Sometimes there isn’t a good day to retry payments because the likelihood of the customer having sufficient funds is really low. By enabling the failure filter you can let GoCardless use its intelligence to only allow retries where the likelihood of success is high, resulting in a better experience for both you and your customers.
What is a failure filter?
With intelligent retries, you can now apply a failure filter which will prevent retries if the likelihood of the payment failing is very high (90% or higher).
Is this feature right for my business?
We appreciate that this is not a straightforward decision to make as there are pros and cons to consider:
Pros
- It could provide a better experience for your customers:
Some UK and European banks charge payers for returned direct debit transactions with fees that can vary from £5 to more than £25. By applying the failure filter, you essentially help your customers avoid these fees which gives a better experience for them.
- It could save you time:
If the failure filter is applied, you’ll be able to see when a payment has been not retried because of its higher chance of failure from your GoCardless dashboard (see below). This provides an opportunity for you to reach out to the customer and understand when would be a better time to retry the payment or arrange a different way to collect the payment you are due.
- It could save you from paying failure fees:
Some contracts (custom packages with volume-based pricing) will have fees for failures built in. By enabling failure filter you’ll only retry payments that are more likely to succeed hence helping you avoid charges associated with payment failure.
Please note: if you are based in Germany, Austria, or the United States, you may have failure fees as part of your pricing. You can check if these fees apply to you here: https://gocardless.com/pricing
Cons
- There is a chance that you will miss some successful retries:
Enabling failure filter will require you to consider a trade off between removing these payments from being retried and the fact that you may miss a successful retry. Even payments that we predict are very likely to fail can succeed from time to time so there is a small chance of missing a successful payment (or payments).
Why have we chosen a 90% as the threshold for the failure filter?
The algorithm behind intelligent retries, trained on millions of recurring payments processed by GoCardless, can estimate the likelihood of failure for each payment. For example, a likelihood of 90% means that if we retry this payment 100 times, on average it would fail 90 times and 10 times it should succeed. Having carefully evaluated the tradeoff between chances to successfully retry payments, possibility to reduce the time you have to wait as well as the chances of your customers incurring fees for returned direct debits, we have selected 90% as the threshold for the failure filter.