Recover helps subscription companies reduce accidental customer churn and boost top-line ARR by 5+% using machine learning to automate the recovery of failed payments. The system retries failed payments to rescue transactions and customers that would have been lost due to accidental payment failure.
For a typical $100M subscription business, Butter finds $10M in failed payments, delivers $5M back to the business, and keeps hundreds of thousands of users connected with the products and services they love.
Recover leverages patented Machine Learning (ML) models to analyze dozens of variables and deeply understand the real-time state of any given payment, allowing optimization for authorization success.
Using dynamic error code handling, insights on factors such as banks, cards, currencies, and geography, coupled with optimized retry timing down to the second, Recover is able to formulate the best recovery approach for each individual payment. New models are constantly being tested against each other to consistently improve the performance of Recover.
By focusing on processing legitimate failed payments and minimizing retries, Recover saves customers, revenue, and reduces fees and chargebacks. Instead of losing a customer due to a failed payment, Recover allows you to continue billing that customer for subsequent billing periods, boosting their lifetime value.
Recover offers a multitude of configuration options for personalizing its operation for your business.
A Dunning Cycle is the length of time determined by the business in which a failed payment will be attempted to be recovered. Once this period is over, the payment and customer are considered lost.
A common length many merchants settle on is 30 days. The longer time goes on without a payment being recovered, the less likely it is to be recovered at all. Many businesses find that 30 days is a good balance between giving enough time for a good chance at recovering a failed payment balanced with the higher costs (in the form of payment processing fees) of continuing to retry a payment that is likely unrecoverable.
With Recover merchants can configure the Dunning Cycle length that they would like Recover to operate with and the underlying models will dynamically adapt their recommendations to account for this time period.
While Recover optimizes its retry strategy to limit unnecessary retries and reduce associated fees, many merchants opt to set a ceiling on the maximum number of retries that should be made. Recover is equipped to take this configuration into account and adjust its strategy to stay within the limit set.
While the optimal value differs based on the unique aspects of each business, there are often diminishing returns after a certain number of retries where the likelihood of recovery decreases.
It is better to start with a higher number of maximum retries and let Recover do the work of reducing retries vs starting with a lower number and potentially leaving money on the table.
Somewhere around 7-8 max retries works well for most merchants, and Butter is happy to work with your team to configure an option that works best for your business.
Recover via the Recover Enterprise API offers the ability to pass custom fields that are relevant to your business over to Recover. These can then be used for filtering out specific transactions that your business does not want to retry and for reconciliation purposes when information is sent back to your system.
For more details on how to pass custom fields please look to the API Reference for the send payment information method of the Recover Enterprise API below.