Facebook. You can’t throw a stone in the marketing space without hitting something related to the social media giant. There’s a good reason, though. Over 80% of social ad spend in 2020 was on Facebook — a topic we dove into in our most recent webinar, Predictive targeting strategies to optimize and scale Facebook ads.
With Covid-19 continuing to rear its ugly head, uncertainty still abounds for brands considering retail space, and the logical choice for a brand forgoing brick and mortar is social media ads. For Faraday, many of our direct-to-consumer brands might not have brick-and-mortar locations anymore — the modern landlord is Facebook.
The question that’s front-of-mind for any brand advertising on Facebook is, “how can I stand out from my competitors?” It's all about Facebook audience targeting. Here are three ways to knock it out of the park.
Facebook targeting with core audiences
The first of the main targeting strategies used on Facebook are core audiences. Core audiences are developed within Facebook itself, and are generally pretty broad: I want to target every female age 20-35 in the country. Core audiences use the shotgun approach, where you place very few limits on your targeting so that your cost per impression remains low while you focus on the quantity of people you’re targeting.
On the flip side, because you’re targeting so broadly, many of your leads are bound to be low-quality consumers who aren’t likely to interact with your brand — especially as your campaign runs longer and Facebook’s algorithms start to bring in consumers at the far reaches of interest for your brand.
While the shotgun approach of broad core audiences tends to be fairly inexpensive, this targeting strategy has its drawbacks — like bidding for the same audiences as your direct competitors. Our approach: geotargeting.
Predictive geotargeting can drastically improve ROAS for campaigns targeting broad audiences by making use of the Faraday Identity Graph (FIG) and predictive modeling. Using these custom predictive models, you can score the country based on who’s most likely to make that leap to interact with your brand, providing you geographic locations of these high-propensity buyers. In this approach, your brand can keep the benefit of broad reach, but also ensure that your ad spend is focused on higher-quality leads.
Facebook targeting with custom audiences
A custom audience is an audience that you, as a marketer, bring into Facebook with the intent of marketing to. These often come with more limits and boundaries than core audiences, which means that the cost per impression will be higher.
The iOS update 14 hit these audiences in more ways than one: not only are the audiences taking longer to get out of the learning phase, but they’re also fatiguing faster due to the now-limited data available on Facebook.
Faraday takes a hand-in-hand approach with data and attribution. After surfacing unique attributes in FIG and using predictive models to determine the most likely buyers with minimal overlap over your Facebook audiences, we’re then able to attribute transactions to these buyers that we surfaced to provide a deeper, more meaningful understanding of their customer profile.
By doing this, you’re able to create a more complete view of both the why and the how for a buyer’s interactions with your brand.
Facebook targeting with lookalike audiences
With lookalike audiences, you’re supplying a group of people to Facebook — a seed audience — and telling their algorithms that you want them to find more buyers that look just like your audience. Your seed audience is likely going to be a group of buyers who have taken a certain action: bought a specific product, made a purchase within the past month, and so on.
This decision is critical because it’ll be what Facebook’s algorithms hinge on for optimization. With lookalikes, you’re also likely to run into an exhaustion issue similar to custom audiences, where you then expand the audience to include less likely buyers which results in weaker performance of the campaign.
Parallel to the exhaustion issue, using lookalikes in this way leaves you with an inability to personalize to these buyers because you don’t really know who they are. Your creative and messaging aren’t necessarily aligned to the people you’re trying to reach, which is an issue because consumers expect personalization. When they see an ad that isn’t personalized, it stands out (and not in a good way).
With Faraday, building personas and persona models helps you to fill this gap so that you can utilize Facebook lookalike audiences more effectively. When using a predictive persona-based audience as your seed, you can reach new customers that are a great fit for your brand while simultaneously solving your personalization problem — if you start a campaign with a specific persona in mind, you can hone in on relevant messaging from the get-go.
Putting it all together
Your ads can only be successful if they're getting in front of the right audience in the first place. Facebook has several potential pathways to success in marketing campaigns, and by making use of predictions to power them, you'll find it's completely possible to break through performance ceilings.
Wondering what predictive audiences can do for your team? Schedule a personalized demo to discuss your goals with a growth expert.