What the policy prohibits for rental housing
The covered category is housing, explicitly including for-sale or for-rent housing listings, plus real estate services. In the US and Canada, the policy says personalized ads may not be selected or targeted based on gender, age, parental status, marital status, or ZIP code. Those are five separate inputs. A ZIP code in ad text is creative content; a ZIP code in a targeting panel is an audience signal, and this policy addresses the targeting choice. A rental listing can be allowed while one of these audience settings is not. The retrieved policy states the restriction without explaining its rationale. Source: Google publisher policies and AdSense program policies.
Housing, employment, and credit: one list, different exceptions
The same restricted list covers employment services—recruitment ads, job boards, job listing sites—and credit-related products, banking products and services, and certain financial planning and management services, with examples such as credit cards, loans, bank and checking accounts, and debt management products. A recruitment ad for leasing help or a credit-card promotion is therefore not a housing ad, yet it shares the same prohibited signals. The table shows where the retrieved text documents an exception.
| Restricted category (per retrieved policy) | Prohibited personalized-ad targeting signals in the US and Canada | Documented exception in the retrieved policy text |
|---|
| Housing, including for-sale or for-rent housing listings, and real estate services | Gender, age, parental status, marital status, ZIP code | None identified in the retrieved text |
| Employment services, such as recruitment ads, job boards, and job listing sites | Gender, age, parental status, marital status, ZIP code | A pre-identified subset of US government advertisers may target limited audiences when targeting is based on the actual occupational qualification of the government role (BFOQ under US law) |
The exception is narrow: a pre-identified subset of US government advertisers may target limited audiences for employment ads when targeting reflects the actual occupational qualification of the government role (BFOQ under US law). No housing equivalent appears in the retrieved text, and no permitted housing alternatives are listed.
Disclosures that travel with personalized housing ads
Targeting choices are only half the check. The retrieved policy requires anyone using Google platform products or Google ad code for personalized ads to hold all necessary rights to audience data such as cookie lists; add an in-ad notice that the ad is interest-based, for example the Ad Options icon; and comply with applicable industry codes, such as the Digital Advertising Alliance’s Self-Regulatory Principles for Online Behavioral Advertising or the IAB Europe EU Framework. The codes are named, not reproduced, so read them at their source. Publishers must also maintain and follow a privacy policy clearly disclosing Google-related data collection, sharing, and use, including cookies, web beacons, IP addresses, or other identifiers across sites, apps, connected TV, consoles, or email platforms.
Sensitive data and precise location: the listing-page traps
Listing pages create a second data surface. Personalized advertising may not use inferred or actual sensitive information; the retrieved text explicitly names negative financial-status information, such as signals that a user has a low credit rating or high debt burden. Keep screening outputs and financial inferences out of ad targeting rather than turning an application field into an audience input. Precise location has its own trigger: if a publisher collects, processes, or discloses information that identifies or can infer a user’s precise location—for example GPS, Wi-Fi, or cell-tower data—the policy requires interstitial or in-context notice, explicit consent before collection, encrypted transmission to Google, and disclosure in applicable privacy policies.
Ad copy risk: promises the listing cannot keep
Adjacent Google guidance for AdSense for Search product-integrated features treats concrete promises that cannot be fulfilled as egregious violations, citing free or cash offers, promising loans, and promises outside the publisher’s control such as “no credit check.” That is a different product context and a risk illustration, not a universal rental-ad rule. The transferable point is to avoid absolute promises you cannot control. Separately, Google’s publisher content policies prohibit misrepresentation about the publisher, creator, purpose, or content, including falsely implying another party’s association or endorsement.
What to verify before you launch
Build the check in this order. First, audit every audience setting against the five named signals and remove selections based on gender, age, parental status, marital status, or ZIP code for housing or real estate services. Second, add the interest-based notice and confirm your rights to audience data. Third, update the privacy policy to disclose Google-related collection, sharing, and use, including cookies, web beacons, IP addresses, and other identifiers. Fourth, if you handle precise location data, document notice, consent, encryption, and policy disclosure. Fifth, confirm permitted targeting alternatives in Google Ads documentation, because the retrieved policy names only prohibitions. Sixth, recheck the English policy at the cited URL; the retrieved version was localized, and Google says it may revise policies at any time and that publishers must keep current. Record the review date. Finally, have a qualified professional review fair-housing and state landlord-tenant obligations, which the retrieved sources do not cover.
研究の知恵