For B2C brands on Meta, broad audience targeting consistently outperforms interest-based targeting. Meta’s algorithm has access to billions of behavioural signals, including purchase history and cross-platform activity, that make manually defined interest stacks redundant. Giving the algorithm room to operate produces stronger conversion efficiency than restricting delivery to narrow audience segments.
Why Narrow Interest Targeting Hurts B2C Campaign Performance on Meta
Interest targeting made sense when Meta’s data infrastructure was limited. That is no longer the case. When B2C advertisers stack interest categories, they create an artificially small audience pool that prevents Meta’s delivery system from finding buyers outside those predefined boundaries. The algorithm cannot optimise properly when it is working with a restricted sample.
The practical result is higher CPAs. Meta’s Performance 5 framework recommends broad audiences over interest-targeted approaches for most campaign objectives, a position supported by the platform’s algorithm design which performs best with larger audience pools to optimise against. When you constrain delivery, you are essentially telling Meta to ignore the behavioural signals it has already collected on people who would buy your product but do not fit a particular interest label.
Interest categories also age poorly. Someone who purchased running shoes six months ago may no longer engage with fitness content, but they remain a high-probability buyer for athletic apparel. Interest targeting misses them. Broad targeting, anchored to conversion signals, does not.
The compounding effect matters too. Narrow audiences exhaust faster, forcing budget into frequency before Meta has enough conversion data to properly calibrate delivery. This creates a cycle of diminishing returns that broad audiences avoid by maintaining a larger, refreshing pool of potential buyers. B2B advertisers face a similar dynamic, where narrow Meta audiences can drive a significant CPM premium by restricting the algorithm in the same way.
How Meta’s Machine Learning Algorithm Actually Finds Your Buyers
Meta’s Advantage+ and broad audience delivery systems draw on a data set that no advertiser can replicate manually. This includes on-platform purchase behaviour, off-platform browsing via the Meta Pixel and Conversions API, app activity, video engagement patterns, and cross-device behaviour. The algorithm uses this to build lookalike profiles in real time, without requiring an advertiser to define the audience upfront.
When you set a broad audience (typically defined as 18-65+ with minimal layering), you are giving Meta’s system the latitude to test and learn across a wide pool. It identifies which users share characteristics with your existing converters and shifts spend toward them dynamically. This is fundamentally more sophisticated than a static interest stack.
Consider a DTC skincare brand running a broad audience campaign with strong Conversions API integration. Meta’s system identifies purchase signals from users who have never interacted with beauty content but have demonstrated high purchase frequency in adjacent categories. A narrow interest audience built around skincare and beauty would have excluded them entirely.
The key input that makes this work is signal quality. Broad audiences perform best when paired with accurate, real-time conversion data flowing back to Meta through server-side tracking. The algorithm is only as good as the feedback it receives. How you measure that feedback also matters — Meta’s default attribution window can distort campaign performance data in ways that lead to misguided optimisation decisions.
How to Use Creative as Your Primary Targeting Strategy on Meta
When you remove interest targeting as the audience qualification mechanism, creative takes on that role instead. This is a meaningful shift in how B2C teams should approach campaign structure. The ad itself becomes the filter. Copy and visuals that speak precisely to a specific customer problem, lifestyle, or motivation will naturally attract the right buyers and repel unqualified clicks.
A useful framework: treat each creative as a message to one type of person, not a message to one type of audience segment. A B2C furniture brand might run three broad audience campaigns simultaneously, each with creative speaking to a distinct buyer motivation (first apartment, home renovation, aesthetic upgrade). Meta’s algorithm distributes each ad to users whose behaviour aligns with that context, without the advertiser manually defining who those users are.
This approach also generates cleaner performance data. When creative is doing the targeting work, you can read performance differences as genuine signals about customer motivation rather than artefacts of audience overlap or segment size. Weak creative performance tells you something real about messaging fit, not just audience selection.
Practical execution means testing creative variables systematically: headline angle, visual format, offer framing, and social proof type. Keep audience settings broad and consistent across tests so creative is the isolated variable. This produces learnings that compound over time and give Meta’s algorithm consistent, high-quality signals to optimise against.
Key Takeaways
- Stacking narrow interest categories restricts Meta’s algorithm from accessing its full purchase intent and behavioural signal data, driving up CPAs for B2C brands.
- Meta’s machine learning already holds purchase history, browsing behaviour, and cross-platform conversion data, making broad audiences a more effective delivery input than manually defined interest stacks.
- For B2C advertisers, creative that speaks directly to the ideal customer replaces the need for interest targeting, with a clean conversion signal doing the heavy lifting on audience qualification.
Frequently Asked Questions
What audience size should B2C brands use for broad targeting on Meta?
For most B2C campaigns, a broad audience with minimal demographic restrictions (typically 18-65+ in your target geography, with no interest or behaviour layering) gives Meta’s algorithm enough volume to optimise effectively. Audiences below 2-3 million can limit delivery and slow the learning phase, particularly for campaigns with higher CPAs.
Does broad audience targeting work for B2C brands with niche products?
Yes. Meta’s algorithm can identify buyers for niche products more accurately than manually defined interest stacks, provided the Pixel and Conversions API are tracking purchase events cleanly. The algorithm finds patterns in converter behaviour that often extend well beyond obvious interest categories, reaching buyers a narrow audience would exclude.
How important is Conversions API setup when running broad audiences on Meta?
Conversions API integration is critical. Broad audience delivery relies on accurate conversion signals to calibrate and improve over time. Without server-side tracking, signal loss from browser privacy changes and iOS restrictions degrades the algorithm’s ability to find buyers, which reduces the performance advantage that broad targeting provides.
Should B2C brands stop using retargeting audiences entirely in favour of broad targeting?
Not necessarily. Retargeting audiences built from high-intent site visitors or cart abandoners still serve a function, particularly for products with longer consideration cycles. The shift away from interest targeting applies primarily to prospecting campaigns, where broad audiences with strong creative outperform interest-stacked cold audiences for most B2C categories.
If your Meta campaigns are still structured around interest stacks, there is likely room to improve delivery efficiency and reduce CPAs. Explore our free paid media audit or review our performance marketing plans to see how we structure Meta campaigns for B2C growth.