You split ad money between the two platforms as if they were interchangeable, and you put the neighborhood work on Facebook because that is where the audience is. Housing is a restricted category there, and the smallest circle it will let you draw is bigger than the neighborhood you are farming.
Buried inside a guide about YouTube ads is the sentence that decides where hyperlocal money should go: "Dial in Google's geo‑radius to as little as one mile around a hot subdivision—something Meta still won't let housing advertisers do (Facebook forces a 15‑mile minimum)."
One mile against fifteen. On Google you can draw the circle around a single subdivision. On Meta, by this account, a housing advertiser cannot go under fifteen miles, so a campaign aimed at one neighborhood is delivered across a circle you did not choose and cannot shrink.
He pairs the tight radius with audience layers rather than a wider circle: in-market audiences, life-event segments such as likely to move, and customer-match retargeting lists built from your newsletter.
One thing the article does not do is date the claim. This is the author's statement of platform policy rather than a citation to Meta's own rules, and ad-platform minimums move.
“something Meta still won’t let housing advertisers do (Facebook forces a 15‑mile minimum)”
What this costs you is most of the money you spend farming one neighborhood on Facebook. You pick the subdivision, the platform quietly widens it, and the report comes back with impressions that look fine, because impressions always look fine. You never see the households you paid for who live nine miles away.
What changes is that you split the budget by job instead of by preference. Broad awareness can stay where it is. Anything aimed at one subdivision, one condo building, or one school zone moves to Google, where the radius goes down to a mile. Confirm the current minimum in Meta's special ad category rules before you rebuild around it, because these rules do change.