Audience segmentation

Marketing Analytics for Business

Sarah DeAtley

Principal Data Scientist

Audience segmentation basics

group of multi-colored people icons

Audience segmentation: identifying groups within an audience to deliver tailored marketing

Many different ways to segment audiences:
- Demographics (age, geography, gender)
- Psychographics (values, opinions, lifestyle)
- Behavioral data (engages with ads, abandons shopping cart)
Segments can be very granular; check that marketers can reach a segment

Icons and descriptors of demographics, behaviors, interests, psychographics, and audience size

Cluster analysis: statistical model where data in a group (cluster) is more similar than data in other groups (clusters)

Automatically groups data, but analyst determines how many clusters are needed
Analysts can look for common themes in clusters (like business product affinity)

network graph visualization with various nodes

¹ "20111214-NodeXL-Twitter-User jowyang network graph no jowyang edges" by Marc_Smith

Marketers can scale segment definitions across channels and campaigns
DSPs and DMPs manage segment definitions
- Enable retargeting and look-alike model tactics

bullhorn on left and multiple people on right

¹ mohamed Hassan from Pixabay

Retargeting: target audiences based on behavioral segments

Look-alike models: target large audiences with similar characteristics to segments

Retargeting with people on top then cart then email icon. Look-alike modeling on bottom with people then larger group of people

Marketing Analytics for Business