08 Aug Why Interest Targeting Fails (And the Narrow Detail Fix)
When advertisers want to target a specific type of person — wealthy people, for instance — the instinct is almost always the same: search detailed targeting for luxury brand interests. Maserati, Prada, Ferrari. It feels precise. It usually isn’t.
Interest categories reflect the past, not the present
Here’s the problem with that instinct: liking a page doesn’t reflect current ability or intent, it reflects a moment in the past. Someone might genuinely like pages for Maserati, Ferrari, and Porsche — plenty of people who can’t afford any of them do. Liking the page and being able to buy the product are two completely different things.
The same distortion shows up in less obvious ways. Someone planning a wedding engages heavily with wedding-related content — photographers, planners, venues. Meta categorizes them as interested in weddings. After the wedding happens, most people don’t go back and “unlike” all of that content. So Meta continues showing them as interested in weddings long after the actual intent has expired. Targeting “people interested in weddings” months or years later is targeting a stale signal, not a live one.
This is why detailed targeting built purely on interest categories is often a lot less precise than it feels. It’s not that it doesn’t work — it’s that a meaningful share of anyone you reach through it no longer matches the intent the category implies.
The fix: stack two layers instead of relying on one
Instead of leaning on a single broad interest category, the more reliable approach is what’s often called narrow detail targeting — using Meta’s “Narrow Audience” option to add a second, more specific layer on top of the first.
Here’s how it works in practice. Start with a broader interest — say, people interested in property or investment. On its own, that’s a large, loosely-defined pool. Then narrow it using a second layer built from a more durable, behaviorally-grounded status: married, divorced, or parent. The result is the intersection of both — people who show property/investment interest and match a specific life status — which is a meaningfully more targeted group than either layer alone.
Not all detailed targeting is equally reliable
It’s worth knowing there are two different kinds of detailed targeting, and they’re not equally trustworthy.
Keyword-based interests — what comes up when you type and search a term, generally labeled “additional interest” — are the broader, less reliable category. This is the Maserati/Ferrari trap described above.
“Browse” categories — demographics, life events, relationship status, education — tend to be more behaviorally grounded and generally more accurate. Relationship status (married, divorced, engaged), parental status (parents, parents with toddlers, parents with pre-teens), and life events (recent anniversary, upcoming birthday) are pulled from more concrete signals than a page-like ever was. When building the second, narrowing layer of this technique, Browse-category options are usually the stronger choice.
Going one layer further: device exclusion
There’s a third lever worth knowing inside detailed targeting: device-based inclusion or exclusion. A business selling iPhone cases, for instance, only wants to reach iPhone owners — or at minimum, wants to exclude Android device owners entirely, since an Android user has no use for the product no matter how interested they otherwise seem. Stacking a device exclusion on top of an already-narrowed interest and status combination tightens the audience even further, cutting out a segment that was never going to convert regardless of anything else about them.
Don’t overdo it
This technique is powerful specifically because it’s targeted, not because narrower is always better. Stack too many layers and the audience shrinks to a size that struggles to deliver consistent results or drives CPM up more than the improved relevance is worth. Use narrow detail targeting deliberately for niche or high-value offers where precision genuinely matters — not as a default habit applied to every campaign.
A quick self-check before you build a targeting layer
- Am I relying on a single interest category that might reflect stale, past intent rather than current reality?
- Have I considered a second, narrowing layer built from a more durable status — married, parent, education level — rather than another loose interest?
- Am I using Browse-category options where possible, since they tend to be more reliably accurate than keyword-searched interests?
- Would a device exclusion meaningfully tighten my audience for a product that only applies to specific hardware?
- Have I stacked so many layers that my audience has become too small to deliver consistent results?
This breakdown is written by Jason Gan, a Meta Certified Professional and Badged Meta Business Partner who has personally audited over 1,000 advertiser accounts since 2010. You can see real account breakdowns on the Jason Gan YouTube channel.
Frequently Asked Questions
Why is basic interest-based targeting on Meta often unreliable? Because interest categories reflect past engagement, not present intent. Someone who liked a luxury car brand’s page may not be able to afford one. Someone who engaged with wedding content while planning their own wedding is still categorized as “interested in weddings” long after they’ve married and moved on.
What’s the “Narrow Detail Targeting” technique? Stacking two layers of detailed targeting instead of relying on one broad interest category. The first layer is a broader interest (like property or investment). The second layer, added through Meta’s “Narrow Audience” option, filters that group down using a more durable status like married, divorced, or parent — targeting the intersection rather than either group alone.
What’s the difference between keyword-based interests and “Browse” category targeting? Keyword-based interests (what shows up when you type and search, labeled “additional interest”) tend to be broader and less reliable. The “Browse” categories — demographics, life events, relationship status — tend to be more behaviorally grounded and generally more accurate, since they’re based on more concrete signals like relationship status or education level.
Can I exclude specific device types from my targeting? Yes. Device targeting sits inside the detailed targeting options, and it works for exclusion as well as inclusion. A business selling iPhone cases, for example, can exclude Android device owners specifically, tightening the audience toward people who actually own the product the ad is designed for.
Book a Meta Ads Audit — $80 / 30 Minutes →
No login required. If your detailed targeting feels like it’s missing the mark, we can rebuild it together with a layered approach that actually holds up.
If you want ongoing help refining targeting as you test new offers, that’s exactly what 1-on-1 coaching is for.