While it costs retailers far less to retain an existing customer versus growing their new customer base, new customers are essential for business growth. And we have found that paid social advertising is a great way to reach new audiences at a reasonable cost.
Not long ago, Adlucent recently took over paid social advertising on Facebook and Instagram from the internal team of a client who manufactures and sells high quality bedding and other home décor items. This retailer understood their customer profiles pretty well, but needed creative insights to help them more effectively represent their brand to potential customers who fit their profiles. To help them gain those insights and optimize their ads, below are two tests we conducted.
1. Identify the Right Messages
We got started right away—developing a rigorous account structure that allowed us to test creative across audiences without driving up advertising costs through internal competition. This was easier said than done, but we were happy that we put in additional work to set up a proper A/B test because the insights were pretty interesting.
As you can see from the chart above, the retailer’s “Exclusive Brand Line” and “Pets” creative categories were the most compelling at the top of the funnel while also driving the most revenue at the best efficiency. The test provided compelling evidence for our client that their future creative efforts should focus on these two categories. The test also provides launch points for additional analysis and further testing. For example, the “Bedding” category drove plenty of interest, but its Revenue and ROAS weren’t much to write home about. This could have been because of a disconnect between the ad and the landing page in which shoppers liked the bedding ad but couldn’t find what they were looking for on the site. The top-of-funnel/bottom-of-funnel performance difference in this creative category could also indicate, or be counter-balanced by, above average remarketing performance for the bedding category. The devil lies in the details in the digital advertising world and the paid social channel is no exception. So the big takeaway we have from this test is: Perhaps it’s time to create a carousel ad (typically a high performer) using the best performing single images in the top creative categories.
2. Let Data Drive Your Paid Social Audience Tactics and Budget Allocations
Another great test to run is focused on audience. But we must preface this by saying that we recommend being careful with your testing tactics–running too many tests at once can add confounding variables which obscure insights and create unnecessary performance inefficiencies. In this retailer’s case, we were disciplined with our architecture and we managed to produce insights that are relevant to all paid social advertisers who use lookalike audiences. Test lookalike audiences if you don’t already use them, as they often have the best e-commerce return.
As you can see in the above chart, Facebook seems to have its lookalike algorithms calibrated like a fine Swiss watch. 1% was clearly the best audience during this test, and we see performance tiers forming with 1% or 2% occupying the top tier, 3% or 4% occupying the medium tier, and 5% or 6% occupying the lowest tier. Although our data seems to present an open and shut case, we recommend you apply our insights with a nuanced understanding of your business challenges. For example: we don’t necessarily recommend you avoid lookalikes at 3-6% just because they don’t perform as well as 1% and 2%. An account with aggressive traffic and gross revenue growth goals likely has a greater appetite for 3-6% lookalike audiences compared to an account with strict CPA and efficiency goals. We also recommend you do your own testing – results can vary by account and disciplined testing is a great habit to build.
Overall, we recommend letting data lead the conversation, no matter which goal you’re solving and which advertising channel you’re focused on. Socrates once said “The unexamined life is not worth living.” Similarly, we believe that unexamined ad spend is not worth spending.
Have you seen similar results in your creative and audience tests? Let us know in the comments below.