Every PPC specialist has likely experienced this scenario. You create a PMAX campaign, set the tROAS and budget, fill in the assets, and rely on Google’s algorithm to do the work. Sometimes, it works immediately, and everything runs good. However, the campaign’s performance often doesn’t meet expectations, and we need to optimize it. But how do you optimize a PMAX campaign when it’s essentially a black box, and you can’t see enough statistics to make data-driven decisions?
That’s where we use our own custom PMAX script, which we developed in-house. In this case study, we’ll show how this script helped us when our campaign’s performance dropped.
But first, what is the PMAX script?
The PMAX script allows us to view statistics based on where the ads were placed. In other words, we can easily see what percentage of costs went to Shopping, Video, Display, or Search. Beyond costs, we can also view metrics like conversions, conversion value, clicks, and impressions broken down by placement.
This is an example of how the visualization looks. You can see five different graphs, each showing a different metric, with the data color-coded based on the individual placements. The total amount for the selected period is represented in black.
The main advantage of this script and visualization is that it allows us to view data over longer periods, typically 180 days. However, you can adjust the time frame to shorter or longer periods as needed. The data can be analyzed by day, week, or month, giving us flexibility to review both short-term and long-term trends.
Case Study
We noticed a weaker performance in our PMAX campaign and decided to loosen the tROAS. Two days later, we observed a significant spike in costs, so we checked the report and discovered that our video ad spend had surged dramatically.
When investigating the cause, we found that many products had accidentally been removed from the custom label. We immediately resolved the issue, and our YouTube ad costs stabilized in the following days.
In this example, you can see how two changes made at the same time—loosening the tROAS and the unintentional removal of products from the label—created such a cost spike for the video ads.
What’s important here is that we not only identified the problem quickly but were also able to resolve it immediately, so the costs only spiked for one day. Without the script that shows the cost distribution over individual days, it would have been much harder and slower to identify this error. The script saved us a lot of time and saved the client a significant amount of money.
You can download the script here: PMAX SCRIPT
