Payment method performance

Payment method performance is the comparison of how each payment type performs at the job of collecting revenue: how often a charge succeeds, how much is recovered when it fails, how long collection takes, and what it costs to process. A subscription business rarely uses one method, so the blended outcome hides large differences between them.

The gaps are real. Credit cards typically fail on 5 to 10 percent of attempts because cards expire, funds run short, and fraud filters block charges. Direct debit fails on roughly 1 to 3 percent because it pulls straight from a bank account with no expiry to manage. A business that knows these numbers can steer customers towards the methods that hold revenue rather than lose it.

The metric also feeds into failed payment rate and involuntary churn. A failed charge that is never recovered is a customer lost to a payment problem, not to a product one. Measuring performance per method is the first step to fixing that leak.

The core figure is success rate per method: successful charges divided by charge attempts, times 100. A method with 9,500 successful charges out of 10,000 attempts has a 95 percent success rate. But success rate alone misses what happens after a failure, so pair it with recovery rate and time to collection.

For example, cards might succeed on 92 percent of first attempts but recover another 5 percent through retries and dunning, lifting the effective rate to 97 percent. Direct debit might succeed on 98 percent first time but recover almost nothing because a bank rejection is harder to retry. The full picture needs both numbers, not just the headline.

1. 1

   Charge attempts per method

   All attempts on each payment type over the period, including automatic retries. This is the denominator for success rate.

2. 2

   Successful charges per method

   Attempts that completed and settled. Settled, not authorised, since an authorisation can still be reversed.

3. 3

   Recovery rate after failure

   The share of failed charges later collected through retries, dunning, or a switched method. This shows how much a failure actually costs.

4. 4

   Cost and speed per method

   Processing fees as a share of value, and the average days from attempt to settled cash. A high success rate can still be expensive or slow to collect.

A single success-rate gap tells you one method beats another, but not why or what to do. A metric tree breaks performance into the drivers behind it, so a falling rate points to a specific cause: more expired cards, weaker recovery, a region with poor card coverage, or a costlier mix creeping in.

Benchmarks vary by region, customer type, and processor, so treat these as reference ranges for recurring subscription billing rather than fixed targets. Cards and direct debit behave very differently, and the right comparison is each method against its own history and the alternatives available to the same customer.

Improving performance is partly about steering customers to reliable methods and partly about recovering more of what fails. The actions below map to the drivers in the tree, from first-attempt success through to the method mix.

1. 1

   Reading the blended rate only

   A single overall success rate hides which method is leaking revenue. Always split performance by method, or a strong direct debit base will mask a weak card flow.

2. 2

   Counting authorisations as success

   An authorisation can still be reversed or fail to settle. Count settled charges, not approvals, or the success rate will read higher than the cash actually collected.

3. 3

   Ignoring recovery

   First-attempt success alone makes cards look worse than they are, because much of what fails is later recovered. Track effective rate after retries and dunning to compare fairly.

4. 4

   Forgetting cost and speed

   A method can succeed often yet cost more or settle slowly. Read processing cost and days to collection alongside success, or you optimise for the wrong outcome.