How to Use Targeted Advertising
Targeted advertising questions test whether you can build audiences that match intent and relationship, then avoid the overlaps and exclusions that quietly waste budget.
Know the audience families
- Core: demographics, location, language, device, job title, company size.
- Interest and behaviour: what people engage with or do.
- Custom / matched: your CRM lists, customer emails, site visitors, video viewers, lead-form fillers.
- Lookalike or similar: modelled expansion from a high-value seed list.
- Retargeting: people who already interacted but have not converted.
- Exclusions: customers, competitors, employees, existing opportunities.
The tension is always reach versus precision: too narrow and delivery stalls; too broad and cost per qualified lead climbs. First-party seeds matter most, so consent, list hygiene and match rate drive performance.
Worked example
An ecommerce brand runs three ad sets: a 1% lookalike of purchasers for prospecting, a seven-day cart abandoner audience for conversion, and a past-180-day customer audience that is excluded from the prospecting set. Reasoning: each set has a distinct relationship to the brand, and the exclusion prevents paying to reacquire people who already bought.
Traps that cost marks
Leaving overlapping ad sets live so they bid against each other; building audiences so small they never exit the learning phase; using 365-day retargeting windows and calling it efficient; putting customers in prospecting campaigns; assuming a bigger lookalike percentage is better without testing; ignoring match rate on CRM-based lists.