A lot of people ask me how automated campaigns like PMax and Advantage+ are doing in terms of performance, and whether they’re likely to be incremental.
The results we’ve seen are by no means conclusive, but I have seen outsized results as measured by the platform. Unfortunately those results don’t always translate to incremental uplift, partly because these automated campaigns are great at optimization.
They seek the cheapest conversions, which usually come from brand or retargeting people who are already likely to buy. The problem is that the system can’t estimate incrementality: i.e. how many of those people would have bought anyway had they not seen the ad. This is a common bias that it’s important that marketers correct for by using independent measurement like MMMs.