Facebook’s Audience Insights tool is being removed from the platform in July 2021, and as a result many advertisers will see their main source for targeting inspiration disappear. How can you build prospecting audience lists for Facebook advertising without the Audience Insights tool?
How can you build prospecting audience lists for Facebook advertising without the Audience Insights tool?
Facebook’s Audience Insights tool is being removed from the platform in July 2021, and as a result many advertisers will see their main source for targeting inspiration disappear. It will be harder to build new prospecting audiences and find new customers for your eCommerce brand.
Whilst Facebook is recommending a switch to Facebook Business Suite Insights where advertisers will be able to review and analyse metrics, it is worth making a note of other strategies for building out your prospecting audiences.
This update will make the process of setting up new audience groups for targeting more difficult and abstract. Luckily, there are many alternative approaches you can take when building out new customer targeting to ensure your paid social ads are reaching the right users and ultimately converting and driving additional revenue.
Google Analytics houses a whole host of data from your site traffic that you can utilise to help create your audience targeting. Outside of the obvious age and gender data, using the in-market / affinity audience data can help us to understand exactly what our new customers look like.
You can review the top revenue driving in-market or affinity audiences, and use this to target the related brands and interests within Facebook. This should give a similar audience to what you would previously have generated from the Audience Insights tool.
Your site’s referral data in Google Analytics can also tell us a lot about the interests of your site’s users, as we can see the other sites that your users also visit. These sites can be used as targets within Facebook as well to help us understand their general interests and find new similar customers.
Using data from other channels such as Google Ads can give further insight into how to best reach your target audience. Search terms from Google Shopping can be especially useful - these can be relatively broad but give a good idea of how your target audience is searching.
Audience Insights data from Google Search can also be a good source of information, with the brands you are competing with in these auctions likely to be relevant interests to target.
Ultimately, with the changes coming around third party tracking, first party data is going to become more and more valuable. Segmenting first party data into lists of the most valuable customers, or customers who engage most with your brands content, will allow you to generate lookalike lists to target. The users in these lists (in Facebook’s eyes) will be the most similar to your most valuable customers.
Surveying current customers can also help develop an understanding of how to build and expand your Facebook targeting - questioning customers on other brands they like, how they found your brand, other brands they would consider buying from etc. These questions can inform interests to target within Facebook.
There are other areas where this data could be useful too, such as understanding which aspects and USPs of your products to push in messaging, or new products to launch.
Although Facebook’s liquidity approach means we no longer have to narrow down targeting by interests, demographics, locations etc. - this is likely what will have led to the removal of the Audience Insights tool.
With our clients, we have seen some of our strongest paid social performance for new users coming from targeting extremely broad audiences, potentially entire countries or even continents, and letting the algorithm feed off the huge amounts of data it can gather very quickly. The algorithm can then serve the right ads to the right people.
Letting Facebook’s algorithm do the work sounds scary and counter intuitive, but with the right strategies and the right amount of control, you can balance its machine learning with your brand knowledge.
In order to deliver the most efficient performance from Facebook, testing will be key as there is no ‘one size fits all’ approach that will work for all brands. Testing the above approaches will give an understanding of if Facebook’s liquidity approach can deliver efficient performance vs. using alternative data sources to build out more specific targeting.
Once top-performing targets have been identified, these should be combined to ensure Facebook’s algorithm can combine as much data as possible - only splitting if ever needed to run different messaging per group etc.