We've tested broad vs. narrow targeting across dozens of accounts over the past three years. Broad wins almost every time — not because interest targeting is bad, but because Meta's AI is better at finding buyers than any manually built audience stack. The accounts clinging to hyper-segmented interest stacks are fighting the algorithm instead of working with it.
The Shift That Changed Everything
For years, Meta Ads success meant building precise audience stacks: layering interests, behaviours, demographics, and lookalikes into tightly defined segments. The skill was in the targeting. Broad audiences were for brands with unlimited budgets; everyone else needed to be precise.
That era is over. Meta's AI — powered by years of behavioural data across Facebook and Instagram — has become genuinely better at finding likely buyers than most advertisers are at manually defining them. The algorithm needs volume to learn, and narrow audiences starve it of the data it needs.
The shift happened gradually, then all at once. iOS 14 destroyed signal fidelity for interest-based targeting. Meta's Advantage+ systems absorbed more campaign decisions. And the results data became undeniable: broad audiences, given the right creative and enough budget, consistently outperform narrowly defined stacks.
Broad targeting on Meta means running an ad set with no interest, behaviour, or demographic targeting beyond basic location and age. You're handing the algorithm a blank slate and saying: find the people most likely to convert. It sounds scary. It works.
Why the Algorithm Beats Manual Targeting
Scale of signal
Meta processes billions of signals daily — what people watch, like, share, click, buy, and ignore. No manually built audience can approximate the pattern-matching capability of a system operating at this scale. Your interest-based audience is a proxy for buyer intent. Meta's AI is working with the actual behavioural signal.
Real-time adaptation
Interest audiences are static. You set them, they don't change until you change them. Meta's broad targeting adapts in real time — shifting delivery toward users exhibiting the conversion signals your campaign has learned to recognise. If purchase behaviour shifts on a Tuesday, your broad audience adapts. Your interest stack doesn't.
The creative becomes the targeting
This is the most important mindset shift. When you use broad targeting, the creative itself acts as the audience filter. Someone scrolling past your ad self-selects based on whether the creative resonates with them. The algorithm learns who stops, who clicks, who buys — and finds more of them. Your job shifts from audience architect to creative director.
How to Structure Broad Targeting Campaigns
Campaign architecture
The Darlington approach for broad targeting campaigns is simple: one campaign, one or two ad sets, multiple creative variations. The consolidation is intentional — it gives the algorithm maximum data to work with and avoids the audience overlap that fragments learning across multiple ad sets.
| Layer | Setup | Rationale |
|---|---|---|
| Campaign | 1 campaign per objective | Clean data, no internal competition |
| Ad set | 1–2 ad sets, broad targeting | Consolidate learning, avoid fragmentation |
| Ads | 3–5 creative variations per ad set | Let the algorithm find the winning creative |
| Budget | CBO (Campaign Budget Optimisation) | Meta allocates budget to best-performing ad set |
Minimum budget thresholds
Broad targeting needs data to work. The minimum daily budget for broad targeting to function effectively is roughly 3–5x your target CPA. If your target CPA is $50, you need at least $150/day — ideally $250. Below that threshold, the algorithm doesn't get enough conversion events to find its footing.
Creative is your only real lever
When you remove targeting as a variable, creative becomes everything. This means testing more creative, testing it faster, and building a feedback loop between what performs and what you make next. The accounts winning on broad targeting today are running 15–20 creative tests per month. The accounts losing are running 2–3.
Test across four dimensions: format (static, video, carousel), hook (first 3 seconds for video, first visual for static), angle (problem-led vs. outcome-led vs. social proof), and offer (discount vs. free trial vs. free guide). Each dimension is a separate test — don't change multiple variables at once.
When Interest Targeting Still Makes Sense
Broad isn't always right. There are specific situations where interest targeting is still the better choice:
- New accounts with no conversion history. The algorithm needs a starting signal. Interest targeting gives it initial direction while it builds conversion data.
- Very niche products with genuinely small addressable markets. If your buyer is genuinely a narrow demographic (e.g. professional beekeepers), broad targeting will waste budget on non-buyers before finding signal.
- Top-of-funnel content distribution. If your goal is reach and video views (not conversions), interest targeting can help you find an engaged initial audience for content amplification.
Outside these cases, default to broad. The data consistently supports it.
Is your Meta structure holding you back?
We'll audit your campaign architecture, targeting setup, and creative strategy and tell you exactly what to change.
Broad Targeting vs. Advantage+ Shopping
Meta's Advantage+ Shopping Campaigns (ASC) take broad targeting to its logical extreme — no manual targeting at all, with Meta handling audience, placement, and budget allocation automatically. For e-commerce brands with strong creative and conversion volume, ASC often outperforms manually structured broad targeting campaigns.
The tradeoff is control. With ASC, you can't structure ad sets, can't isolate creative tests cleanly, and have limited visibility into where your budget goes. For brands that need that granularity, manually structured broad campaigns are the better fit. For brands that want to maximise performance with minimal management overhead, ASC is worth testing seriously.
The Bottom Line
Broad targeting isn't lazy. It's a deliberate structural choice that lets Meta's AI do what it does better than any human: find the right buyer at the right moment. The shift this requires isn't technical — it's philosophical. Stop trying to control who sees your ads and start controlling what they see when they do.
Build better creative. Give the algorithm room. Trust the conversion signal. That's the meta game on Meta right now.