A campaign can show a healthy click-through rate, generate plenty of add-to-baskets, and still fail the only test that matters: whether it produces profitable new customers at a level you can scale. That is why the debate around broad targeting versus interests is not really about choosing a favourite Meta Ads setting. It is about deciding how to give the platform enough room to find buyers while retaining a testing framework your team can trust.

For established eCommerce and lead generation businesses, the answer is rarely “always broad” or “always interests”. It depends on your account history, creative strength, conversion signal quality, offer, budget and the economics behind every acquisition. The right structure is the one that creates reliable learning and supports profitable growth, not the one that looks smartest in an account screenshot.

What broad targeting actually asks Meta to do

Broad targeting generally means keeping audience restrictions light. You may set location, age where genuinely relevant, and gender only when the product requires it, then allow Meta’s delivery system to find likely converters across a wide eligible audience.

The premise is straightforward. Meta has access to far more behavioural and contextual signals than any advertiser can select manually in an interest field. With sufficient conversion data and compelling creative, it can identify patterns among people who are likely to buy or submit a lead, then serve ads accordingly.

This approach has become more effective as platforms have reduced the precision and transparency of interest targeting. Interest labels can be inconsistent, broad, outdated or based on signals that do not reflect current purchase intent. A person interested in fitness is not necessarily shopping for your protein powder this week. Conversely, a high-intent customer may never appear inside the interest audiences you assumed would work.

Broad also gives delivery more flexibility. Instead of forcing spend into a narrowly defined segment, Meta can move towards lower-cost conversion opportunities as auction conditions change. For brands with mature pixel data, clean server-side tracking and a steady flow of purchases, that flexibility can be a serious advantage.

Where interest targeting still earns its place

Interest targeting is not obsolete. It remains a useful way to create a meaningful starting point when an account has limited data, a niche offer, or a clear audience context that creative alone cannot communicate efficiently.

A specialist B2B lead generation campaign is a good example. If the offer is built for operations directors in a specific industry, using relevant professional, industry or software-related interests may help reduce wasted early spend. The same applies to eCommerce products that serve a distinct enthusiast category, such as technical outdoor equipment or hobby-led products with recognisable adjacent brands and publications.

Interest audiences can also provide structure for creative testing. If a brand has two genuinely different customer motivations, targeted ad sets may reveal which message earns stronger response from each group. One angle may resonate with price-conscious buyers, while another appeals to customers motivated by quality, convenience or status.

The mistake is treating interests as proof of intent. They are hypotheses, not customer lists. A strong result in one interest cluster may reflect the creative, placement mix, attribution window or temporary auction conditions as much as the audience itself.

Broad targeting versus interests: the real trade-off

The core trade-off is control versus flexibility. Interests provide a clearer theory about who should see your ads. Broad gives the platform greater freedom to act on its own signals. Neither is automatically more profitable.

Broad campaigns often need stronger foundations. Meta needs enough conversion events to learn, and your tracking must send reliable signals back to the platform. If purchase values are inaccurate, duplicate events are firing, or consent configuration is suppressing key data without a considered measurement plan, broad delivery can optimise towards the wrong outcome.

Interest campaigns can feel safer because they are easier to explain. A marketing team can point to an audience and say, “These are our people.” But that apparent precision may come at a cost. Tight audience restrictions can limit reach, increase frequency, make scaling harder and prevent delivery from finding unexpected but valuable customer segments.

For this reason, performance should be judged beyond top-line platform ROAS. Review new customer acquisition cost, contribution margin, conversion rate, average order value, lead quality and post-click behaviour. For lead generation, sales acceptance rate and booked-meeting rate often tell a far more useful story than cost per form fill.

When to start broad

Broad is often the sensible first choice when you have a mass-market or moderately broad product, stable conversion tracking, and enough budget to generate consistent purchase or lead volume. It is especially effective when your creative clearly qualifies the audience.

Creative is the targeting mechanism many advertisers overlook. A video that demonstrates a product’s specific use case, a clear price point, and a direct statement of who it is for will naturally filter out many irrelevant users. Someone with no need for the product is unlikely to engage, while the right buyer gets a faster reason to act.

For example, a premium skincare brand does not need to rely solely on beauty interests. Its ads can communicate the skin concern, product format, ingredient story, price position and proof points. Those elements help Meta recognise who responds, while helping potential customers decide whether the brand is relevant.

Broad is also worth prioritising when an interest audience has become too small to support delivery. If frequency rises quickly and acquisition costs climb, expanding the eligible audience can give the algorithm more room to work.

When interests are the better testing tool

Start with interests when there is a credible reason to believe broad delivery will struggle to find enough qualified users at the current stage of the account. This could be a new brand with little conversion history, a restricted budget, a specialist product, or a lead-generation offer aimed at a narrow professional profile.

Keep the structure disciplined. Rather than building dozens of tiny ad sets, group related interests into a small number of distinct audience themes. One group might represent direct category interests, another competitor or adjacent-brand signals, and a third relevant lifestyle or professional context. The goal is to test meaningful differences, not create an unmanageable account.

Avoid stacking every plausible interest into one audience. When many interests are combined, you lose the ability to understand what is driving performance and may create a group so broad that it offers little practical distinction from broad targeting anyway.

It is also wise to separate prospecting from remarketing. Retargeting results can mask weak acquisition performance, particularly when warm audiences are small and heavily exposed. A prospecting test should show whether you can create demand efficiently among people who do not already know the brand.

A practical testing framework for paid social

A fair test needs consistency. Launching broad with new creative while running interests with proven winners does not tell you which audience approach is better. It only tells you that different variables were used.

Begin with one broad prospecting campaign and one interest-based prospecting campaign. Use the same conversion event, comparable budgets, matching geographic coverage and the same core creative concepts. Let each test run long enough to gather meaningful conversion volume rather than making decisions after a day of volatile results.

Then assess performance at three levels. First, look at efficiency: cost per acquisition, ROAS and cost per qualified lead. Second, look at quality: average order value, repeat purchase behaviour where available, lead-to-sale rate and margin. Third, look at scalability: whether performance holds as spend increases, not merely at a small daily budget.

If broad wins on volume but loses slightly on first-purchase ROAS, do not dismiss it immediately. It may be bringing in customers with a higher lifetime value or giving your account a larger pool of future retargeting opportunities. Equally, if interests produce a lower cost per purchase but cannot absorb more budget without rapid deterioration, their value may be limited.

At Lightspeed Digital Media, we treat these results as inputs to a wider growth plan, not isolated campaign verdicts. Data should drive decisions, but the data must connect back to commercial outcomes and the constraints of the business.

The factors that matter more than audience labels

Audience selection has influence, but it is rarely the largest lever in a scaling account. Creative fatigue, weak landing-page conversion, unclear offers and poor measurement can all make an audience look like the problem.

Before repeatedly rebuilding targeting, check whether your ads give people a compelling reason to act. Review the product page or lead form experience on mobile. Confirm that pixel and conversion API events are deduplicated and correctly valued. Compare platform reporting with your analytics, CRM or backend sales data so you understand where attribution differs.

A weak offer will not become profitable because it is shown to people interested in the right topic. Equally, strong creative and a strong offer can often outperform narrow targeting by making relevance obvious at the point of impression.

Build the account around learning, not certainty

The most productive approach is usually a blended one. Keep a broad campaign active to give Meta room to find demand, while using selected interest tests to investigate specific customer hypotheses or support newer products and offers. Shift budget according to verified results, but leave enough space for exploration.

As the account gathers cleaner data, revisit assumptions. An interest group that worked six months ago may have saturated. A broad campaign that once struggled may improve after better creative, stronger conversion tracking or increased purchase volume. Paid social is not static, and neither should your audience strategy be.

The helpful next step is not to ask which setting is universally best. Ask what your current data can support, what you need to learn next, and whether every pound of spend is moving you towards sustainable, scalable growth.