If you are running any significant ad volume on Facebook, you probably run into a few problems over and over:
1. Audience filters that don’t match your audience (or ones that match don’t seem to actually be a good fit).
2. If you run sophisticated ad testing, you risk bidding against yourself if you test them against the same audience.
3. You are frustrated with the options for interest targeting (no “AND” targeting within interests)
Thankfully there are now certain tools being launched in private beta that help to solve all of the problems above. These tools will improve testing efficiency, expand potential audiences while refining targeting options, and drive down COCAs for customer acquisition efforts. All of these tools are unavailable to the public currently, but are live on our client accounts. If you are running $1,000+/day of media you should be able to ask your Facebook rep to make these features live (however #2 is only available through Epsilon). Read on for more details about why these tools are important and how they can improve your customer acquisition efforts.
1: Audience Segmentation:
A critical aspect of Social Fulcrum’s methodology is copy testing, which is used to find the optimal input for every facet of copy. Although the effectiveness of our approach is proven, there is one drawback: running a creative test to one audience means some overlap in delivery between audience members. While Facebook ensures that this does not increase cost per impression, it can increase the amount of time it takes to achieve statistical significance, which ultimately affects net spend.
Now, with Facebook’s new audience segmentation tool, we have the ability to eliminate overlap completely. Facebook takes its entire population and breaks it into groups of 5%, 10%, 20%, 25%, 33%, and 50% segments. So for example, if we ran a picture test with 10 images to an audience of 100,000 people, we could run one photo to each 10,000 person segment, ensuring no overlap. Net effect? The same, effective results delivered in less time and at a lower COCA.
2: Epsilon Audiences:
Most of Facebook’s behavioral audience targeting is supplied by third party data vendors and derived from online activity. While this is certainly useful, the targeting groups available through the Facebook platform are limited to a small number of relatively broad categories. So, we decided to partner with Epsilon, the largest database marketing agency in the world, in order to increase our behavioral targeting options by 10x.
Epsilon has an enormous database of offline consumer transactional, survey, compiled, logically derived, and financial data. Thanks to our new partnership, Social Fulcrum – and our clients – now have access to that database too. This means more data, more filters, and more effective targeting.
3: Overlapping Audiences (Boolean Targeting):
This one gets relatively technical, but it’s also the biggest deal. Facebook’s traditional targeting uses “and/or” functions. When one adds multiple interests or behaviors to an ad set in Facebook, they are using an “or” function. When one combines an interest with a behavior, they are using an “and” function. While this is an effective tool, it can be limiting for our clients who need their targeting to be further fine-tuned.
With our new targeting tool, such limitations are a thing of the past. We now have complete control of the “and/or” functions and how many layers of functions we use. This greater flexibility in filtering will lead to more accurate targeting, which in turn, leads to lower COCAs for our clients. See the example for a generic sportswear e-commerce company below:
Old Facebook Targeting –
Ad set 1: [interest: soccer or football or basketball] AND [[behavior: fitness] Audience Size: 1M
Ad Set 2: [interest: soccer or football or basketball] AND [[behavior: Online buyer] Audience Size: 1M
New Facebook Targeting –
Ad Set 1: [interest: soccer or football or basketball] AND [[behavior: fitness] AND [[behavior: Online buyer] Audience: 500k
Ok that’s the update from us on stuff that has us excited. If you’ve made it to the bottom of this blog, congratulations! You are as big a nerd as we are 🙂