Optimizing for the Right Conversions on Meta

The Challenge

Tracksmith, an independent running brand that sells performance running gear, was determined to make Meta a more effective cog in their overall marketing program. According to Rockerbox, the multi-touch attribution modeling platform that Tracksmith relies on, the ROAS attributed to Meta ads was falling well short of their brand goals. 

The core problem was that Meta was optimizing off of the wrong conversion data. The “Add to Cart” conversion was being used, under the assumption that driving more upper funnel conversions would give Meta more information to optimize from. In reality, this approach was dramatically depressing their Meta campaigns’ actualized ROAS and capping their ability to scale—to the extent that the client was questioning whether Meta was even an effective channel for their business. 

ADM knew that Tracksmith had the potential to find success on Meta: We just had to recalibrate their campaign conversions. 

The Solution

When the end goal of a campaign is to increase purchases on the client website, it needs to optimize for that specific conversion. Although optimizing towards Add to Cart events is a common tactic for low-conversion-volume accounts, it leads Meta to deliver users who may complete that action without actually following through on a purchase. This drives up CPA and depresses actualized ROAS.

Optimizing for the end conversion is a best practice, though it’s only doable if that conversion delivers enough volume. Our team analyzed the number of conversion events for Purchases across a 30-day period and determined that they exceeded the minimum threshold needed to inform Meta’s algorithms. We began converting campaigns to this approach, running pre/post-tests in phases across the account to minimize short-term impact.