Look-alike Modeling

Written by Ido Tanne

What is Look-alike Modeling? Look-alike Modeling is the method of finding out who to advertise what to.

It can be categorized into three different categories. Each category has specific details to it as well.

Demographic has:

• Age

• Sex

• Income Level

• Marital Status

Psychographic has:

• Hobbies

• Interests

• Habits

• Habits

Behavioral has:

• Videos Watched

• Social Media

• Articles Read

• Websites Visited

• Items Bought

• Keyword Searches

Machine Learning AI helps to make behavioral targeting much easier since it can work around the clock and learn while doing so. For example, after spectating 100 different users machine learning AIs can easily predict what a user would like to purchase.

Each User has a Cryptographic Hash, which is a privacy protected small file that identifies a user and the things an advertiser knows about them. It is the tool to identify their behavioral lookalikes. It never expires and it can’t be lost, it is also portable, so it can be used on social media, physical mail, and email.

By looking at one user it’s possible to categorize them under a section to where an advertiser can submit their ads to that category.

How do these advertisements work?

1. Retargeting: After the user visits the site their hash key is submitted to the site and from there they’ll receive ads related to that site. When your hash is spread to more than one site the advertiser will produce a higher income.

2. Create Behavioral Lookalike Audiences:

3. Leverage Behavioral Lookalikes from Multiple Platforms: