Marketing teams with growing sales targets are always looking to reach larger audiences. Because of look-alike modeling’s ability to reach audiences beyond a marketer’s CRM, it is a solution every marketer wants in their toolbox. But what is this tool and how is it used? In this post, we pull back the curtain by exploring the what, why, and how of look-alike modeling for digital marketers.
What is Look-alike Modeling?
As defined in Advertising Age, look-alike models are used to build larger audiences from smaller segments to create reach for advertisers. The larger audience reflects the benchmark characteristics of the original audience. In the context of marketing, look-alike modeling can be used to reach new prospects that look like a marketer’s best customers.
Why Use Look-alike Modeling?
Look-alike modeling is an established and well-proven strategy. A study by Exelate found 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 the tactic.” – Exelate Study
A recent study conducted by Forrester Consulting on The Total Economic Impact™ Of LiveRamp Connect found that “sales driven by new customers were 2.5x higher following data onboarding than in previous campaigns” due to look-alike modeling that enabled the organization to target advertising to prospects that resembled its best customers.
Similarly, in a second case study, a bank’s top customer data was used to build a look-alike model using 3rd party financial data from Datamyx. LiveRamp onboarded the modeled audience from Datamyx into the bank’s DSP, Collective, so that the additional prospects could be reached with display ads. This resulted in a 150% response rate improvement and 40% ROI improvement over campaigns towards a general audience.
How does Look-alike Modeling work?
The look-alike modeling process typically involves the use of data enrichment to expand the set of attributes that are used to create the modeled audience. Using third-party data within a data provider or a DMP (Data Management Platform), the smaller seed audience of customers can be enriched with added attributes. This enriched seed data can then be analyzed for patterns of similarity with the data provider’s total audience. The larger audiences based on this enriched data will then “look like” the seed audience.
This process can surface valuable attributes that lead to a higher performing model than general marketing segment buys such as gender or age. By finding audiences that the marketer would otherwise be unable to identify, look-alike modeling becomes a key marketing tactic for new customer acquisition.
How Do You Get Started With Look-Alike Modeling?
Given the many benefits of look-alike models, how do you get started implementing the strategy? Marketers use two common approaches to look-alike modeling, both involving 3rd party data.
- When building look-alike models from email or direct mail lists, marketers can work directly with a 3rd party data provider to build offline model audiences.
- Alternatively, when using online customer data as the seed, such as a highly active website audience, marketers can build look-alike models through data management platforms that house 3rd party data.
But even when starting with offline seed data, digital marketers can reach their model audiences within digital channels such as display, video, and mobile. In the case with Datamyx above, a bank’s offline seed data was used to build an offline look-alike model. The Datamyx model audience was then onboarded by LiveRamp for targeting on digital display in Collective. In another approach, a marketer’s offline seed audience might be onboarded to a data management platform such as Adobe Audience Manager, where it can be enriched by online data to create a look-alike audience for targeting on digital channels.
The advent of data onboarding creates exciting new possibilities for digital marketers interested in look-alike modeling. By engaging an onboarding service, marketers can use offline seed data and 3rd party data modeling to target larger audiences on digital channels. These model audiences will have the characteristics of the brand’s most valuable and engaged customers, leading to higher engagement and ROI than general campaigns.