Build a model from actual transaction mix

A single blended fee obscures the differences between credit and debit, ticket sizes, channels, and fixed charges. Start with the same reporting period for sales, transaction counts, fees, and refunds. The categories in part one explain why interchange alone is insufficient.

An illustrative comparison

The following inputs are invented for calculation, not published network rates or a UniSight quote. Assume $1,000,000 of monthly card sales across 25,000 transactions, an average ticket of $40.

Illustrative contractCalculationMonthly total
A: 2.0% + $0.10 per transaction + $100 fixed$20,000 + $2,500 + $100$22,600
B: 1.8% + $0.15 per transaction + $200 fixed$18,000 + $3,750 + $200$21,950

Under those assumptions B is $650 cheaper per month, about 2.9% of A's total cost. Changing ticket size, volume, or extra fees can change the result. The example assumes the stated prices include all other charges; actual quotes need explicit coverage and exclusions.

Measure routing rather than promising a percentage

The Federal Reserve publishes covered and exempt debit measures by network. These are reference data, not a guarantee of your processor's rates or the savings available on your mix.

Ask for an eligible-route comparison including interchange, assessments, processor markup, and any minimums. Check whether the routing feature is enabled on your specific hardware and contract. Compare settled costs after implementation against an equivalent baseline.

Keep operational costs in view

  • Review statement descriptions, authorization and settlement charges, reversals, refunds, and chargebacks.
  • Check equipment, gateway, support, and cancellation terms.
  • Confirm how checkout tenders map to settlement reports.
  • Use financial records to reconcile deposits and fees.
  • Separate payment cost from software subscription cost.

Continue: review customer-facing payment policies →

Review customer-facing payment policies

Consider requirements and customer effects before changing checkout.

Continue to Part 3 →