What the policy actually prohibits
Google's publisher policies separate content rules from delivery rules, and the targeting restriction sits in the second group. When a site uses Google platform products or Google ad code for personalized advertising, the operator must not select or target those ads, or collect and use audience data such as cookie lists, on the basis of a defined set of signals. Two entries matter to rental businesses: housing, described as information about homes for sale or for rent, and real estate services. In the US and Canada those categories may not be personalized using gender, age, parental status, marital status or postal code. Credit, banking and some financial planning services fall under the same clause, which matters because rental screening brushes against credit information.
Why the rule reaches rental listings, not just home sales
The policy pairs housing with real estate services and never limits the housing entry to purchases. A page promoting available apartments, a leasing availability feed or a neighbourhood landing page is housing information in the ordinary sense, so the same five signals are unavailable there. The restriction attaches to the personalized ad request rather than your internal product logic: saved searches, waitlists and applicant records are yours to run, but once audience data from those systems feeds a personalized ad request through Google products or code, the rule applies. Segments built from renter profiles must be stripped of gender, age, parental status, marital status and postal code before the request leaves.
What else the same policy takes off the table
The prohibited-signal list runs longer than the housing clause, and part of it is uncomfortably close to rental data. Personalized ads may not be selected or targeted on health or medical records, information about negative financial circumstances such as low credit rating or heavy debt, ethnicity or race, religion, political affiliation, union membership, sexual behaviour or orientation, or past activity on adult, gambling or government sites. A platform that knows an applicant defaulted, or infers family composition from a co-applicant field, should not turn that knowledge into an ad audience. The policy publishes prohibitions rather than a complete list of permitted inputs. Context drawn from page content is not among the banned categories, but treat any new segment definition as a review item rather than an assumption. Fair-housing duties sit in a separate legal layer this policy does not address.
The disclosures that travel with the targeting rule
Removing prohibited signals is only half of the personalized-advertising obligation. The policy also requires the operator to hold every right needed to use audience data, including cookie lists; that ads carry a notice stating they are interest-based, such as the Ad Options icon; and that all parties follow applicable industry codes such as the Digital Advertising Alliance's Self-Regulatory Principles for Online Behavioral Advertising or, in Europe, the IAB Europe framework. Any site running Google products must also maintain a privacy policy clearly disclosing the collection, sharing and use those products cause, with detail about cookies, web beacons, IP addresses and other identifiers. Rental sites that pass enquiry data to leasing partners need that wording to cover the sharing, not just the collection.
Accuracy duties that quietly catch rental marketers
Google's misrepresentation policy requires that information supplied to its ad systems be very accurate and complete, with no misleading omissions and no deceptive framing. Named examples include seriously incomplete personal or payment information, inaccurate site or app details such as ads.txt or app-ads.txt entries, and deceptive ad request data such as incomplete or inaccurate URLs or app IDs or incorrect geographic information. Geographic accuracy is the field rental teams most often break, because city-level inventory gets mapped to metro targeting by hand. A listing page tagged with the wrong city, or an app ID that resolves to a different property than the one displayed, is a policy problem rather than a rounding error.
Placement mistakes common on rental listing pages
Three screen-level rules hit directory-style rental sites hardest. Ads may not appear on screens carrying no publisher content or only low-value content, that are still under construction, or that exist for reminders and navigation. Ads and other paid promotional content may not occupy more than the publisher content on the screen. And ads may not overlap or sit adjacent to navigation items in ways that invite accidental interaction, nor be styled so they read as menus, navigation elements or download links. Pagination pages, "coming soon" city shells and thin filter results are where these conditions cluster, and content must never fully or partially obscure Google-served ads.
Enforcement and market differences
Google's stated remedies are contractual rather than advisory: it may suspend delivery of its services, require an operator to stop or change how it uses them, or exercise rights under the applicable service terms. Scope shifts by market as well. The housing clause is written for the US and Canada, so a purely American rental site faces that restriction while an operator serving EU traffic also picks up the European user consent policy. Publishers in Crimea, Cuba, the so-called DNR and LNR, Iran, North Korea and Syria cannot use Google publisher products at all.
A review sequence before you ship a renter audience
Start by classifying each page: housing or real estate services, or something else on the site? Then list every personalization input in the build and delete gender, age, parental status, marital status and postal code from anything aimed at US or Canadian housing users. Screen remaining segments for sensitive-category inference, especially credit distress and inferred family status. Confirm you hold the rights to the audience data, that the ad carries an interest-based notice, and that the privacy policy names the practices and sharing partners. Audit ads.txt or app-ads.txt entries plus the geographic and URL fields in the ad request. Finally walk the layout: content-to-ad ratio, navigation adjacency, thin and unfinished pages. Policy text changes, so schedule a re-read, and route fair-housing questions to qualified counsel rather than to platform documentation.
研究の知恵