The table below illustrates Facebook ad results for one of the companies I work with. They run website lead generation ads for their franchise locations across Canada and the United States. For the purposes of this post, I’ve highlighted a few key metrics: reach, website traffic, click-through rates, and leads generated for Android smartphone and iPhone users.
At the click-through level, we see very similar response rates for Android and iPhone users. Click-through rates are nearly identical at 0.52% and 0.51%. However, while 3.47% of clicks from Android users convert to leads, only 0.37% of iPhone users do. This discrepancy is a bit odd.
Underreported iPhone conversions
Is it likely that Android users convert at ten times the rate of iPhone users, or could this be a tracking issue? It’s likely that Facebook’s ad platform is not effectively tracking iPhone conversions. This is a consequence of Apple’s privacy updates – specifically, the introduction of AppTracking Transparency (ATT) in iOS version 14.5 and beyond. This is summarised below from Meta’s business blog.
On April 20, 2021, Apple confirmed that the release of general availability of iOS 14.5 will begin next week and it will start requiring apps in the App Store to obtain permission to “track” users across apps and websites owned by third parties for advertising and measurement purposes through its AppTrackingTransparency framework. Apple’s prompt requirement will prohibit certain data collection and sharing by businesses unless people opt in. This will create limitations for businesses that use Facebook advertising for mobile apps on iOS devices, as well as those that drive business results through their website.
What does this mean?
First, your results will likely be underreported, especially in regions where iPhone usage is prevalent. If we assume that iPhone users convert at the same rate as Android users, it would mean that approximately 261 leads were not reported in the data presented above—about 90% of the leads generated from iPhone users and 18% of the adjusted total.
The second issue is that Meta’s delivery systems are operating based on incomplete data. Using the table above as an example, the data shows that advertising reach is more than twice as high for Android users compared to iPhone users. Why is Meta doing this in markets where the iPhone has a larger market share than Android? Meta’s delivery system is probably optimising towards showing ads to Android users because of the limited amount of conversion data from iPhone users that’s being fed into its algorithms / machine learning models.
What should you do?
Work with your developer to implement server-side tracking. Mechanisms such as Facebook’s Conversions API can help capture more accurate data from iPhone users by communicating directly with your server, instead of relying on browser-based pixels. Unlike browser-based tracking, which depends on cookies and can be blocked by privacy settings, server-side tracking sends data directly from your server to Facebook, thereby bypassing browser-based tracking limitations.
Disclaimer: Web development is not my area of expertise. This recommendation is based on research and my experience collaborating with web developers, not on my personal experience in implementing these solutions.
