Performance Max: what you do and do not get to see

Performance Max, usually shortened to PMax, bundles nearly all Google channels into one campaign. You supply budget, goals, copy, images and a product feed, and Google decides where and when you appear.

It works well in many accounts. The problem is that it is hard to establish why it works, and that makes steering difficult.

What you do not get to see

• No full breakdown per channel. You do not get a clean split of what went to YouTube and what went to Search.
• Limited insight at placement level. You can see where ads appeared, but not always with cost and conversions attached.
• Search terms yes, but in less detail than in a regular search campaign.
• No separate bids per channel. You steer on the whole.

That is not a fault in your account, that is how the campaign type is built. Anyone promising they can pry this open is promising something Google does not return.

What you can steer

Asset groups. Think of them as ad groups. Make several, and keep products or services that have nothing to do with each other apart, or you will not be able to separate anything later.

Audience signals. These are hints, not targeting. Google may go beyond them. They mainly help in the first weeks.

Account-level exclusions. Negative keywords for PMax are available in limited form and differ per account and per period. Brand and placement exclusions do exist. Check what your account offers rather than assuming what was possible last year.

The feed, if you run a webshop. For Shopping traffic, the quality of your product titles and images is one of the few levers that genuinely has an effect.

The trap with your own brand name

PMax happily picks up searches on your own brand name. That traffic almost always converts well, because those people were already looking for you. Your campaign therefore looks excellent while part of that revenue would have arrived without an ad.

To know what PMax really delivers, you want to keep brand traffic separate. That can be done with brand exclusions if your account has that option, or with a separate brand campaign that catches the traffic. Without that separation, you do not know what you are measuring.

How to judge it

Do not look at PMax performance on its own, but at your account total before and after it started. If PMax rises while your search campaigns fall by the same amount, traffic has moved rather than been added.

Give it time as well. The first weeks are expensive, and that is the learning period, not the end result.

What AdWarp does here

We show what comes out of PMax through the API: search terms with cost, conversions and the change over time, with the exact period attached. Where Google returns no cost per term, there is no amount rather than an estimated one. That is why an AdWarp report sometimes shows a row without a euro figure, with a note that cost per term is not available.

That is less tidy than a fully populated table, and it is the only way you can trust the rest of the table. See also the honesty rules.