How Smart Routing Improved Our Authorization Rate by 12%

When we onboarded a mid-market e-commerce platform processing $40M annually, their authorization rate was sitting at 81%. That number looked normal compared to industry averages, and their previous provider had told them it was "within expected range."

Within 90 days of implementing smart routing, their authorization rate hit 93%. On $40M in volume, that 12-point improvement represented roughly $4.8M in additional approved transactions per year.

Here is what changed.

The starting point

The merchant was processing all traffic through a single acquirer. Every transaction, regardless of card network, issuing country, or transaction amount, went to the same processor through the same connection. Their configuration looked like this:

All transactions → Acquirer A → Authorize → Done

Simple. Reliable. And leaving millions in revenue on the table.

When we analyzed their decline data, clear patterns emerged:

• Cross-border transactions had a 68% approval rate (vs. 89% domestic)
• Amex transactions approved at 74% (vs. 85% for Visa/Mastercard)
• Transactions over $500 declined at nearly double the rate of smaller orders
• Retry-eligible soft declines were not being retried at all

The routing strategy

We connected two additional acquirers and built a routing configuration with three layers:

Layer 1: Geographic routing

Transactions from European cardholders were routed to a European acquirer with local acquiring capabilities. This alone moved cross-border approval rates from 68% to 82% for EU traffic, because the transactions were now processed as domestic rather than cross-border.

Layer 2: Network-specific routing

Amex transactions were routed to the acquirer with the strongest Amex relationship and best negotiated rates. Amex approval rates improved from 74% to 87%.

Layer 3: Cascade retry

When a transaction received a soft decline (processor timeout, temporary issuer unavailability, or generic decline codes that indicate a retry might succeed), the system automatically retried on an alternate acquirer. This recovered an additional 3-4% of transactions that would have otherwise been lost.

Transaction → Route by rules → Acquirer X → Approve/Decline
                                    ↓ (soft decline)
                              Retry → Acquirer Y → Approve/Decline

Results by the numbers

After 90 days of optimization:

• Overall authorization rate: 81% → 93%
• Cross-border approval rate: 68% → 85%
• Amex approval rate: 74% → 87%
• Revenue recovered from retries: ~$1.2M annualized
• Processing cost change: Net decrease of 8 basis points (routing to cheaper acquirers for eligible traffic)

The processing cost decrease was a bonus. The primary goal was authorization rates, but when you can route to the acquirer with both the best approval rate and the lowest cost, you get both.

The biggest unlock was not any single routing rule. It was having the data to see which transactions were failing and the flexibility to route them differently.

What we learned

Three takeaways from this implementation:

1. Your "normal" authorization rate probably is not normal. Industry averages include businesses that have already optimized and businesses that have not. Your rate relative to your potential is what matters.
2. Start with decline data, not routing rules. Analyze your declines first. The data tells you exactly where to focus. Do not build complex routing logic before understanding where transactions are failing and why.
3. Measure continuously. Authorization rates shift as issuer policies change, card portfolios evolve, and seasonal patterns cycle. A routing configuration that is optimal in Q4 might need adjustment in Q1. Review monthly at minimum.

Smart routing is not a set-it-and-forget-it feature. It is a feedback loop: analyze declines, adjust routes, measure impact, repeat. The merchants who treat it as an ongoing practice consistently outperform those who configure it once and move on.