Broad Targeting Versus Lookalikes: Which Wins?
A supplement brand can spend weeks rebuilding audiences while the actual constraint is creative, landing-page conversion, or an offer that does not earn a first purchase. That is why broad targeting versus lookalikes is the wrong debate when it is framed as a universal winner. The right question is which setup gives Meta enough room to find qualified customers while keeping acquisition economics under control.
For DTC supplements, the answer changes with account maturity, creative volume, customer data quality, and the claims your ads can safely make. Broad can scale remarkably well. Lookalikes can still be highly useful. Neither strategy fixes weak fundamentals.
Why this decision matters more for supplement brands
Supplement buyers rarely purchase on a single signal. They may care about ingredients, flavor, price per serving, third-party testing, subscription flexibility, or whether a product fits a training goal. Meta has to infer those signals from behavior, and it does so more effectively when the account sends clean conversion data and supplies a wide range of compliant creative.
At the same time, supplement advertising comes with narrower messaging boundaries than many ecommerce categories. You cannot rely on aggressive disease claims or before-and-after promises to force cheap conversions. That puts more pressure on the media system, the offer, and the creative angle to find buyers who will convert without creating account risk.
A narrow audience can appear efficient in a short attribution window while quietly limiting delivery and increasing frequency. A broad audience can look expensive during its first few days, then stabilize once it receives enough conversion volume. Founders who judge either approach too early often turn normal learning volatility into a scaling problem.
What broad targeting actually does
Broad targeting means giving Meta minimal audience constraints, typically using location, age, and any necessary exclusions while allowing the platform to optimize toward a purchase event. The audience is intentionally large. Rather than telling Meta exactly who should buy, you give it conversion signals, ad creative, and a clear optimization objective, then let delivery find likely customers.
This approach works best when the account has enough purchase data and a meaningful creative testing engine. A hydration brand, for example, might run separate ads around endurance, everyday energy, travel convenience, and clean ingredients. Broad delivery can identify which angle resonates with different pockets of buyers without artificially separating them into dozens of audiences.
The upside is scale. Broad audiences generally provide more inventory, lower overlap between ad sets, and fewer manual decisions that restrict the algorithm. It also reduces the temptation to keep rebuilding targeting instead of solving the real performance issue.
The downside is that broad is unforgiving when the rest of the system is weak. If your pixel or Conversions API is poorly configured, your purchase event is inconsistent, your landing page converts at 1%, or your ads all communicate the same generic benefit, Meta has little to optimize around. Broad targeting then becomes an expensive way to discover that the business is not ready for cold scale.
When broad is the strongest choice
Broad is usually a strong starting point for established accounts with consistent purchase volume, a stable Shopify funnel, and at least several genuinely distinct creative concepts ready to test. It is especially effective when a brand has a product with wide appeal, such as protein, electrolytes, creatine, greens, or daily wellness formulas.
It can also outperform audiences built from stale assumptions. A buyer who purchased a sleep product two years ago may not resemble your best current customer. Broad delivery can react to the current market, current creative, and current auction instead of being anchored to an old customer snapshot.
That does not mean every campaign should be completely unrestricted. Geography, age thresholds, existing customer exclusions, and compliance considerations still matter. Broad is not careless targeting. It is strategic restraint.
Where lookalikes still earn their place
Lookalike audiences use a source list, such as purchasers, high-value customers, subscribers, or repeat buyers, to find people who share similar characteristics. Their value depends almost entirely on the quality of that source. A lookalike built from 5,000 customers who placed a single discounted order is very different from one built from high-LTV subscribers who repeatedly buy at full price.
For supplement brands, the most valuable source lists often reflect customer quality, not just customer count. A first-purchase buyer may be a deal seeker. A customer on their third subscription renewal has demonstrated product-market fit. If you can segment data by contribution margin, subscription retention, or 60- and 90-day repeat behavior, your lookalike testing becomes more commercially meaningful.
Lookalikes remain useful in several scenarios. They can help a newer account create a more informed cold prospecting test before Meta has enough recent purchase data. They can support a specialized product where the buyer profile is meaningfully narrower, such as a niche endurance formulation. And they can provide a controlled comparison when broad delivery becomes unstable after a major creative or offer change.
The limitation is that lookalikes do not create demand. They simply model the customers you already have. If your source list is small, low quality, or built from an old acquisition strategy, you may be asking Meta to find more of the wrong people.
The common lookalike mistake
Many accounts fragment budgets across 1%, 2%, 3%, 5%, value-based, purchaser, subscriber, and email-list lookalikes, all with similar creative. The result is auction overlap, underfunded ad sets, and too little conversion volume for any audience to learn efficiently.
A better approach is to select one or two source types with a clear business rationale. Test a purchaser or high-value customer lookalike against broad under comparable conditions. Give each enough budget and enough time to generate a credible number of purchases. Then judge the result on blended economics and downstream quality, not just a single day’s ROAS.
Broad targeting versus lookalikes: the practical test
The cleanest way to settle broad targeting versus lookalikes is not by following platform folklore. Run a structured test that controls the variables you can control.
Use the same campaign objective, attribution setting, geography, placement strategy, offer, and creative pool. Keep existing customers excluded from prospecting when the goal is net-new acquisition. Allocate sufficient budget to each ad set or campaign so neither is starved of delivery. For a brand spending $10,000 a month, that may mean a focused test rather than five simultaneous audience experiments. For a brand spending $100,000 or more, you can create a more rigorous split while preserving scale.
Evaluate performance after enough spend to move beyond random variation. The exact threshold depends on your average order value and conversion rate, but a handful of purchases is not evidence. Look at cost per new customer, new-customer ROAS, conversion rate, average order value, and early subscription attachment where applicable. If possible, connect the analysis to 30-, 60-, or 90-day repeat behavior.
This matters because the lowest front-end CPA is not always the most profitable customer. A broad audience may produce buyers who take a little more to acquire but subscribe at a higher rate. A lookalike may win initial ROAS while attracting customers who churn after one shipment. The media buyer’s job is to expose that difference before scale magnifies it.
Creative is the bridge between targeting and scale
Meta increasingly uses creative as a targeting signal. The opening hook, product format, creator profile, use case, visual environment, and proof point all influence who sees an ad. That is especially relevant for supplements because one product can serve several legitimate customer motivations.
A magnesium product could be positioned around a nighttime routine, muscle recovery, ingredient transparency, or a stress-conscious lifestyle, provided every claim is policy-safe and substantiated. A single broad campaign with these distinct angles may reach more relevant buyer groups than a narrowly targeted campaign with repetitive ads.
This is why creative testing should precede major audience changes. If the same three ads have fatigued, broad targeting is not the issue. If a new creative concept lowers CPA across broad and lookalike tests, the insight is not that one audience suddenly improved. It is that the market responded to a better reason to buy.
At HelloPerformly, we treat audience strategy as one part of an acquisition system that includes compliant messaging, creative throughput, offer architecture, and conversion rate optimization. A strong media setup cannot compensate indefinitely for a weak product page, unclear bundle, or checkout that gives buyers too many reasons to hesitate.
Make the decision based on account stage
Early-stage brands should avoid overcomplicating the account. If purchase data is limited, start with a small number of broad and high-quality lookalike tests, supported by creative built for clear use cases. Focus on collecting clean conversion data and learning which message produces qualified first-time buyers.
Growing brands with reliable conversion volume should usually make broad a central part of the prospecting mix. Keep a small number of lookalike tests active when there is a strong source list or a distinct product segment, but do not let audience segmentation dilute budgets and slow learning.
Mature brands should think beyond the front-end auction. Broad can be the scale engine, while value-based lookalikes, retention cohorts, and product-specific audiences become diagnostic tools. The question shifts from “Which audience gets the cheapest sale?” to “Which acquisition path produces the best contribution margin and customer lifetime value at higher spend?”
The useful closing thought is simple: do not reward an audience because it won last week. Reward the system that keeps finding qualified customers at a CPA your business can afford, while giving your team enough room to scale the next winning creative.