Understanding Look-Alike Modeling

Marketers are constantly looking for different ways to reach larger audiences. One useful tool that you can add to your audience targeting arsenal is look-alike modeling. Find out what it is, why you should use it and how you can get started.

What is Look-Alike Modeling?

Savvy advertisers use look-alike modeling in conjunction with other audience targeting methodologies, such as IP targeting. Look-alike modeling is the process of building larger audiences from smaller audience pools in order for marketers to reach more people.

To give you an example, let’s say you are a clothing retailer and you’re having a sale and are aiming to increase online purchases. Essentially, what you want to do is divide your audience into groups of attributes (this can be anything including gender, age, location, or other data such as websites they visit, the music they listen to, etc.), and then determine any common behaviors you can find that rank most highly among the people with a tendency to purchase clothing online. Then, you look for people whose attributes match these behaviors and “look like” your current audience and advertise to them. Online advertising platforms, such as Google and Facebook, already allow advertisers to create models using online behavior.

While this may sound simple enough, in reality, the process is highly systematic and complex, and requires tremendous amounts of computational power. However, this is advantageous in that you will have the ability to find audiences that you would otherwise be unable to discover – people who behave in the same way as your current customers but have not yet purchased from you. Look-alike modeling can become an important marketing tool for new customer acquisition.

While Facebook and Google have the resources to use online data for audience/behavioral modeling, using offline data, which are usually available from 3rd party data experts, can provide you with a better picture of the consumers’ purchasing behavior.

Why Should You use Look-Alike Modeling?

In a nutshell, one of (if not the only) reason why you should use look-alike modeling is that it is a proven strategy. A study conducted by Exelate revealed that  look-alike modeling “results in double or even triple the results of standard targeting, according to the 30 percent of advertisers and more than half of agencies who reported using this tactic.”

In addition, in an article by Mashable, Facebook revealed that an unnamed online retail partner has experienced a 56 percent lower cost-per-fan acquisition cost and a 94 percent lower cost per checkout using Facebook Lookalikes Audience targeting.

How to Use Look-Alike Modeling

Look-alike modeling is an established practice and is advantageous to marketers looking to develop their audience pool. But how do you get started executing the tactic? The answer is 3rd party data providers. Whether you are building look-alike models using offline seed data and want to build larger audiences on digital channels, or using offline customer data as seed to create an offline look-alike model, you can do so by working directly with reputable 3rd party data providers.

Because model audiences will have the attributes of your current customers who are most engaged with your brand, you can expect an increase in engagement and ROI after utilizing look-alike modeling.