How to optimize payments using Dynamic Routing

In the fragmented e-commerce payments industry, there’s a meaningful number of failed transactions, and prices vary significantly. But what are the factors driving these issues? And how can we use technology to solve this challenge?

Why Transactions Fail

In the e-commerce market, there’s a significant percentage of card transactions that are declined. Four-party schemes like Visa, Mastercard, and China Union Pay work with multiple independent entities both in the Issuing and Acquiring side (e.g. Issuing Banks; Acquirer Banks; Acquirer Processors; etc.). This model unlocked a set of economic incentives for every stakeholder involved, which provided the scale necessary for network effects to kick in and ultimately lead these three companies to dominate the market.

But this growth has come with a cost. Information asymmetry generates authentication/authorization issues, and this leads to transaction rejections. One useful indicator of such phenomena at play is the fact that acceptance rates in Three-Party Schemes like American Express are higher than the ones from Four-Party Schemes. In Three-Party Schemes, the Issuing and Acquiring operations are managed through the same infrastructure: fewer participants lead to fewer communication problems.

Regardless of the Card Network’s scheme structure, Merchant’s business model, Customer’s profile, card type, or any other transaction parameter, our transaction data shows that between 5% and 20% of all e-commerce transactions fail due to a broad set of external factors.

This happens mostly because of the following reasons:

Technical Issues

• A stakeholder involved in the transaction communication is down: these can be Payment Service Providers (PSPs), acquirers, acquirer processors, or any other stakeholder involved in the transaction flow.
• Communications between stakeholders fail: timeouts may occur between any given pair of stakeholders’ systems, which leads to transaction failures.

Risk Assessment

• Acquirer risk parameters: Risk standards are configured on the Acquiring side to ensure compliance with either Card Schemes regulation, or internal risk policies. These standards may include parameters such as transaction amount, issuing country, or currency.
• Issuer risk parameters: risk assessment can depend on different factors such as the Merchant’s business model (MCC). Merchant Accounts configured with a High-Risk Merchant Category Code tend to have lower acceptance rates. They also may depend on the fact that transactions were not put through a sufficiently high level of authentication, as is the case of the 3DS (Three-Domain Secure) protocol.
• Lack of specific input parameters: if authentication inputs are not correctly collected and communicated by the payment gateway, the transaction will not be authorized by the issuer.

Customer Reasons

• Wrong authentication parameters: whether because the checkout process is not explicit enough for the customer, or it lacks fill-in orientations, or it doesn’t have a form validation system to proactively detect invalid data formats, the customer may provide the wrong payment credentials and transactions won’t go through.
• Lack of available funds: one of the most common reasons for transaction failure. We don’t need to go long further into this: no funds available in the customer’s account means no transaction.
• Checkout abandonment: besides reasons like high additional costs, checkout processes can also cause customer churn due to their complexity. When customers are not educated to perform actions like 2-Factor Authentication (2FA) or to finish the checkout process in a redirected webpage, checkout abandonment is the most certain output.

Why Transaction Costs Vary

On top of the acceptance challenges described above, the fragmentation of stakeholders in the payment card industry is an open door for price disparities. These may occur because of multiple reasons:

Negotiation Power

• Economies of scale provide opportunities for negotiation: the bigger the Merchants’ processing volume, the higher their ability to set lower processing fees with Providers.
• Market knowledge fosters competition: Merchants who are more educated in the payments industry — namely on the different stakeholders available throughout the world — can choose the alternatives that will provide them with the lowest transaction cost.

Risk Management

• Different business models represent different risk profiles, which ultimately reflect on processing rates: dispute processes embedded in any Card Network enable transaction reversibility, which implies that any Provider responsible for fund management can risk losing disputed transactions’ amounts. For this reason, Merchant Account opening processes include underwriting measures that account for the Merchant’s transaction history, as well as its business model and any additional information that might lead to a more accurate risk profile. Fund reserves may be placed as a guarantee, and prices are set higher to offset any potential losses.

Value-Added Services

• Cross-Border Processing: PSPs might charge additional fees for Merchants that want to operate internationally. Instead of entering into direct commercial relationships with foreign financial institutions, Merchants use their existing PSPs’ connections to perform funds’ collection, which might add FX Rates and settlement fees to their existing payment cost structure.
• Additional Services: Any other value-added service typically provided by PSPs can have an impact on transaction costs. These could be payouts, split payments, Merchant of Record (MOR) setups, authentication services (e.g. 3DS), Risk services, or Analytics tools.

Transactional Flow

• Different geographic transaction patterns are priced differently: The possibilities created in transactional flows depending on the geography of the Issuer and the Merchant will set distinct costs to transactions set by Card Networks, mostly materialized as Interchange Fees.
• Payment methods differ significantly in pricing: A Credit Card transaction can cost 1% in a common e-commerce purchase but the same transaction might cost 10% to the Merchant if the customer uses, for example, a Paysafecard, a prepaid online payment method based on vouchers.
• Acceptance environments have different prices: In an e-commerce environment — where the payment instrument is not present (Card Not Present) — the authentication method is potentially weaker than the one from an in-store transaction (Card Present). Consequently, there’s a higher probability of fraudulent events, which as seen above, may lead to a higher processing cost. By considering cost as a function of the risk involved in the transactional flow, Providers might negotiate pricing according to the authentication method. For example, transactions authenticated using 3DS, or the more recent EMV Tokenization standard used by mobile wallets such as Apple Pay, Google Pay, and Samsung Pay, might be offered cheaper processing costs.

How to aim for acceptance and cost optimization using Dynamic Routing

As seen above, the fragmentation of the card payment industry presents a set of challenges for Merchants, but it can be seen as an opportunity if we use the right tools.

In the most common setup, Merchants are bound to communicate with a single Provider, having no other Card Network alternative to send transaction messages in case they fail, or in case they’d have lower costs elsewhere.

Flagship Merchants have been increasingly adopting payment orchestration layers like Switch, to decouple their payments infrastructure and get access to the entire payments value chain. By doing this, Merchants can tap the potential of a free market at play, with all the benefits of capitalizing on multiple Providers to optimize their payment operations.

Merchants can leverage this setup by using our Dynamic Routing Application. The purpose of Dynamic Routing is to build upon Switch’s existing channel offer, by offering our clients the ability to route transactions to multiple Providers in real-time based on pre-set rules:

• Fallback rules, which increase acceptance rates to any payments operation by retrying transactions in different payment providers in real-time. For a certain issuing country, merchants can choose to use a provider but, if for any reason, the transaction fails, authentication parameters can be reused and sent to a subsequent Provider without any disruption on the checkout flow. This feature has already helped our clients to recover a combined 28% of declined transactions.

Fallback rule example in issuing country of Brazil

• Split rules allow merchants to divert a percentage of their transaction volume to specific providers. This feature enables them to maintain minimum amounts to get access to special rates, or simply leveraging the ability to rapidly switch traffic to another alternative as a means of redundancy. For example, merchants can set the weight of transactions at 30% on one provider and 70% on another in order to send most transactions to the one that will maximize success.
• Filter Rules allow Merchants to take into account particular pricing structures and route transactions based on specific transaction parameters. Merchants can opt for a specific Payment Service Provider to handle the transaction funds if its value is set above 100$ or if the currency of processing is the USD. Split and Filter rules have led our clients to a 17% reduction in processing rates.

Split rule and Filter rules example

By setting Dynamic Routing rules, Payment Managers can look into the different Channels’ performance on the Switch Dashboard and decide which ones are being more efficient in optimizing acceptance and processing rates. Just like Tokenization, Dynamic Routing plays an important role in maintaining an efficient payment operation, which supports business growth.