The Unexpected Risks of Performance-Based Billing for Retention Software
"Only pay if it performs"–Is this a risk-free deal?
Nobody wants to pay for ineffective software.
This is why performance-based billing models have a "risk-free" appearance. It's all upside, right?
When it comes to failed payment recovery, or any customer-retention tool, there are common pitfalls to performance-based billing. The results can be frustrating, as they drive business decisions that come with outsized hidden costs (both in the form of lower retention, and higher fees).
After reading this short post, you'll learn how to avoid the biggest risks—and find a better way to measure changes to your customer retention.
Recovery Rate is the metric often tied to performance-based billing.
Recovery Rate is defined as the percentage of failed recurring payments that are recovered before a customer cancels their subscription, or before they are written-off as churned.
Recovery Rate isn't the best way to measure before & after performance (more on that later). But if decision-makers are measuring performance by this standard, let's look at how to get it right.
Don't trust, verify
There's only one thing you need to do: verify the data.
Export the customer data specifically used for your “before” measurement. Without the data, you can’t verify the results you are seeing. Without data, all you can do is trust results from the tool that wants to bill you based on performance improvements.
You may be surprised by how difficult it is to gain access to this data. When billing is performance-based, this will often be the case (by contrast, this data is available with a button click in Churn Buster).
The good news is, you only have to do this once to establish an accurate "before" baseline to compare performance against.
And once you have the data, it’s easy to spot check customers to see if any have been mis-categorized as “churn” when they still have an active subscription.*
*Note: our team helps with these spot checks when asked, and has found differences of over 100% in “before” baselines (for example, seeing a 25% baseline recovery rate when 50% of payments were verifiably recovered).
Wait... that's all?
Yes, all you need to do to compare recovery rates based on verified results.
With this approach you won't be over-charged by performance-based billing. Or fooled by incorrect reporting that hides the damage being done to your business.
Our team has helped companies visualize hidden losses that amount to hundreds of thousands of dollars annually. These findings are tough lessons, but necessary ones.
Below you'll see the limitations of recovery rate as a metric, and the calculation method our data team uses when verifying results.
Recovery Rate Measures More than Passive Churn
Your overall churn rate varies month-to-month fairly dramatically.
Here’s our own customer churn rate, ranging from 0.78% to 2.48% per month:
Top to bottom, that's a difference of 104%.
Recovery rate is connected to churn rate, and you will see similar swings up and down month-to-month.
When using recovery rate to measure dunning performance, you are also measuring the many other factors that affect churn rate:
- Random variance
- Price points
- Billing cycles (annual subscribers have a lower recovery rate than monthly)
- Customer acquisition (promotions, no-card-required trials, etc)
- Service quality
These factors all contribute to churn rate.
And after a payment fails, if retries to the card-on-file aren’t successful, you have to ask a customer to take action to correct the issue.
And that’s when they make a choice: “Do I update the card? Ignore the issue and passively churn? Or actively cancel?”
The choice they make will show up in your recovery rate.
In this way, recovery rate measures more than the performance of your dunning tool.
Looking for a better way to compare before/after performance?
Our team has developed a non-biased method for comparing dunning tools, after helping thousands of subscription brands understand their retention.
This approach uses 3rd-party payment processor data, and accommodates any differences between two retention tools for neutral, transparent results.