Every audience we review looks simple from a distance and varied up close. A single list can hold students and parents, two big cities, and people at very different stages of life. Sending them all one message means most of them get the wrong one.

Segmenting before you spend is the cheapest improvement most campaigns can make.

Split by city first

Location is the most reliable field in most audiences and the one that changes your message most: office address, event venue, delivery promise, even the examples you use.

City splits in reviewed audiences
AudienceLargest locationSecond locationWhat it means
HannahOlalaHo Chi Minh City 52,734Hanoi 33,976Two national campaigns, then regional ones
IELTS Forecast GroupHo Chi Minh City 2,675Hanoi 2,237Near-equal budgets for both cities
IPL ScholarshipHo Chi Minh City 3,589Hanoi 886Lead with HCMC and the southern provinces
Columbia UniversityNew York City 2,900Manhattan 1,768Split NYC, New Jersey and Connecticut
KeapCanada 236 (country level)Ontario 192Separate US and Canada campaigns
Motorbikes parked at a street stall in Hanoi
Hanoi and Ho Chi Minh City audiences often respond to different offers, venues and examples. Photo: Unsplash.

Then split by age

Age data is thinner, because only some profiles include a birthday, but it is still useful. In the IELTS Forecast Group review, 55.6% of profiles with a birthday were aged 18–30 and 33.8% were 30–60. The younger group is preparing to study; the older group is more likely preparing for work or migration. Same test, different reason, different message.

Use life stage where it exists

Relationship status is a simple proxy for life stage. In the HannahOlala audience, 42.5% of profiles that list a status are married, which opens family and home categories alongside beauty. In the IPL Scholarship audience, 67% are single, which fits early-career offers.

A busy street corner in New York City
In the Columbia University audience, one in three profiles is in the New York tri-state area. Photo: Unsplash.

Four rules for segments

  1. Keep segments big enough. Ad platforms set minimum audience sizes. Check yours before splitting too far.
  2. One change per segment. Change the city, the headline or the offer, not all three at once.
  3. Name segments clearly. "IELTS_HN_18-30" is easier to read in a report than "Audience 3".
  4. Report by segment. A campaign average hides the segment that is paying for the rest.

Segments make testing possible. Next: Testing Outreach Without Burning Your List. Or explore audiences and check their city splits.