For years, advertisers built tight audiences by stacking interests and demographics. Meta's machine learning now often outperforms that approach, because it uses far more signals than interest targeting can capture. With broad targeting, the creative itself attracts the right viewers, and conversion data teaches the system who actually converts.
Broad targeting works best for accounts with solid conversion tracking and enough volume to learn, and when the creative clearly speaks to the intended customer. It's less suited to niche B2B audiences or very small budgets, where some guidance — through Advantage+ audience suggestions — still helps.
Example
A national e-commerce brand removes all interest targeting and runs a broad U.S. audience aged 25–65, letting conversion data guide delivery — and its cost per purchase drops.
How Sloss Digital uses it
We use broad targeting when the data supports it and let creative do the targeting — testing it against suggested audiences rather than assuming either way.