Ad Targeting: How We Reach People Who Actually Buy

Good targeting puts your ad in front of the people most likely to buy, at the lowest possible cost. On today's ad platforms, especially Meta, that doesn't come from picking twenty interests and narrowing the audience as far as it goes. It comes from three things working together: correct basic settings (country, city, age, language), data about your real customers, and creative that clearly speaks to the person you want.

The part many business owners find surprising: the ad itself is now a big part of targeting. When a video opens with a parent worrying about their child's final school exams, the algorithm learns from who stops and who engages, then looks for more people like them. So we treat targeting and creative as one decision, not two.

Why over-narrowing raises your costs

Narrow the audience too much and you take away the platform's room to search. You end up competing for the same small group, the price of reaching them goes up, and they see the same ad until they're tired of it. Meanwhile, some of your real buyers never show up inside the interests you picked, because the platform simply doesn't know that about them.

That doesn't mean launching to everyone without thought. It means setting the limits that actually matter and letting creative and data handle the rest. A shop that only delivers in Amman has no reason to show ads in Irbid. That is a real limit. Stacking "cooking" and "restaurants" and "healthy eating" on top of each other usually isn't.

The audience layers we work with

We split audiences into three layers, each with its own message.

1. Cold: people who don't know you

Usually a broad audience within the right area and age range, sometimes a clear interest group if the product is very specialized. The message explains the problem and introduces the solution, because this person has no idea who you are yet.

2. Warm: people who engaged

People who watched most of your videos, engaged with your page, messaged you without buying, or visited your site. We build Custom Audiences from them, and the message moves closer to the decision: customer reviews, an answer to a common objection, or a time-limited offer.

3. Existing customers

Your customer list (phone numbers or emails collected with consent) is used two ways: excluded from cold campaigns so you don't pay to reach people who already bought, or used as the seed for a Lookalike audience. If the product is bought repeatedly, they may get their own ads too.

Exclusions: the half of targeting people forget

A lot of waste comes from people who should never have seen the ad. Before any launch we review exclusions:

• Recent buyers, when the product isn't repurchased quickly.
• Staff and owners, whose engagement distorts the numbers.
• Areas your service doesn't reach, even nearby ones.
• In location settings, people who live in the area rather than anyone passing through, when the product depends on where people live.

When a client sells in more than one country, such as Jordan, Kuwait and Saudi Arabia, each country gets its own campaign or ad set. Costs differ a lot between markets, and if they're combined the platform tends to spend most of the budget wherever reach is cheapest, even if the buyers are elsewhere. Splitting also lets us write in each market's dialect with clear local prices and delivery details.

Example: a private school in Amman opening registration

Say a private school in West Amman wants to fill next year's first-grade seats. The old approach: target "mothers", "parenting" and "education" across all of Amman with one ad showing the building.

How we would approach it:

1. Set the real geography: the neighborhoods most current students come from, based on the school's own records.
2. Write for parents specifically: a video opening with the question every parent asks, "What actually differs between schools in the first year?", answered by the first-grade teacher.
3. Launch broad within that area and parent age range, and let the content filter. A single 22-year-old scrolls past. A father of a five-year-old stops.
4. Build a warm audience from people who watched most of the video, and return to them with a classroom tour and a link to book a visit or start a WhatsApp chat.
5. Exclude current parents from cold ads and give them a separate message about sibling registration.

Targeting here comes from three sources: geography based on real data, creative that selects its audience, and warm audiences that move people toward a decision.

How we know targeting is working

We don't judge targeting by audience size or reach. We judge it by numbers tied to sales:

• Cost per conversation or per lead: falling or rising over time?
• Conversation quality: the client's sales team tells us whether people messaging are serious. That matters more than any dashboard figure.
• Frequency: if people see the ad many times with weak results, the audience is too small or the creative is tired.
• Who actually buys: ages and areas of buyers, not just engagers.

More on reading these numbers in how we read your account data, and on placing audiences into campaigns in how we structure Meta campaigns.

How we handle this at Wave

We start from the client's own data: who bought, from where, and what they asked before buying. We set only the limits that matter (area, age, language, exclusions) and write several content angles for different buyer types, as covered in content angles. After launch we build warm audiences as they grow and check lead quality with the sales team. If an ad brings plenty of conversations but no sales, the message is the first thing we review, before touching the settings.