This post summarises an A/B test I recently ran on Meta ads to compare the performance of broad versus narrow audiences. With a lot being said in PPC spaces about “targeting being dead” following Meta’s algorithm updates, I wanted to test whether a broad Advantage+ audience would outperform niche targeting. For context, the campaign was for an education conference in Europe aimed at people in specific medical professions.
In setting up the “narrow” ad set, I found that Meta’s library of behaviours and interests didn’t include options that match the event’s topic exactly. While I could find broad interests related to “medical education” and “medical research”, there were no interests specific to the subject matter (e.g. “multiple sclerosis”). So instead, I leaned on Meta’s demographic targeting, selecting people who work in the health industry, along with job titles that closely matched the intended audience.
The estimated audience sizes for the narrow and broad groups were very different. The niche audience’s estimated reach was between 88,700 and 104,400 people, while the broad audience ranged from 40.2 million to 47.3 million. So, which do you think performed better?
By the end of the test, the narrow audience delivered a cost per result that was approximately five times lower than the broad targeting. It was a clear win for further restricting your target audience.
What makes this interesting is that it goes against the current narrative. There’s no shortage of content claiming that, in 2026, targeting is outdated and that success with Meta ads now comes from focusing on creative variety and broad audiences. To add to that, the Meta platform itself actively discourages defining your targeting audiences too precisely. It gave the winning ad group a low optimisation score and flagged it with a warning that deviating from Advantage+ could increase costs.
So what gives? This experiment doesn’t prove that narrow audiences work better than broad targeting. It only shows that, in this specific scenario (this market, this budget, this timeline) narrow targeting worked best. And that’s really the point.
Digital advertising advice tends to become dogma very quickly. And once enough people repeat the same idea, it starts to feel like a universal truth. But marketing has never been one size fits all. Every situation is unique and the only data that truly matters is the data from your own campaigns. Industry advice can give you ideas. Platform recommendations can give you direction. But neither should replace a strategic approach to AB testing.
