Facebook ads offer a variety of marketing methods to choose from, but which is best? The method we found to be the game changer is the customer file upload with a lookalike audience. We recently ran two major social media campaigns with one key difference, the way we targeted our paid advertising. Both were B2B Facebook ad campaigns with a similar offer and the same goal: Influence prospective customers to fill out a form to request more information. Even with all the similarities between the two, one significantly outperformed the other on all metrics.
Facebook Ad Results
|Campaign A||Campaign B|
|Click Through Rate||0.23%||0.52%|
With campaign A, we had a list of the businesses we wanted to target and manually selected their company pages and employees. As you can see in the table above, the rates were unsatisfactory. So, we went back to the drawing board for campaign B.
For campaign B, we had an actual customer file with business contact names, email addresses, etc. We uploaded this file to Facebook and created a “look-alike” audience with the same characteristics as our customer file. The results speak for themselves.
After conducting this experiment, we realized we hadn’t really learned anything new. Rather, we failed to apply the things we already knew to the first Facebook ad campaign. There are professionals (like our data processing team at ANS) who can take an ordinary customer file and produce a “model” file of prospective customers who match the traits of your current customers. When a campaign targets these model contacts, the response rate skyrockets. Being in the direct marketing industry for more than 20 years, we have repeatedly seen the “magic” that lies inside a modeled customer file. It manifests itself in an impressive ROI. The Facebook lookalike audience applies these same ideas for social media marketing.
In contrast, many marketers make an educated guess at who their target customer is, much like we did with campaign A. In our experience, the results are about the same with any campaign. Using a customer file and modeled data proves to be more effective than making an educated guess.
For more information on data modeling or to request a data model, email email@example.com