Conversion Modeling
What is conversion modeling?
When we talk about conversion modeling, we are referring to the use of machine learning to help us quantify and estimate attribution to specific marketing campaigns when a cohort of conversions cannot be tracked.
On iOS 14, this involves taking the behavior of users who have consented to ATT and projecting it to users who have opted out, in order to estimate campaign performance. This is accurate to a high degree of probability. This approach is fully compliant with Apple’s guidelines and is based on aggregated data only.
Why is conversion modeling important?
To accurately assess the effectiveness of your marketing campaigns, conversion modeling will become increasingly important in the coming months and years. It enables you to accurately measure campaign performance with the aggregated and anonymized data available to you on iOS 14.
Is a user’s consent still important with conversion modeling?
Yes. To accurately model conversions, you will need to have a large cohort of users who have opted in, and the more that opt-in, the more accurate your data. A developer’s primary focus should be to optimize for ATT first to increase the opt-in rate, then optimize for users who have opted out.
