimizing for Purchase; it is hyper-focused on the single event you select.
However, the best solution to get all the data you need for better optimization on your long funnel is to implement the Facebook Conversions API via a server-side tagging solution.
The long answer is:
Your intuition about the low volume of Purchase conversions is correct, and optimizing directly for Purchase with very little data will almost always keep your campaigns in the “Learning Limited” status, leading to poor and inconsistent performance.
Facebook’s machine learning needs a steady stream of data – ideally at least 50 optimization events per ad set per week – to reliably find the right people.
Since you have a long funnel, your Purchase volume is too low, so you need to move “up the funnel” to a more frequent, yet still high-intent, event.
For e-commerce, this is often Add To Cart or Initiate Checkout.
For your specific long funnel, you should look for the high-volume event that immediately precedes the point of a customer being sales-qualified.
This gives the algorithm more “food” to learn with, allowing it to exit the learning phase and optimize more effectively.
You should run a conversion campaign targeting this higher-funnel event, while still tracking the actual Purchase event for reporting and to monitor your conversion rate from your chosen mid-funnel event to the final sale.
The core challenge is that the algorithm will only optimize toward the single event you choose; it will not automatically consider leads a positive step in a campaign set to optimize for Purchase.
The machine learning model is designed to find users most likely to complete that exact specific action, and it ignores other standard events for that primary optimization goal.
This is why setting up the Facebook Conversions API using a server-side solution like Google Tag Manager’s server-side container hosted on a service like Stape or Google Cloud Platform is an excellent solution for your exact problem.
The Conversions API allows you to send event data directly from your server to Facebook, bypassing issues like ad blockers and browser restrictions that cause your browser-side Pixel to miss up to 20-30% or more of your conversions.
By implementing server-side tracking, you increase the reliability and accuracy of all your event data, including your low-volume Purchase events.
A more complete and accurate count of your Purchase conversions is crucial because it significantly improves the Event Match Quality, which is a score Facebook uses to determine how well it can attribute the conversion back to the user who saw your ad.
Higher match quality on all events, especially Purchase, gives the algorithm better data to train on and increases the chances that your campaigns can successfully optimize for Purchase – even if the raw volume remains relatively low.
You can also leverage the Conversions API to send deeper-funnel and offline data back to Facebook, for example, sending a custom event like Qualified Lead from your CRM after a sales call, which is a much stronger signal than the initial front-end website Lead event.
This enriched data can further improve the algorithm’s understanding of a high-value customer, which is the ultimate goal in a long sales cycle.
About The Author
