Payment approval rates are subject to manipulation, bias and false interpretations like any other metrics. To avoid being misinformed, as a business, you should be able to benchmark acceptance across multiple payment service providers on your side.
At Ecommpay, we base our decision-making on hard evidence and propagate this approach to our clients. That's why we decided to outline the main data attributes and techniques for anyone striving for representative approval rate benchmarking.
Preparing the data for approval rate benchmarking
Data is the cornerstone for unbiased approval rate analysis. The interpretation, however, could be tricky. It gets even more complicated when you benchmark approval rates, mixing and matching figures from several payment service providers.
If you want to make your analysis clear, concise and usable for further decision-making, stick to the following data attributes.
Consistency
All the data used in the analysis should be consistent. That's why you must pay utmost attention to the card decline messages and codes and uniformly arrange them.
For example, suppose you see one payment service provider having a decline message and code for a failed 3DS authentication or its antifraud system while another doesn't have such a code. In that case, it doesn't mean the latter is not subject to such declines.
Try figuring out every rejection code and message across all the providers you compare. A good approach is to organise codes and messages in the following groups:
- General authorisation declines, e.g. insufficient funds, do not honour etc.
- 3DS-associated authorisation declines associated with the incorrect or incomplete 3DS authentication
- Risk-associated authorisation declines like provider's antifraud system decline, risk thresholds, and AVS decline
- Payment gateway authorisation declines, e.g. the ones associated with the incorrect/incomplete authorisation request.
At the same time, you should also consider the geography of payments. It is always better to benchmark the same regions across multiple payment service providers as the approval rates for the identical merchant category code (MCC) widely differ from country to country.
Volume
Comparing two payment service providers based on 100 operations from the former and 10 000 from the latter will not result in representative benchmarking. Therefore, you should always gather as much transaction data as possible but adjust the volumes before proceeding to the actual analysis.
Veracity
Using accurate and trustworthy data is key to getting actionable insight from your analysis. You should have a clear understanding of every decline reason or code means. In case you encounter aggregated or mapped decline reasons and codes, ask the payment service provider for clarification on the unmapped decline reasons. Otherwise, you will compromise the quality of your research.
Approaching the approval rate analysis
Once the data is collected and refined, you can proceed to the comparative analysis. There are three common methods of doing it that could be combined to get more valuable insight. Let's have a quick look at each one of them.
Gross approval rate
Comparing the gross approval rate is the most general approach you can take. In practical application, you include all the decline reasons in your benchmarking. Taking this approach could be beneficial for understanding the overall performance of the acquirer.
Approval rate by customer purchases
When you focus on getting your customers through the payment process, it could be wise to add up subsequent successful or declined purchases. For instance, if there are five identical payments in a row and only the final one is successful, we count it as a single successful transaction instead of four declined and one successful.
This approach seems rather tricky to implement, but the insights are worth it, as we are getting quite a realistic view of how customers interact with the acquirer. However, you should be sure that this approach fits your business and monetization models focusing on receiving the payments regardless of the number of attempts.
Soft and hard declines
When you seek to understand the share of failed transactions that could come through if you improve your payment flow, soft and hard decline distinction is the way to go.
Divide all the declines into two groups:
- Hard declines that consequent payment attempts can't resolve. For instance, declines associated with pick-up or expired cards and closed accounts.
- Soft declines that could be resolved by additional actions on the customer's side. For example, declines associated with insufficient funds, generic declines from the issuing bank, and declines due to the limits of the anti-fraud filter.
Comparing those groups will show you if you have room to improve your customer service and how clients interact with your checkout page.
Each approach discussed above is helpful, but if you want to maximise the value of your benchmarking insight, you should combine all of them, if possible. This way, you will get a comprehensive picture of the acquirer's approval rates with more data that can be used for further improvement.
Wrapping it up
The more information you factor into your analysis, the deeper understanding you will develop. Drill down approval rate benchmarking by adding data like issuing country, issuing bank, and even BIN if needed. Doing this will reveal the effectiveness of the acquirers you research from the geographical standpoint. If you notice approval issues with a specific issuing bank and/or BIN associated with a particular issuer, we advise asking the acquirer to contact this bank and investigate.
At the same time, don't forget to analyse the approval rates by authentication flow - challenge or frictionless. It may help you to pinpoint flaws in your acquirer's authentication strategies and discuss possible alterations to maximise frictionless transactions.
