The Benefits and Limitations of Look-Alike Modeling

Demand Gen Report recently published a white paper describing the benefits of using look-alike modeling powered by artificial intelligence (AI) to improve lead generation performance. The white paper argues that B2B marketers can use "AI-fueled" look-alike modeling to get more qualified leads that convert at higher rates.

The principles underlying look-alike modeling aren't new. For years, astute B2B marketers have been identifying important attributes of their best existing customers and using those attributes to create a profile of their "ideal prospect." Then, they would use this ideal prospect profile to identify target audiences for outbound lead generation programs and otherwise guide lead generation efforts.

The current incarnation of look-alike modeling does essentially the same thing, but in a more sophisticated way using AI-powered data analytics.

Several technology providers now offer solutions that include or support look-alike modeling, and most of these solutions take similar approaches to the look-alike modeling process.

• They extract data regarding a company's existing customers from the company's internal technology systems including, but not necessarily limited to, the CRM and marketing automation solutions.
• Most solution providers have developed or obtained access to extensive databases regarding business organizations. The modeling solution will combine the company's internal customer data with any additional data regarding these customers in the provider's database. This enables the solution to create a more detailed picture of the attributes of the company's existing customers.
• The modeling solution then uses an algorithm to analyze the combination of internal and external customer data to identify the attributes that the company's existing customers have in common. The result of this analysis is usually called a customer data model.
• The solution then runs the company's customer data model against the provider's database of businesses to identify companies that resemble the model.

The major advantage of AI-powered look-alike modeling is that it incorporates far more data points than humans can realistically use when the process is done manually. Therefore, AI-powered modeling enables marketers to build a richer and deeper customer data model, and it does a better job of identifying companies that are likely to be good prospects.

Look-alike modeling can be an effective tool for improving B2B demand generation performance, but like any business tool or methodology, it has some limitations.

First, for look-alike modeling to be effective, a company needs to have enough existing customers to build a customer data model that's reliably predictive. One provider of look-alike modeling has indicated that a company needs at least 500 existing customers to build a reliable model. While 500 may not the the absolute minimum, effective look-alike modeling does require a company to have a substantial number of existing customers, and a start-up or young business may not be able to meet this requirement.

Second, look-alike modeling can be less effective when a company is marketing new products or services. If a new product or service appeals to a different type of customer than the company's other products or services, a customer data model based on the company's existing customers may not identify the right prospects for the new product or service.

The important point here is that look-alike modeling is a powerful tool for improving demand generation performance, particularly when it's enhanced with artificial intelligence. But B2B marketers should also remember that like any business methodology, look-alike modeling has a few important limitations.

Image Source:  Flickr.com