In June of this year, the world was introduced to Facebook’s Value-Based Audiences for advertisers. This means we’re not only able to indicate which users are our customers, we’re also able to give Facebook some indication of who are our better customers. If you haven’t read up on the basics of LTV components in Custom Audiences, check out this post before you venture any further.
The biggest impact this LTV component can have is on the Lookalike Audiences you create from lists that utilize the LTV metric. As Facebook puts together its 2M+ audience based on your LTV Custom Audience it’s able to put more weight on those users with higher lifetime values as indicated by your values and suppress those with lower ones to, in theory, create an even more valuable lookalike model.
Sounds great, right? Unfortunately, in some cases…
Customer Lifetime Value is Hard to Find
I’ve had numerous clients be very excited about the potential of an even smarter Lookalike Audience model, until it came time to actually calculate customer lifetime value. For eCommerce stores, this can be relatively easy as you’re usually tracking the order values for each of your customers. Even if the CRM or Marketing Automation platforms are a mess, you can typically formulate a fairly close answer to lifetime value.
For lead generation, however, it can be an entirely different story. For better or worse, many lead gen companies simply don’t have the ability to determine individual lifetime values for customers. This can be caused by inconsistent pricing practices, business model changes, lack of follow through on manual tracking, or any other number of reasons. In other instances, I’ve worked with companies that simply don’t have enough customers for a big enough list.
The good news?
None of these issues mean that Facebook Value-Based Audiences are out of reach. But before we get to the fun stuff, let’s get a sense of the rules.
Guidelines for Value-Based Audiences
A little creativity can open up a world of possibility with Facebook value-based audiences, but there are some limitations. Since the strategies we’ll be talking about (or any others you come up with) utilize this new functionality a bit outside of it’s normal parameters, it’s important to keep Facebook’s guidelines in mind while putting together your value-based audiences.
It’s All Relative – Don’t Include Ratings/Rankings
In an attempt to make things a bit easier than finding actual lifetime value, it might seem like a good idea to use ratings or rankings for your custom audience.
- Ratings: Similar to what we do with stars, creating a scale of 1-5 and giving each customer in the list a value of 1-5.
- Rankings: If you have 50 customers, this would mean putting them in order and giving them a score of 1 to 50.
Although these both seem like an easy way to accomplish to the goal, the ease of setup will most likely hurt your returns. And there’s a simple answer as to why:
Relativity.
Facebook needs a sense of how valuable each customer is in comparison to another. In the rating system example, an easy way to test this setup is to ask the question:
“Do five 1-star customers equal the value of one 5-customer?”
More than likely, the answer is going to be “No”. Either those 5 customers are going to be more or less valuable than the single customer for whatever reason based on your business pricing model. And that’s totally fine, but it means your rating system isn’t doing justice to your customer’s values and the results with value-based custom audiences and/or value-based lookalike audiences might suffer for it.
When developing a scoring system, consider the relativity of each score to show Facebook that not only is one user of higher value than another, but that it’s on a scale that makes sense relative to each other.
Leverage a Full Range of User Values
Don’t include only your top customers. As with nearly everything in digital advertising, the more data you give Facebook the better. With indicators to show higher and lower value customers, Facebook will still consider the whole list when developing new Value-Based Lookalike Audiences, but it will emphasize those with higher values and suppress those with lower values.
Don’t Use Negatives
Depending on how you determine your lead scores, there is such a thing as a negative lifetime value. Facebook indicates that we should not use negative values, so if we decide to use a metric that ends up with a negative in place, we’ll have to get a little creative. In an example below, I’ll show you how we were able to overcome this fairly simply for one client.
Now that we’ve set up our guidelines for the substitute lifetime value metrics, it’s time to dive into what those could possibly be. The easiest place? Your CRM.
Using Other Metrics from Your CRM:
What do you do when customer lifetime value isn’t something you’re able to easily find or find at all? There are a number of ways you can still utilize value-based audiences by getting a little creative with your user list. Here are some other examples:
1.) Length of Client Relationships
Some companies charge based on a regular subscription fee, typically on a monthly basis. It can be fairly simple to take the monthly recurring revenue and multiply it by the number of months someone has been around to get an estimate of customer lifetime value. Likely you would use the date they became a customer, the current date, then create a formula to count the months in between. Easy.
But some companies also offer discounts for annual plans. In terms of bottom line financials, this annual plan might be lower perceived revenue than 12 months of MRR, but there’s bound to be a certain level of churn found in the monthly recurring plans. For example, here is pricing scale for Unbounce:
Users are given a 20% discount when they sign up for Annual vs. Monthly billing.
Although I’m not sure on Unbounce’s thoughts on the matter, many companies value users who sign up for an annual plan higher than recurring revenue, even if the value has the potential to be slightly lower.
A bird in the hand…if you will.
So for these types of calculations, it may be valuable to first segment your audience into monthly vs. annual plan subscribers. For monthly users, simply identify the number of months someone has been a client and utilize those values to determine customer value. It also doesn’t have to be a linear scale of 1 month = value of 1, 2 months = value of 2. You can give credits when someone hits a 6 month, 1 year, or 2 year mark. Maybe everyone over 6 months gets a 2 point boost, everyone over a year gets a 5 point boost and everyone over 2 years gets a 10 point boost to value them even more heavily.
2.) Lead Scoring
One client we work with has a list of just over 155,000 active, paying customers. This is a fantastic list to use with a wide range of customer values. But another company we work with has only around 100 paying customers, which would create a very small custom audience. So what do we do? Rather than forfeit use of value-based audiences, we get a little creative. For this client, we decided to include their entire CRM database (a total of 2,900 matched users) and utilized Lead Score as their lifetime value component. We had just one problem:
Lead scores ranged from -1000 to 1450.
Now what?
According to the guidelines above, our numbers cannot be negative, so we have to adjust them numbers to be positive. By adding 1000 to every user score, we were able to keep the relative scores intact, utilize the full range of users, and avoid any negative numbers.
Additionally, we wanted to give Customers and Opportunities a boost as their Lead Scores weren’t always proportional to their positions in the lead funnel.
Customers were given an additional 2,000 point boost and Opportunities were given extra 1,000 points to differentiate them from SALs, MQLs, and Leads. It’s still not a perfect science, but these boosts allowed us to better indicate our higher value list members to Facebook.
3.) Lead Stage
Not all companies employ a Lead Scoring method like the one laid out above, but most have some sort of Lead Funnel with lifecycle stages that can be leveraged in some way. Effectively for this type of list, you’ll be developing a rough Lead Score to use in your value-based audience. Your grading system can be as simple or as elaborate as you like.
A simple model based solely on a 5 stage cycle could look something like this:
You might develop the scores by each stage based either on the average value of each stage or by using conversion rates to say that “10% of Opps become Customers”, “50% of MQLs become SALs”, etc. to help you develop your scale. There’s no right or wrong answers here, but we will touch on some guidelines shortly.
If your list is big enough, you can get a little more sophisticated than the example above. If you’re not only looking at the individual stages but also if they’re still qualified (Q) or disqualified (DQ), you can add an additional wrinkle into your scoring methodology.
Again, there’s no right or wrong answer here, so get creative and utilize all that your CRM is giving you insights into.
4.) Account Size Indicators
I’ve written before about one account where we’re trying to generate leads for companies with larger accounts and how the targeting wasn’t producing the results we wanted. So in another attempt to shape our targeting to be more focused on larger potential accounts, we started leveraging the value-based lookalike audiences.
For our custom audience upload, we utilized the Employee Count reported by the customer as the value in our audience in hopes that our Lookalike Audience would then result in higher employee counts on average. So far so good, and although our original goal was to not only focus on higher account sizes, we’re also noticing that we’re seeing more of those higher accounts convert into customers. So with this strategy, we’ve killed two birds with one stone: better conversion rates AND bigger account sizes.
5.) Your Own Component
The list I’ve laid out today certainly doesn’t even begin to cover all of the creative ways you could leverage your database in the value-based audience model, but hopefully I’ve given you some things to think about. Don’t limit yourself to only a true lifetime value component. Think about those indicators that make one customer more valuable than another and turn those differences into relative numeric values that can be used in this new functionality.
If you find something that works well the first time, wonderful! If you need to go back to the drawing board, retool, and try again for better results, don’t be afraid to test different pieces and see how they impact your results.
Conclusion
Calculating actual lifetime value can be a complex art and science, so don’t get too bogged down in that process. Instead, develop a value system that can work for your customer based (or entire CRM list) that follows the best practices Facebook provides.
What other ways have you found to utilize the Lifetime Value component in Facebook without actually using LTV? Share with us in the comments!