The Rule Behind Your Audience Settings
Google's personalized advertising policy draws an unusually concrete line for financial advertisers. In the United States and Canada, gender, age, parental status, marital status, and postal code may not be used to select or target personalized ads for credit-related products and services, banking products and services, or certain financial planning and management services. The policy names credit cards and loans, bank and checking accounts, and debt management products as examples of what counts. Note what the restriction attaches to: personalization, not advertising itself. A card offer can still run; the audience-building step is what changes.
Why a Credit Card Offer Is Named, Not Implied
Because credit cards appear by name in the example list, there is little room to argue that a rewards card, a secured card, or a balance-transfer offer sits outside the category. The same sentence covers bank accounts and debt management products, so a campaign straddling a card and a deposit account should be treated as covered. The five-signal limitation also belongs to a wider US and Canada block covering housing and employment advertising, which reveals the pattern: life circumstances are not treated as fair stand-ins for financial need.
The Five Signals, Translated Into Media Plans
Each one removes a familiar shortcut:
- Gender: no male-only or female-only card targeting, and no gendered lookalike builds.
- Age: no 25–34 brackets, no age bands assembled from birth-date fields.
- Parental status: no "parents of young children" segments, however they were sourced.
- Marital status: no married-versus-single splits.
- Postal code: no neighborhood-level ZIP targeting, even for a single-city campaign.
Together these signals describe life stage and household structure — precisely the inputs a planner reaches for when guessing who needs a new card. The policy removes that instinct from the setup, not from the creative work.
Two Plan Sketches, One Blocked and One Allowed
A blocked sketch: a travel rewards card aimed at "single professionals, ages 28 to 35, in three trendy ZIP codes, no kids." Four of the five prohibited signals carry that plan, and strong creative cannot rescue it.
An allowed sketch: the same card aimed at people who recently searched international flight deals or read frequent-flyer content, with an interest-based ad notice on every placement. The audience is built from expressed interest rather than assumed life stage.
The contrast is the whole lesson: the second plan asks what someone did, the first asks who someone is.
What Is Still Available, and the Strings Attached
Interest and behavior signals become the natural substitute, and the policy does not name custom intent, affinity, in-market, or remarketing lists among the prohibited inputs. They still carry conditions. You must hold all rights needed to use audience data such as cookie lists; the ad must state that it is interest-based, which Google illustrates with the AdChoices icon; and all parties must follow self-regulatory codes such as the Digital Advertising Alliance principles or the IAB Europe framework.
A separate exclusion cuts across every category you sell. Personalization may not rest on the past or current activity of users known to be under 13, on activity in adult, gambling, or government sites and apps, or on inferred and actual sensitive information — the policy explicitly names negative financial information, such as a low credit rating or a high debt burden. A "credit-challenged" or "high-debt" segment is therefore not a workaround for the age ban; it is its own violation.
Claims and Landing Pages Carry a Second Set of Traps
Targeting is only half the exposure; the system also polices what the card copy and the click path promise. Publishers may not misrepresent or conceal information about themselves, the purpose of their content, or an affiliation or endorsement they do not have. Deceptive promotion is barred, including get-rich-quick framing. Traffic sources must describe the destination accurately, must not advertise an offer that is missing or hard to find, and every landing page must match what an organic visitor would see. One example in Google's policy documentation concerns a promise outside the advertiser's control — "Open a PNC high-yield savings account with no credit check!" — filed under egregious rather than ordinary violations. Card content that borrows a bank's name or implies an approval it cannot guarantee sits in the same territory.
A Pre-Launch Sequence You Can Repeat
- Confirm the market. US and Canada trigger this restriction; other countries follow different local rules and need their own review.
- Inventory every audience input and delete the five signals before the first draft goes live.
- Verify rights to each audience list and confirm the interest-based notice renders on all placements, including app inventory.
- Screen for sensitive inferences, treating financial-distress signals as a hard stop.
- Read creative and landing page side by side: claims, disclosures, and parity with organic traffic.
- Log the decision and route product classification to compliance, since "certain financial planning and management services" is not defined in the policy text.
Where This Reading Stops
This is a reading of a published policy summary, and Google can revise both the policy and its enforcement. The text also leaves "certain financial planning and management services" undefined, so hybrid products deserve case-by-case classification. Nothing here substitutes for consumer credit law, state rules, or the disclosures a card issuer owes applicants; those require qualified legal and compliance advice.
Holding the line is easier when it is written into the brief. Before the next card campaign is assembled, list the five prohibited signals by name, require a justification for each audience input, and get compliance sign-off on classification. The plan that survives will be the one that describes behavior rather than people.