When creating a lookalike audience on Facebook, it’s important to start with a strong foundation. With a good custom audience, Facebook will find people who closely resemble our most valuable customers. If done correctly, we then have a list of names who have the potential of becoming customers.
During your research on Lookalike Audiences, customer lifetime value is a term you’re likely to have encountered. As Lookalike Audiences has grown in popularity, the confusion surrounding this term seems to have grown somewhat. In this guide, we want to uncover the basics behind customer lifetime value as well as explain how we can utilize this metric in future campaigns.
Essentially, this term describes the net profit a given customer brings to a brand over the course of their interaction. Therefore, there are a number of different factors that contribute to customer lifetime value:
One of the most important aspects of this metric is how it relates to building the custom audience list in lookalike audience marketing. By sending this data to Facebook, it will use this list of elite customers in order to find others who have the potential to provide the same value. Since they have similarities to your existing core customer base, you should know how to target them and convert them into reliable and long-lasting customers.
Thanks to Facebook, we actually have some advice from the platform itself regarding the process of gathering valuable customers. In particular, they advise AGAINST ranking and rating.
In the past, plenty of marketers have ranked their customers according to their worth for the brand. If we have 50 customers and they all have a worth of between $400 and $1,000, they might rank them from 1 down to 50. According to Facebook, the main problem with this is that ranking is never proportional. It’s all well and good seven being ahead of eight, but what if the top seven are very similar and the eighth is worth half as much as this elite group?
All you have to differentiate between customers is a single rank; it will never display the extent to which two customers are different. Ultimately, the difference between each rank can range from less than 1% to a much higher percentage.
Elsewhere, other marketers decide to rate their customers. Unfortunately, we have the same problem where customers aren’t given the appropriate rating. In order to compare two customers, they tend to be factored by different extents. For example, a $100 customer could be factored by 20 to get a rating of ‘five’ and a $10 customer could only get factored by five to get a rating of ‘two’. While the $100 customer provides ten times as much value, they only have a rating of five compared to two.
Since Facebook is the one handling the data, it seems wise to listen to its words of advice. So, what does it say?
For some marketers, they feel as though the best results come with only the highest value customers included in their custom audience. In reality, we need to have a broad range in order for Facebook to assess the quality of a good customer. When Facebook doesn’t have a good range, it cannot learn how to distinguish between average customers and fantastic customers (we all want the latter, right?).
Facebook can do lots of things, but it cannot bring several different currencies into a universal currency. With this in mind, make sure that you only use one currency or scale when offering value in this way.
Next, it’s important to understand that Facebook only accepts positive numbers for customer value. We understand that you want to avoid certain customers, but this doesn’t mean you can use negative numbers in order to show the type of people you don’t want in your lookalike audience.
Lastly, Facebook cannot work with separators and punctuation marks. Don’t worry, decimal points are absolutely fine for cents.
How do we set up a Lookalike Audience campaign with customer lifetime value (which we’ll now refer to as CLTV) as the main metric? We know that Lookalike Audiences works, and we could point to thousands of stats to prove this. However, this doesn’t mean we can neglect the process and hope for the best; we need to take control and make good decisions.
Every Lookalike Audiences campaign starts with the custom audience. In the Business Manager, head to the dashboard and choose ‘Audiences’. At this stage, many people make a mistake by choosing Lookalike Audience. Before that, we’re more interested in creating the Custom Audience, so select this option. Some will say that you can do it the other way around but going for Custom Audience first allows you to choose CLTV from the beginning.
Next, you’ll be asked how you want to create the audience and we recommend ‘Customer File’. For some marketers, they utilize the options to base the custom audience on website traffic or activity on your app. In our opinion, it’s better to upload your own file since this will bring a list of names that most closely resemble your most valuable customers.
After this, you’ll see an option to ‘Include LTV for better performing lookalikes’ and you should choose this. Although it’s time-consuming, you will be able to assign customers with a value; those operating on eCommerce stores will have access to this sort of data. Thankfully, Facebook provides very good instructions and help so that you can get this right the first time.
Once you’ve edited the data mapping and chosen a customer value type, it’s time to let the creative juices flow by creating an ad that will get conversions in your new audience. However, there’s an element of fear here because you’re nervous about appealing to people similar to your customers. Many marketers feel this pressure that they need to get it right or face missing out on real value. However, we view the similarity and ad creation process as an advantage. You know what converted your existing customers, so you have an advantage when reaching out to this new audience.
As a word of advice, don’t forget about the different funnel stages. Just because this audience has similarities doesn’t mean they’re a guaranteed conversion. Always be mindful about the content of your ad and the calls to action you include. If you ask them to buy straight away, you aren’t likely to see good results. Most won’t have even heard of your brand before, so asking them to spend right off the bat will likely backfire.
Like a normal campaign, start by generating awareness. Initially, you can expect your audience to know about your industry and to even have some knowledge of the products you provide. With this in mind, aim for the middle of the funnel with your creatives and copy. How will you encourage them to take their interaction with the brand further? We recommend coupons, webinars, lead magnets, and other relevant content.
Don’t forget, your work isn’t over just because the campaign is live. We know that split testing has a reputation for being time-consuming and difficult, and we know there are a million things calling for your attention in your business, but Facebook keeps it simple. With this market in particular, they aren’t brand aware which makes split testing all the more important. Test different creatives and gather the results to see which resonates better with your audience.
Finally, not everybody will convert…but this doesn’t mean you lose them forever. If you’re split testing, it makes sense that some people will see creatives and copy that were eventually replaced by more effective solutions. In the Ads Manager, why not create a second custom audience based on the people who didn’t convert?
We recommend creating the audience through people who have interacted with your page; exclude people who interacted with the CTA, include people who engaged with posts and set this campaign to run parallel to the first. With the two working together, you’ll be converting while also catching a percentage of those who slip through the first net.
Discover valuable tips, tricks, and industry news with a specific focus on the role of artificial intelligence in social media advertising.