Demographic Targeting Optimisation | Age and Income

Google Ads reports performance broken down by age, gender, parental status and estimated income. That data can be used both to narrow targeting and to differentiate bids.

Report first, decision second

Before touching demographic settings you have to read the current distribution: which age group converts better, which spends without converting.

The most common mistake here is narrowing on assumption without looking at data. In accounts that closed an age group believing “our customers are not young”, I have later found that group converting well.

The limits of the data

Google does not know these attributes for certain, it estimates them. Estimates are more accurate for signed-in users and behaviour-based for others. A meaningful share of users also falls into “unknown”.

That unknown group often makes up a sizeable part of traffic. When excluding demographics you also have to decide what happens to it — closing it can cost a lot of volume.

When exclusion is right

If the product genuinely targets a specific group, exclusion makes sense: age-restricted products, services tied to a life stage. There it protects budget and is also the correct behaviour policy-wise.

But most products have no hard demographic boundary, and narrowing cuts volume needlessly. Narrowing has a second cost too: as targeting tightens, smart bidding has less data and learns more slowly.

Different rules in restricted sectors

In areas such as health, finance, employment and housing, Google restricts demographic targeting. In those sectors you cannot target by age, gender and similar attributes.

Campaigns built without knowing this can be disapproved or quietly limited. On the policy violation side this is a frequent cause.

How do we start? First I map your real performance across demographic breakdowns and compare it against your assumptions. I send the narrowing or bid adjustment plan in writing. You can reach me from the contact page.

Frequently Asked Questions

Is income targeting reliable?

It is estimated and supported by regional data; it can suggest direction but should not be taken as fact. It can help for luxury products, but should not be the sole basis for targeting.

Should I exclude unknown demographics?

Usually not. That group is a large share of traffic in most accounts and contains well-converting users. Check its performance before excluding.

I use smart bidding. Should I touch demographics?

Usually not; the algorithm already uses these signals. Manual intervention should be limited to cases where the product genuinely serves one specific group.