The Solution to iOS 14.5 Restrictions

Meta's advanced matching features gained much popularity after the launch of iOS 14.5. Like the Conversions API, these features help mitigate the lower volume of events received due to privacy restrictions.

For many professionals, this gave the impression that this tool was developed precisely for this purpose. In reality, the existence of Advanced Matching precedes Apple's launch. Additionally, it neither has the capability nor the intention to "solve" or "bypass" Apple's "blockages."

In fact, Apple is clear in its documentation (in English) regarding the App Tracking Transparency framework (ATT):

So, even if the iOS user agrees to the site's terms and provides their data in a form, they still need to have granted permission to the ATT to be tracked. If the user rejects Meta's app request, providing user data for advanced matching will not be effective for attribution.

Enriching User Profiles on Meta

Another common myth among professionals is regarding the use of user data sent to Meta. There is a common belief that the benefits derived from advanced matching are in the improvement of Meta's database. This would come in the form of a richer internal user "profile," containing information received in different instances where the user interacted with a form.

In other words, this "enriched" profile would contain various different emails or more updated addresses from the same user, all coming from various interactions with sites that sent data to Meta.

This perception is also unfounded, and Meta's documentation is clear on this point:

In other words, the data sent is used solely to identify the user and enable event attribution to a click. Whether or not this attribution is possible, the data is discarded.

"Empty" Conversions and Unidentified Users

One of the most common myths about using or, better put, not using advanced matching is that many professionals believe that recent privacy restrictions have reduced Meta's ability to attribute events occurring on the site to a click occurring on Meta. However, they also believe that Meta registers conversions in its reports without identifying the users who performed them. This problem, they think, would be solved with the use of Advanced Matching.

This belief is incorrect because, for a conversion to be displayed in the report, Meta needs to know which ad, ad set, and campaign to attribute the conversion to. This information is obtained by determining a click that led to the site visit. From this click record, it is also possible to determine the user.

One factor that fuels this belief is the Modeled Conversions feature, which uses statistical modeling in conversion attribution. Contrary to what many believe, this feature does not create phantom records of "empty" conversions in reports. Instead, real events with partial or missing data are used. Combined with data from user groups with similar behaviors, it is possible to estimate the origin details of this conversion.

Conclusion

These, among many others, are some of the main myths about advanced matching that persist in digital marketing communities. Misunderstanding how this feature works creates difficulty in estimating the impact of its implementation. It is common for professionals to make inaccurate decisions, believing that one tool or another can solve problems that were never the focus of their development.