Interchange isn’t the real cost driver. Fee drift, mis-tiering, and weak data inflate payment costs. Learn how to detect, recover, and cut spend.

Oct 16, 2025 (Last Updated: Oct 28, 2025)

Your “payments are expensive” problem probably isn’t interchange. It’s the quiet stuff: misapplied tiers, wrong MCCs, inconsistent AVS/3DS evidence, and missing L2/L3 data. Those pennies don’t shout; they seep—across statements, fee files, and chargebacks—until your neat net effective rate looks honest but incomplete. Here’s how to see (and stop) that seep without switching PSPs.
Fee drift (the bps you never approved).
Scheme assessments, cross-border adders, and processor surcharges are legitimate—but they often drift from contract logic when geography or currency mixes change. If you only check fees monthly, you miss weeks of small deltas that add up.
Mis-tiering & MCC misclassification.
Your category and tier determine qualification and pricing. A wrong MCC or a contract tier not being applied will quietly lift your effective cost. Networks’ rulebooks make clear that assessments and pricing vary by category/region—so classification accuracy is not optional.
Data quality (AVS/3DS, L2/L3).
Returns & refunds tax.
Refunds aren’t “free.” U.S. retail returns were estimated at $890B (≈16.9% of sales) in 2024, and refunds carry assessment fees, lost credits in some scenarios, and handling cost—all of which inflate your real cost of acceptance.
Think of total cost as five stacked layers; measure each weekly:
1. Core processing
Interchange + acquirer margin + scheme assessments (domestic + cross-border). Keep a geography/currency view so mix-shift costs don’t hide in a blended NER.
2. Authorization tactics
Retries, tokenization, and local routing can lift approvals—but also add network/processing fees. Track net margin per 1,000 orders, not approvals alone.
3. Disputes
Attach unit costs (fees + labor + lost goods) to your win-rates. Rising disputes make evidence quality (3DS, descriptors, delivery/usage logs) a cost lever, not just a risk lever.
4. Refunds/returns
Model fee treatment and ops cost per refund; trend by category to shape policy.
5. Data quality & compliance
Measure L2/L3 qualification rates and the bps penalty when fields are missing. (For B2B, this alone can be the difference between “acceptable” and “excellent” cost.)
Detect (daily, not monthly).
Normalize and reconcile fee events against your contract and network rules every day. Alert when a code, rate, or region deviates beyond your threshold. Optimus makes this straightforward with Payment Reconciliation (line-item fee normalization + exception queues) and a Knowledge Base that shows how to wire sources cleanly.
Validate.
For each anomaly, tie the transaction lineage end-to-end—auth → capture → settlement → funding → fees → refund/chargeback—and confirm whether the variance is contractual (ok) or misapplied (actionable). This is where AVS/3DS, descriptors, and delivery/usage logs matter; they turn “he-said-she-said” into a clear audit trail. See our primer: Why reconciliation is the hidden profit center.
Recover.
Batch your claims with evidence and push through your acquirer’s process. We’ve written about the mechanics and patterns in Fee overcharges: the silent margin killer and When fees go wrong—both focus on turning found bps into cash, fast.
Renegotiate.
Use three months of clean, reconciled evidence to reset tiers, fix MCC, and align cross-border pricing. With facts on the table (by market, by card type, by data quality), your contracts get better—and stay better.
Interchange sets the baseline; execution decides your bill. Tighten tiers and MCCs, clean up AVS/3DS and L2/L3, and run daily fee hygiene on reconciled data. Do that, and you’ll drop basis points on the same volume—no replatform, just better control. If you’re ready to quantify it, start with Optimus’ Payment Reconciliation to reconcile every cent and keep fee drift in check.