How to Create Audience Signal for Performance Max Campaigns

An Audience Signal is a hint you give to Google's AI about who is most likely to convert for your business. It is not hard targeting — Google will still show your ads beyond the signal if it believes other users are likely to convert. The signal simply accelerates the learning phase by pointing the algorithm in the right direction from the start.

No. Unlike standard audience targeting in Search or Display campaigns, an Audience Signal does not restrict who sees your ads. Google may serve your ads to users outside the signal if the algorithm predicts a conversion. Think of it as a starting point, not a fence — it guides the AI rather than constraining it.

You can add signals from three main categories: Your Data (customer lists, website visitors, app users), Interests & Detailed Demographics (in-market audiences, affinity segments), and Demographics (age, gender, parental status, household income). Using a combination from all three categories gives the algorithm the richest possible starting signal.

Yes. Each Asset Group within a Performance Max campaign can have its own Audience Signal. This is useful when you have distinct product or service categories — for example, one Asset Group targeting existing customers with a loyalty signal and another targeting new prospects with interest-based signals. Tailoring signals per asset group can improve relevance and performance.

No, a customer list is not required. You can build an Audience Signal using only Interests & Detailed Demographics and Demographics data if you don't have first-party data available yet. However, adding a customer list or website visitor audience under "Your Data" is strongly recommended if available, as first-party signals tend to be the most accurate.

It is a good practice to revisit your Audience Signal when you launch a new product, shift your target demographic, or notice the campaign is consistently reaching an audience that doesn't match your customer profile. Keep in mind that changing the signal resets some of the algorithm's learning, so avoid making frequent changes once the campaign is performing well.